1 
INTRODUCTION 
 
1. ADAPTATION 
 
1.1. Adaptation to novel environments 
 
The natural environment is dynamic and constantly changing, yet plant populations 
show an ability to survive and adapt under such varying conditions (Clausen et al., 
1948; Joshi et al., 2001; Hangelbroek et al., 2003; Reed et al., 2003; Eckhart et al., 
2004).  Individuals within populations respond differently to environmental change, 
due to intrapopulation genetic variation.  If the change is a negative one, for example 
a toxicant, the individuals may respond by adapting and thereby increasing their 
fitness, or not adapting and therefore decreasing their fitness or dying.  An 
organism?s fitness refers to its ability to survive and reproduce (Clausen et al., 1948; 
Lopes et al., 2004).  
 
Understanding the variables that affect an organism?s ability to adapt to a changing 
environment is a critical issue in many research fields.  Adaptation has therefore 
been widely studied (Reed et al., 2003).  Variation in the environment that plants are 
exposed to has been shown to be associated with genetic variability.  This indicates 
a genetic basis for the ability to adapt to varying environments (Lesica and 
Allendorf, 1999; Joshi et al., 2001; Kittelson and Maron, 2001; Hangelbroek et al., 
2003; Rossum et al., 2004).  
 
Adaptation is not always genetically based, however, often being associated with a 
high degree of physiological variation within a single genotype.  This is termed 
phenotypic plasticity and is an example of adaptation without genetic variation 
(Lesica and Allendorf, 1999; Joshi et al., 2001; Kittelson and Maron, 2001; Price et 
al., 2003; Eckhart et al., 2004; Ernande and Dieckmann, 2004).  The environment 
only alters the phenotypic expression of the genotypes, for example by increasing 
enzyme activities, and does not alter the genotypes themselves (Lopes et al., 2004).  
This ability may, however, be genetically controlled.  Most highly adaptable plant 
species show some degree of phenotypic plasticity together with a high degree of 
 2 
genetic variation (Lesica and Allendorf, 1999; Joshi et al., 2001; Eckhart et al., 
2004; Ernande and Dieckmann, 2004).   
 
Phenotypic plasticity usually confers fitness advantages, through allowing 
adaptation to harsh environments that would otherwise be uninhabitable (Price et al., 
2003; Ernande and Dieckmann, 2004).  For example, wild plants that were adapted 
to low-nutrient soils and subsequently grown in phosphorus-deficient soils grew 
better and had higher phosphorus levels in their shoots than crop plants grown in the 
same phosphorus-deficient soil.  This was found to be due to a slower growth rate in 
the wild plants, rather than an increased uptake of phosphorus from the soil by these 
plants (Marschner, 1986).  Adaptive plasticity is characterised by phenotypic 
variation in the direction of optimising character traits for the particular 
environment, for example smaller leaves in drier areas (Eckhart et al., 2004).  Non-
 adaptive plasticity may also occur.  For example, plants may show reduced growth 
when fewer nutrients are present, thereby reducing competitive ability as phenotypes 
move away from the optimum for the environment (Hangelbroek et al., 2003; 
Eckhart et al., 2004).   
 
Plant strains that have adapted to a chosen environment outperform strains newly 
transplanted to the same conditions, indicating that adaptation leads to increased 
fitness in the local environment (Joshi et al., 2001; Hangelbroek et al., 2003; Reed et 
al., 2003; Eckhart et al., 2004).  However, adapted genotypes show decreased fitness 
in environments that they are not adapted to, indicating that some degree of 
specialisation is necessary for adaptation to occur (Hangelbroek et al., 2003).  This 
response has also been shown to occur in animals (Reed et al., 2003; Lopes et al., 
2004).  In Drosophila species, populations that were inbred in a variable 
environment were shown to be more adaptable and have higher fitness and stress 
tolerance than those that were inbred in a constant environment (Reed et al., 2003).  
 
Local adaptation to particular environmental conditions may also occur within 
populations, within short distances.  Local adaptation has been reported at small 
geographical scales in response to abiotic factors such as heavy metals, wind, and 
elevation, and to biotic factors such as interspecific competition and herbivory 
 3 
(Hangelbroek et al., 2003).  For example, within-population adaptation has been 
shown to occur in plants that have colonised different levels of sand dunes, as a 
response to varying competition and water-availability levels (Knight and Miller, 
2004).  
 
1.2. How does adaptation occur? 
 
If a population does not adapt to a changing environment, it is likely to become 
extinct.  Therefore what conditions will lead to local adaptation rather than 
extinction?  In many cases, local adaptation follows colonisation of a novel 
environment through dispersal of seeds from surrounding areas (Bush and Barrett, 
1993; Campbell et al., 2003; Price et al., 2003; Jacquemyn et al., 2004).  Different 
types of plants have different potentials for colonisation of a new environment, for 
example those that have wind-dispersed seeds have a higher immigration rate, and 
therefore a higher rate of colonisation, than those that have seeds that are dispersed 
over short distances (Campbell et al., 2003).  The probability of colonisation also 
depends on the distance between the two areas (Jacquemyn et al., 2004).  The prior 
environment experienced by the colonising individuals can have a large influence on 
the ability of plants to adapt to novel conditions (Reed et al., 2003; Jacquemyn et al., 
2004).   
 
Population genetic diversity is a fundamental component of biodiversity and is 
crucial for ensuring the long-term sustainability and survival of a species, as well as 
influencing the rate of evolutionary change (Faugeron et al., 2001; Bagley et al., 
2002; Wimp et al., 2004).  Genetic diversity in plant populations has been shown to 
directly influence diversity in dependent animal species, further emphasising the far 
ranging implications of genetic diversity (Wimp et al., 2004).  The role of genetic 
diversity in determining the ability of plants to adapt to novel environments will be 
discussed further. 
 
After colonisation of a new environment, some plant species form ecotypes, i.e., 
evolve ecological adaptations suited to the new environment, but not reproductive 
barriers (Clausen et al., 1948; Gibson and Pollard, 1988; Fiedler, 1992; Macnair, 
 4 
1992).  If ecotypes evolve further to form reproductive barriers, a new species has 
formed (Macnair, 1992).  These ecological races approach a dynamic equilibrium 
with the new environment, an equilibrium that is maintained by the balance between 
the forces of genetic variation and natural selection (Clausen et al., 1948; Clausen, 
1951).  Genetic diversity in traits that will assist the organism in the new 
environment is therefore vital for adaptation to occur (Clausen et al., 1948; Clausen, 
1951; Bagley et al., 2002; Reed et al., 2003). 
 
Any particular environment contains a dynamic balance between forces that increase 
genetic diversity, primarily mutation and gene flow, and those that reduce it, 
including genetic drift and directional selection (Clausen, 1951; Eckhart et al., 2004; 
Knight and Miller, 2004; Rossum et al., 2004; Sol? et al., 2004).  Gene flow refers to 
the transfer of genes, gametes, or individuals between populations.  The extent of 
gene flow that occurs is influenced by the breeding, pollination, and dispersal 
systems of the species (Lesica and Allendorf, 1999).  A large amount of gene flow 
will reduce or even prevent adaptation through the continuous influx of non-tolerant 
genotypes (Antonovics et al, 1971; Faugeron et al., 2001).    
 
In every population there are random changes in reproductive success from one 
generation to the next.  These changes cause genetic drift: shifts in allele frequency 
from generation to generation.  Genetic drift results in the genes transmitted to 
offspring not being representative of the allele frequency present in the previous 
generation.  Genetic drift becomes greater as the population size decreases due to the 
reduced number of reproductive adults available (Lesica and Allendorf, 1999).  The 
long-term effective population size is therefore important as it determines the levels 
of genetic drift the population is subjected to (Bagley et al., 2002; Reed et al., 2003).  
The effective population size is the size of an ideal population, i.e., one that meets 
all the Hardy-Weinberg assumptions (Storz et al., 2001).  A population in Hardy-
 Weinberg equilibrium is one where random mating occurs and where there are no 
disturbing forces such as selection, mutation, and migration (Weir, 1996).   
 
Genetic drift causes a loss of genetic diversity during colonisation due to the small 
number of individuals present, and later results in changes in allele frequency and 
 5 
causes genetic differences between subpopulations (Lesica and Allendorf, 1999; 
Tribsch et al., 2002; Sn?ll et al., 2004).  These allele frequency changes take the 
form of both a loss of alleles, which is affected by the size of the colonising 
population, and reduced heterozygosity, which is affected by the length of time that 
the population remains small following colonisation (Bush and Barrett, 1993).  Gene 
flow levels also change, becoming restricted due to geographic isolation of the 
colonising population (Bush and Barrett, 1993).   
 
Controversy exists over the relative importance of natural selection versus genetic 
drift and gene flow in the adaptation process (Kittelson and Maron, 2001).  Genetic 
drift and gene flow are not directly influenced by the environment, and therefore 
result in changes in diversity throughout the genome.  These factors therefore 
influence adaptive ability through increasing or decreasing overall genetic diversity.  
In contrast, natural selection influences the selected genome: the diversity of genes 
that are selected by the environment, or genes closely linked to those genes.  Natural 
selection therefore influences adaptation by influencing diversity in genes directly 
responsible for survival in the new environment (Clausen et al., 1948; Clausen, 
1951). 
 
Following colonisation, a combination of directional selection with at least short-
 term opportunity for population growth has been suggested to be necessary for 
adaptation to occur (Reznick and Ghalambor, 2001; Price et al., 2003).  Other 
studies, however, suggest that population growth may be more important, as without 
it directional selection may cause population decline and local extinction (Reznick 
and Ghalambor, 2001).  Natural selection has various possible effects on 
populations.  Strong selection pressures may result in deterioration in some aspects 
of performance, along with an increase in desirable traits.  Rapidly fluctuating 
selection can cause a reduced phenotypic expression of genetic variation (genetic 
canalisation) (Valladares et al., 2002).  Natural selection may also reduce the 
effective population size, increase genetic drift, and cause an increased rate of loss 
of genetic variation (Lesica and Allendorf, 1999; Bagley et al., 2002; Reed et al., 
2003).   
 
 6 
In summary, local adaptation can occur when selection pressures are intense and 
gene flow is of a sufficiently low level that it does not overwhelm selection.  In 
small populations, where gene flow is likely to be significant, local adaptation may 
occur only if the strength of selection is also intense (Hangelbroek et al., 2003; 
Knight and Miller, 2004).  Sharp boundaries can exist between two sites in close 
proximity to each other, for example sand dunes and adjacent grasslands, due to 
dramatic changes in conditions.  These boundaries create strong gradients in 
selection that tend to result in local adaptation (Kittelson and Maron, 2001).  
 
The distribution of genetic variation within and among populations is influenced by 
the characteristics of the species (Hamrick, 1989).  Each plant species has a unique 
combination of life-history traits, historical factors, and habitat preferences that can 
influence patterns of genetic variation within and among populations (Vellend and 
Waterway, 1999).  For example, the mating system of the species determines the 
extent of gene flow in the population, which in turn influences the amount of genetic 
variation within and between populations.  Marked differences in genetic variation 
can therefore occur over short distances in populations, resulting in a non-random 
distribution of genetic variation (Hamrick, 1989).  These forces all interact to 
influence the amount of genetic variation in a population.  Since genetic diversity 
affects the level of adaptation a species is capable of, measuring that diversity 
provides insights into the likely reaction of the species to environmental change 
(Bush and Barrett, 1993; Bagley et al., 2002). 
 
1.3. Studies on adaptation 
 
Local adaptation has most often been demonstrated through reciprocal transplant 
experiments (Joshi et al., 2001; Eckhart et al., 2004).  This technique was used most 
often in older studies, for example in the extensive studies done on Achillea spp. by 
Clausen and colleagues (1948).  These experiments involved growing plants taken 
from varying environments in constant and in differing conditions, and studying 
morphological variation.  Morphological and physiological characteristics directly 
influence the success of plants.  By studying these characteristics in plants obtained 
from different environments that are grown under sets of known conditions, the 
 7 
experimenter can study the physiological adaptations necessary in order for a plant 
to survive in its natural environment (Clausen et al., 1948; Clausen, 1951). 
 
Morphological and physiological characteristics show a high correlation with the 
climate in which the plant population originated.  Characteristics that are shown to 
be inherited and correlated with the environment can be viewed as the 
morphological manifestation of a genetic response to environmental change 
(Clausen et al., 1948; Clausen, 1951).  Physiological variation is found in nearly all 
populations, and typically no single gene accounts for this variation (Hartl and 
Clark, 1989).  Reciprocal transplant experiments do not study genetic variation at 
the molecular level, but rather infer conclusions about the genetic structure of a 
population through studying morphological and physiological characteristics.  These 
experiments therefore view the products of the selected genome, without viewing 
the genes themselves.   
 
Serpentine soils offer restrictions on plant growth due to their unique characteristics, 
for example these soils often contain high levels of nickel and chromium 
(Antonovics et al, 1971).  The serpentine plant Mimulus nudatus shows adaptation to 
drought when compared to the closely related species M. guttatus, from which it is 
thought to have evolved.  Mimulus nudatus was found to grow better on serpentine 
soil than on high nutrient compost.  This was seen as evidence that M. nudatus had 
adapted to grow on serpentine soil (Macnair, 1992).  Generally, species found 
growing both on and off a particular substrate are thought to have undergone recent 
adaptation, while species in which growth is confined to one substrate are thought to 
have undergone adaptation longer ago (Fiedler, 1992).   
 
The earliest studies involving the analysis of genetic variation at the molecular level 
involved the use of allozymes, often coupled with measures of physiological 
variation.  Allozymes are different allelic forms of a protein, detected through 
variation in migration rates (caused by molecular weight, shape, and charge 
differences) during electrophoresis (Weir, 1996).  Spartina patens populations 
seemingly adapted to three different, but adjacent, areas (marsh, swale, and dune) 
were studied.  Differences in morphometric traits, for example seed production, 
 8 
were found between areas.   Physiological differences were also found between 
different areas with regard to salt tolerance, nutrient responses, and drought 
tolerance.  These differences are adaptively significant, allowing the plant to adjust 
to the local environment (Silander and Antonovics, 1979).   
 
In the molecular component of the study, allozyme variation was found between the 
different sites, indicating genetic variability within and among the different 
subpopulations.  Population divergence and differentiation were therefore concluded 
to have occurred, over distances of as little as 200m.  The differentiation observed 
was thought to be due mostly to intense and disruptive selection pressures.  Low 
levels of gene flow, coupled with inbreeding due to a difference in flowering time 
between the areas, could also have promoted subpopulation isolation (Silander and 
Antonovics, 1979). 
 
The serpentine endemic Berkheya rehmanni var. rogersiana was shown to be 
morphologically very similar to the adjacent but non-serpentine Berkheya rehmannii 
var. rehmannii.  Enzyme electrophoresis, however, showed var. rehmannii to be 
genetically variable, and genetically distinct from var. rogersiana.  Unique alleles 
were found in var. rogersiana, providing evidence that negligible levels of gene 
flow were occurring between this variant and var. rehmannii.  This led the authors to 
suggest that these two variants should in fact be different species (Williamson et al., 
1997).  This is an example of the increased levels of genetic variation that can be 
detected with molecular methods when compared to morphological and 
physiological study methods. 
 
Sometimes, however, no evidence of genetic adaptation to a novel environment can 
be found.  Silene dioica plants on serpentine and non-serpentine soils were found to 
differ physiologically, but enzyme electrophoresis showed that the serpentine 
populations did not form a group that was genetically distinct from the non-
 serpentine populations.  It was therefore concluded that there was no evidence of 
genetic adaptation to the serpentine soil at the loci studied (Westerbergh and Saura, 
1992).  This is an example of phenotypic plasticity that does not appear to be 
genetically controlled.  
 9 
1.4. Adaptation to metal-contaminated soil 
 
The colonisation of metal-contaminated soils by plants is a good example of strong 
selection pressures being imposed in a novel habitat (Macnair, 1983; Bush and 
Barrett, 1993).  Heavy metals are toxic to all life forms when present in large 
quantities, causing disease or death (Antonovics et al., 1971).  Some plants are 
capable of growth and reproduction on soils that have been contaminated with heavy 
metals.  These plants either possess or are able to evolve heavy metal tolerance 
(Antonovics et al, 1971; Macnair, 1983; Rossum et al., 2004).   
 
Many species have been shown to be able to develop tolerance to one or more 
metals, with tolerance to multiple metals arising either in separate races or 
simultaneously in the same race (Antonovics et al, 1971).  A few plants show 
constitutive tolerance to heavy metals (Gibson and Pollard, 1988), for example the 
copper mosses, which have even been shown to require high levels of copper for 
optimum growth (Antonovics et al., 1971; Shaw, 1994).  Most plant species, 
however, are constitutionally non-tolerant and instead are capable of evolving 
tolerant races (Antonovics et al., 1971; Gibson and Pollard, 1988).  The ability to 
evolve tolerance varies between individuals within a population.  The level of this 
variation indicates the ability of the species to adapt to a new environment, as if 
variation is high different individuals within a population will be able to adapt to 
different environments (Shaw, 1994).   
 
The evolution of metal tolerant plant populations occurs rapidly.  Most heavy metal 
contaminated mine waste dumps are between 50 and 100 years old, and yet carry 
well-established populations (Antonovics et al, 1971; Bush and Barrett, 1993).  This 
rapid development of tolerance occurs due to the strong selection pressures present, 
which are intense enough to result in adaptation even though gene flow levels are 
often high due to the short distances between contaminated and non-contaminated 
areas (Antonovics et al, 1971; Kittelson and Maron, 2001).  Following colonisation, 
individuals that have the capability to evolve tolerance survive, while those that do 
not die within a short space of time (Bush and Barrett, 1993).  This results in a low 
number of survivors among founders following colonisation, causing a loss of 
 10 
genetic variation due to genetic drift, as discussed earlier (Bush and Barrett, 1993; 
Lesica and Allendorf, 1999; Sn?ll et al., 2004).   
 
Mechanisms of metal tolerance vary depending on the metal/s present, 
environmental conditions, and the organism itself.  Mechanisms that may be 
involved include reducing the uptake of metal ions, removal of metal ions by 
depositing them in a vacuole or removing them from the cell, converting the metal 
into a non-toxic form, increasing the requirement of enzyme systems for metal ions, 
decreasing the permeability of the cell or subcellular units to metal ions, 
hyperaccumulating metal ions in metal storage organs in the leaves which are then 
shed, or chelating the metal to the cell wall (Antonovics et al, 1971; Witkowski and 
Weiersbye, 1998).  In several species, for example in the monkey flower Mimulus 
guttatus, tolerance to a metal, in this case copper, seems to be controlled by one 
major gene system (Macnair, 1983; Macnair et al., 1993; Shaw, 1994).  In other 
species polygenic control is indicated (Macnair, 1983).  Tolerance seems to be 
largely metal specific, with tolerance to one metal not appearing to confer tolerance 
to another (Antonovics et al, 1971; Wu and Antonovics, 1975; Shaw, 1994).   
 
Heavy metal tolerant plants are widely studied as they are of interest for 
phytoremediation, where heavy metal tolerant plants are used to limit erosion on 
toxic soils, thereby preventing the contamination by the metal of surrounding areas 
(Rossum et al., 2004).  The study of species found on contaminated soils is difficult 
as it is often not possible to define the limits of the contaminated area, to assess 
ecological factors that may influence results, or to define the amount of toxicity in 
the area (Antonovics et al, 1971).  Adaptation to heavy metals in populations 
growing on mine waste is therefore of particular interest for study.  These sites have 
very high levels of heavy metal contamination, and are usually made up of toxic 
zones surrounded by relatively non-toxic areas (Antonovics et al, 1971; Shaw, 1994; 
Rossum et al., 2004).   
 
In the original studies of growth on toxic soils done by Bradshaw and colleagues, 
the levels of genetic variation in populations from contaminated and uncontaminated 
sites were not measured directly through molecular methods, but rather indirectly by 
 11 
comparing morphological differences in plants grown in reciprocal transplant 
experiments (Antonovics et al, 1971; Bush and Barrett, 1993).  Plants growing on 
metal-contaminated soils are often characteristic of such soils, usually being smaller, 
more prostrate, and adapted to nutrient shortages and the exposed conditions that 
result from sparse colonisation.  Most of the higher plants found on toxic soils are 
perennial herbs.  Perennial plants colonise novel areas better than annuals due to 
persistence and slower rates of growth, and low-growing plants are better adapted to 
exposed conditions (Antonovics et al, 1971).  
 
Toxin-tolerant ecotypes seem to be highly specialised for growth on toxic soil.  They 
seem to be restricted to toxic soil as they are competitively inferior when grown on 
normal soil, and are rarely found in surrounding non-toxic areas (Clausen, 1951; 
Antonovics et al, 1971).  There is evidence that metal-tolerant plants are 
metabolically different from non-tolerant plants.  They seem to have a higher 
requirement for trace metals, probably due to increased metal inactivation ability.  
Tolerance to any metal also seems to provide plants with increased tolerance to low 
phosphorus or high sulphur levels (Antonovics et al, 1971).  Some species seem to 
be unable to evolve tolerance even though they are abundant in the areas 
surrounding metal contaminated soils.  Certain characteristics may be needed by a 
species to ?pre-adapt? them to toxic soil.  For example, mine wastes contain 
restrictions on plant growth other than heavy metal tolerance, for example low 
nutrient levels.  The possession of tolerance to these factors may assist the 
development of metal tolerance (Antonovics et al, 1971).     
 
In early studies, the presence of tolerance to heavy metals was determined by 
measuring root length of plants grown in solutions with or without the metal of 
interest (Wu and Antonovics, 1975; Macnair, 1983).  This approach was used to 
quantify tolerance to lead, zinc, copper, and nickel in the 1950s and 1960s 
(Antonovics et al, 1971).  Root-length experiments have, however, been shown to be 
inaccurate.  Since root growth is under polygenic control, some of the genes 
influencing it may not be involved in metal tolerance.  This results in inaccurate 
measures of the extent of metal tolerance present in the plant (Macnair, 1983).  
Reciprocal transplant experiments therefore became the preferred method of study.  
 12 
For example, high levels of tolerance to copper and zinc were found in mosses, 
associated with high levels of genetic diversity (Shaw et al., 1987; Shaw, 1994).  
 
The advent of molecular markers based directly on variation in DNA has allowed 
much more in depth studies on genetic variation to be performed.  Polluted soils 
have been shown to have an effect on genetic diversity in plant populations in many 
studies on European plant species.  Studies show discrepancies as to whether 
polluted soils cause a decrease or increase in genetic diversity of plant populations 
growing in them.  In Silene paradoxa populations, chloroplast microsatellites have 
shown that heavy-metal tolerant phenotypes evolve at an unexpectedly rapid rate.  In 
several cases the populations retained a high level of variation, while in a few 
decreased variation was observed (Mengoni et al., 2001). 
 
In Scots pines, heavy metal pollution was found to cause an increase in genetic 
diversity (Prus-Glowacki et al., 1999).  In Pinus sylvestris, however, zinc smelter 
pollution was found to cause a decrease in genetic variation (Prus-Glowacki and 
Godzik, 1991).  Decreased variation was thought to be due partially to an especially 
strong selection pressure found in extreme conditions, as only specialized genotypes 
are able to survive in such conditions (Prus-Glowacki and Godzik, 1995; Mengoni et 
al., 2001).   
 
Research into the toxic effects of heavy metals spans all levels of biological 
organisation, due to the risks to populations and even ecosystems.  Pollutants cause 
changes in the genetic structure of the species.  Studies on both plants and animals 
have shown that distinct changes in gene pools occur following growth in 
contaminated habitats, caused by natural selection for increased tolerance to the 
toxins present.  Physical linkages between genes may result in subsequent changes 
in other traits not directly influenced by selection.  For example, decreased 
competitive abilities have been noted in plant strains that have developed toxin 
tolerance (Shaw, 1994).  The effects of heavy metal pollution on plant populations 
are therefore widespread and far-reaching. 
 
 
 13 
2. MINING 
 
2.1. Mine wastes 
 
Over 40% of the land surface of the earth has been disturbed in some way.  
Although the proportion of this disturbance that is due to mining activity is 
relatively low, consisting of approximately 0.2 Mha in South Africa, this proportion 
is increasing, with a progressive increase in the production of most metals and 
minerals over the last few decades (Cooke and Johnson, 2002).  
 
The effects of mining activities are usually more severe than those caused by other 
forms of land disturbance.  The mining process is made up of extraction, milling, 
processing, refining, and waste disposal.  The production and disposal of waste 
(rock, fine tailings, and slimes) generally cause the most severe and long-lasting 
disturbance as they involve large areas of land, with waste produced by mining for 
heavy metals being the most toxic (Bradshaw, 1995; Leteinturier et al., 1999; Cooke 
and Johnson, 2002).  Mining also destroys ecosystems through their removal or 
burial under waste sites (Cooke and Johnson, 2002).   
 
The metal levels in mine waste tend to be decreasing as a result of improvements in 
mining technology (Cooke and Johnson, 2002).  Modern wastes commonly contain 
approximately 1000ppm (0.1%) of the metal that was extracted.  This number is still 
very high, but is an improvement on the levels of over 1% that can be found in older 
waste (Bradshaw, 1995; Leteinturier et al., 1999).  High toxicity levels mean that 
mining wastes often remain uncolonised and are therefore easily eroded through 
wind and water.  Water courses below mine wastes sites have been shown to be 
contaminated, with alluvial mineralised deposits being found in some parts 
(Antonovics et al, 1971; Leteinturier et al., 1999).  High levels of arsenic, cadmium, 
cobalt, chromium, lead, copper, nickel and zinc have been found in soils downslope 
of gold mine tailings in the North West province, South Africa (Schoeman and van 
Deventer, 2004).   
  
 
 14 
2.2. Slimes dams on the Witwatersrand 
 
Tailings deposits (slimes dams) consist of the pulverised rock slurry left after 
mineral extraction during gold or uranium mining.  On the Witwatersrand alone, 
they cover 8000ha and comprise 3x109 tons of rock slurry.  Slimes dams are divided 
into the slopes of the dams, made up of the slimes themselves, retaining walls, 
which are often partially covered in slimes, and toepaddocks surrounding the dams 
(Figure 1) (Witkowski and Weiersbye, 1998).   
 
Unoxidised slimes
 Oxidised slimes
 Retaining wall
 Toepaddock
  
 
Figure 1:  The different zones that make up a typical slimes dam.  Photographs were 
taken at the South Complex slimes dam near Orkney (North West Province). 
 
Slimes dams are formed by soil and rocks from surrounding areas being used to 
form the initial retaining wall.  Slimes slurry is then piped into this shallow dam. 
The slimes on the outer surfaces dries, allowing the dam to be extended upwards 
(Witkowski and Weiersbye, 1998).  Unoxidised slimes, at the top of the wall, have a 
high pH due to the presence of liming agents added during extraction of the metal.  
On exposure to air, pyrite components in the slimes are oxidised and the slimes 
becomes acidic (Weiersbye et al., 2005).  A second soil wall, the toepaddock wall, is 
made in the grassland a distance from the dam in order to contain runoff, forming 
 15 
the toepaddock (Witkowski and Weiersbye, 1998).  Toepaddock walls are also made 
in the toepaddock perpendicular to the slimes dam in some places.  
 
The retaining wall, toepaddock wall, and toepaddock are the areas of the dam where 
most plant growth occurs.  This is due to soil from the surrounding grassland being 
present in or under the slimes in these areas.  This results in these areas having a 
different and more diverse distribution of particle size, more similar to a typical soil, 
which assists in anchoring seeds so germination can occur.  These areas also have 
increased nutrient levels, and higher numbers of nutrient-cycling microbes.  These 
microbes have evolved tolerance to the relatively acidic and nutrient-poor soil of the 
surrounding areas, and are therefore capable of surviving on the dams.  These areas 
are also not sloped, which allows better water retention and increases the chances of 
seeds germinating (Witkowski and Weiersbye, 1998).   
 
Plant growth is influenced by many soil characteristics.  These include organic 
matter content, pH, nutrient availability, and particle density (Hangelbroek et al., 
2003).  Extensive studies have been done on the slimes dams of the Witwatersrand 
in order to determine the characteristics of the soil in these areas.  The following 
table shows the soil physical and chemical characteristics for the South Complex 
slimes dam at AngloGold?s Vaal River mine (Table 1) (Witkowski and Weiersbye, 
1998).   
 
 
 
 
 
 
 
 
 
 
 
 
 16 
Table 1:  Soil physical and chemical characteristics (mean ? standard deviation) for 
the South complex slimes dam.  Values are means and based on the soil dry mass.  
Old vegetated refers to sites that were vegetated more than 3 years previous to the 
study (Witkowski and Weiersbye, 1998).   
 
Feature Vegetated Oxidised slimes Toepaddock 
pH (KCL) Non-vegetated 
Old vegetated 
3.33+0.28 
3.50+0.60 
3.60+0.10 
3.60+0.00 
Pb (?g/g) Non-vegetated 
Old vegetated 
110+18 
85+2 
41+9 
53+2 
As (?g/g) Non-vegetated 
Old vegetated 
172+52 
103+5 
57+8 
57+3 
Ni (?g/g) Non-vegetated 
Old vegetated 
121+48 
175+34 
75+34 
43+2 
N mineralisation 
(?g/g per 7 days) 
Non-vegetated 
Old vegetated 
6.13+1.37 
-1.74+0.31 
1.34+3.47 
9.01+0.01 
Total N (?g/g) Non-vegetated 
Old vegetated 
174+11 
146+62 
270+23 
124+41 
Total P (?g/g) Non-vegetated 
Old vegetated 
121+15 
100+9 
70+16 
70+11 
 
This compares with average adjacent grassland values, taken 50-100 metres beyond 
the toepaddock wall, in the Klerksdorp area (Vaal reefs and Afrikander leases 
mines) of 23 ?g/g Pb, 12 ?g/g As, and 46 ?g/g Ni.  Grassland nitrogen 
mineralisation values were 18.3 ?g N/g per 7 days.  Average total nitrogen was 774 
?g/g and average total phosphorus was 273 ?g/g in grassland samples.  The 
grassland pH was found to be an average of 6.3.  Mg, Na, Ca, and Cr were also 
found to be higher in the slimes dam areas than in the grassland.  Organic matter 
was much higher in the grassland.  Grassland samples are still affected by the 
slimes, and therefore represent less contaminated soils, rather than uncontaminated 
soils (Witkowski and Weiersbye, 1998).  These general trends persist between 
mining regions, as well as between the mine complexes within those regions, 
although values vary (Witkowski and Weiersbye, 1998).   
 17 
From these values it can be seen that slimes and slimes-contaminated soils are a very 
harsh environment for plant growth.  Although many heavy metals are necessary 
plant nutrients, at high levels they become toxic to both plants and soil microbes 
(Rout and Das, 2003).  It is generally not the total levels of these metals that is toxic 
to plants but the available levels of the metals in solution.  The pH of slimes soil is 
low, due to the oxidation of pyrite to sulphuric acid, and results in increased levels 
of solubilisation of metals.  These metals may therefore be at toxic levels in solution, 
even though the total amounts present in the soil are low (Witkowski and 
Weiersbye, 1998).  Low pH may also severely inhibit nodulation in legumes, 
resulting in a decreased nitrogen cycling ability (Marschner, 1986). 
 
Zinc and aluminium are naturally abundant in most soils.  At high levels, however, 
these elements may become toxic.  Total zinc levels are generally found to be higher 
in the grassland than in the slimes.  The low pH of the slimes soil, however, results 
in such high solubilisation levels that zinc may reach toxic concentrations (~45 ?g/g 
in slimes vs. 0.05-20 ?g/g in grassland soil) (Witkowski and Weiersbye, 1998).  
Zinc is assimilated early during the development of plants and can therefore be 
highly phytotoxic, generally causing stunted growth (Rout and Das, 2003).  
Aluminium also becomes soluble at lower pH levels.  It causes inhibition of root 
elongation, also resulting in growth inhibition (Sledge et al., 2002).  High levels of 
nickel cause stunted growth in plants (Antonovics et al., 1971).  Other components 
of slimes-contaminated soil affect vegetation, for example low levels of nutrients, 
including nitrogen and phosphorus, and high levels of sulphur.  The organic matter 
content of the soil is of extreme importance as it forms stable complexes with metal 
ions, so making them unavailable for uptake by plants (Antonovics et al, 1971).      
 
The concentrations of heavy metals found in mine wastes are therefore sufficient to 
cause growth inhibition or death in plants that lack the necessary mechanisms of 
tolerance to the metals present (Bush and Barrett, 1993).  Individually, these toxins 
and other growth-inhibitory factors are not present at the extreme levels found in 
some natural, non-polluted ecosystems.  It is the fact that so many chemical and 
physical factors act in combination that makes slimes-contaminated soil so 
inhospitable to plant growth (Witkowski and Weiersbye, 1998).   
 18 
High levels of pollutants and heavy metals can also be toxic to microorganisms and 
animals that live on the dams.  For example, cadmium has been found to cause 
lesions in the liver and kidney of mammals (Cooke and Johnson, 2002).  Serious 
disorders and death have been seen in ruminants that graze on forage with elevated 
Mo, Ni, Co, and Se levels.  Therefore even if an area has been successfully 
vegetated, and has formed a stable ecosystem, grazing by animals should be 
prevented until the vegetation has been tested to be safe for animal consumption 
(Witkowski and Weiersbye, 1998). 
     
2.3. Vegetation of slimes dams 
 
The goal of vegetating slimes dams is to assist in some level of restoration of the 
area.  Restoration refers to the process whereby an attempt is made to restore 
ecological integrity to the area.  It involves re-establishing the damaged land?s 
capability to capture and retain resources fundamental to normal ecosystem 
function, including energy, water, nutrients, and species (Bradshaw, 1995; Cooke 
and Johnson, 2002; Campbell et al., 2003).  It is also important to preserve the 
native adjacent communities (Lesica and Allendorf, 1999).   
 
Plants play an important role in changing the nature of toxic habitats.  Plant 
populations that will be viable in the long term will prevent erosion (water and 
wind) through the soil-stabilising effect of root systems, and improve the quality of 
the soil (Witkowski and Weiersbye, 1998; Lesica and Allendorf, 1999).  The 
contribution of organic matter and humus by these plants increases the nutrient 
status of the soil, forms complexes with heavy metals and makes metal ions 
unavailable to plants, and improves soil texture (Antonovics et al, 1971).   
 
If left to natural succession, the development of viable plant populations would 
occur extremely slowly due to the low resilience of ecosystems contaminated by 
slimes (Bradshaw, 1995; Cooke and Johnson, 2002).  Slimes dams generally contain 
no topsoil.  The fine texture of the slimes (mostly silt) and the lack of organic matter 
result in high soil bulk density and soil compaction, low aeration, low water 
infiltration rates, and surface waterlogging (Cooke and Johnson, 2002).  Added to 
 19 
these slimes-specific problems are the fact that Witwatersrand slimes dams are 
generally located in harsh climatic conditions, with low rainfall (600 to 700mm 
mean rainfall in the areas of sampled dams) and an annual range of temperatures of 
?5?C to over 30?C.  These factors make vegetation of slimes dams extremely 
difficult (Witkowski and Weiersbye, 1998).   
 
Attempts have been made to vegetate the slimes dams in the past.  This involved 
liming and fertilising the dams, and introducing mainly northern hemisphere grass 
species.  These efforts were expensive and largely unsuccessful.  The grasses used 
require a continuous supply of large amounts of nutrients and water (Weiersbye and 
Witkowski, 1998).  Vetiver, a grass used in vegetation efforts, needs to be supplied 
with nitrogen, phosphorus, and water, and can only be propagated by breaking up 
clumps and replanting (Knoll, 1998).  Exotic grass species are also likely to compete 
with surrounding indigenous species and affect local ecosystem processes, for 
example fire, as well as causing the loss of locally-adapted genotypes (Lesica and 
Allendorf, 1999).   
 
Lime (calcium carbonate) is added to the soil in an effort to raise the pH.  It does 
not, however, penetrate to the lower levels of the soil (Sledge et al., 2002), and 
together with the addition of fertiliser rapidly destroys any naturally tolerant plant 
and microbial populations already present on the slimes dams and surrounding areas 
(Weiersbye and Witkowski, 1998; Sledge et al., 2002).  Soil microorganisms, such 
as bacteria and fungi, are responsible for most of the biological activity that occurs 
in soil.  Some of the roles of microorganisms include nitrogen cycling, for example 
mineralisation, denitrification, and nitrogen fixation, and the decomposition of 
organic matter.  Fungal hyphae and bacterial and fungal exocellular polysaccharides 
bind soil particles together to form aggregates, thereby stabilising the soil (Haynes 
and Graham, 2004).  
 
In contrast, vegetation of slimes dams using tolerant indigenous species is a more 
cost-effective and ecologically-sound approach to rehabilitation (Antonovics et al, 
1971; Weiersbye and Witkowski, 1998).  Indigenous perennial plants have been 
shown to out-compete and out-persist introduced grass species in vegetation trials 
 20 
(Weiersbye et al., 2005).  Local genotypes have been shown to be useful when the 
disturbance is large, as they will have less effect on the gene pools of local and 
surrounding populations.  The use of indigenous species will also maintain the 
biodiversity of animal and insect species that require native species for food, as well 
as provide specific nesting habitats that may be required by some animals and birds 
(Lesica and Allendorf, 1999).  Locally adapted populations will also contain a great 
deal of adaptive genetic variation correlated with local climatic, topographic, 
edaphic (soil), and biotic differences.  This variation is crucial for long-term survival 
of the plant populations in such harsh conditions (Clausen et al., 1948; Clausen, 
1951; Lesica and Allendorf, 1999).   
 
Indigenous legumes are particularly desirable for vegetation due to their symbiosis 
with nitrogen-fixing rhizobia located in root nodules on these plants.  These rhizobia 
convert atmospheric nitrogen into a usable form, at a rate of up to 400 kg N/ha/year 
(Bradshaw, 1995; Weiersbye and Witkowski, 1998; Morris, 1999; Doignon-
 Bourcier et al., 2000).  Leguminous species should be able to produce all the 
nitrogen needed for their growth, unless nitrogen fixation is inhibited by toxic soils 
(Antonovics et al, 1971).  In addition, many legumes are well adapted to arid 
conditions, and are so successful at nitrogen fixation that they are used in West 
Africa as an alternative to nitrogen fertilisers (Doignon-Bourcier et al., 2000).   
 
The vegetation that has naturally colonised retaining walls and toepaddocks varies 
greatly between dams, as it is influenced by the steepness of slopes, aspect, soil 
physical and chemical characteristics, disturbances, biotic factors, and the overall 
climate of the area.  The slopes of the dams themselves are generally uncolonised 
due to the steep gradient which constrains seed germination and seedling 
establishment (Weiersbye and Witkowski, 1998).  The majority of plants naturally 
colonising slimes dams are semi-woody to woody perennials, even if these types of 
plants are not common in the area.  This suggests that these types of plants are 
naturally more suited for growth on slimes dams (Weiersbye et al., 2005).   
 
 
 
 21 
3. THE AMPLIFIED FRAGMENT LENGTH POLYMORPHISM 
TECHNIQUE 
 
3.1. Advantages of AFLP 
 
Molecular techniques have revealed a higher level of genetic variation in plant 
populations than was previously thought to have existed (Clegg, 1997).  Molecular 
markers have several advantages over the traditional phenotypic markers, that are 
difficult or time-consuming to select by plant-breeders, as they are not directly 
influenced by environmental conditions, they are detectable at all plant growth 
stages and in all tissues, and they have simple inheritance patterns (Lu et al., 1998; 
Bagley et al., 2002; Najimi et al., 2002).  Molecular markers have proved to be 
powerful tools in the assessment of genetic variation both within and between plant 
populations (Muluvi et al., 1999).  
 
Several molecular techniques have been used for the analysis of molecular variation.  
Restriction fragment length polymorphisms (RFLP) are based on detecting variation 
in restriction endonuclease restriction sites caused by base substitutions or small 
deletions/insertions.  This technique requires large amounts of DNA and the use of 
Southern blotting, and is therefore time-consuming and expensive.  Randomly 
amplified polymorphic DNAs (RAPD) detect sequence changes within polymerase 
chain reaction (PCR) oligonucleotide primer binding sites.  These markers tend not 
to be reproducible due to the low specificity of binding of short primers.  
Microsatellites (simple sequence repeats or SSR) are stretches of DNA that consist 
of tandem repeats of a simple sequence of nucleotides, with the number of repeats 
varying between individuals.  They require prior sequence knowledge of the study 
organism, whereas RAPD does not (Hillis et al., 1996).  Although microsatellites 
allow the detection of many alleles per locus, they lack power due to the small 
number of loci analysed in each reaction (He et al., 2004). 
   
The AFLP (amplified fragment length polymorphism) technique is a relatively new 
technique, first described by Vos et al. (1995).  The technique is based on the 
amplification of fragments created by digestion with restriction enzymes.  
 22 
Amplification is carried out through PCR, by which a region of DNA may be 
amplified over a million fold by the use of DNA synthesis initiated by two specific 
primers and carried out by a thermostable DNA polymerase (Vos et al., 1995).  
AFLP is an example of a technique that uses neutral markers.  Neutral markers are 
genetic differences that are assumed to not be affected by natural selection, although 
they may be influenced by selection at linked loci (Faugeron et al., 2001; Sn?ll et al., 
2004).  These types of markers are very powerful in exploring genetic variation 
(Faugeron et al., 2001).   
 
Like other fingerprinting techniques, for example the RAPD technique, AFLP 
requires no prior sequence knowledge of the DNA of interest (Vos et al., 1995; 
Muluvi et al., 1999).  The AFLP technique has, however, many advantages over 
other molecular techniques.  Since it is PCR based, it can be performed with very 
small amounts of DNA (Castiglioni et al., 1999; Najimi et al., 2002).  It is also a 
quick process, providing results in a relatively short space of time (Castiglioni et al., 
1999; Najimi et al., 2002).  AFLP can detect polymorphisms simultaneously in more 
loci of the genome than are inspected by other PCR-based techniques (Thomas et 
al., 1995; Miyashita et al., 1999; Han and Ely, 2002).  This high resolution makes it 
possible to study genetic variation throughout the entire genome (Miyashita et al., 
1999; Gobert et al., 2002).      
 
The AFLP technique is capable of producing a large number of informative bands 
from a single PCR reaction (Castiglioni et al., 1999; Miyashita et al., 1999; Vanhala 
et al., 2004), many more than can be produced with the study of allozymes 
(Terashima and Matsumoto, 2004).  In barley, melon, potato, Arabidopsis thaliana, 
cassava, Eucalyptus, hops, maize, and wheat, AFLP detected a larger number of 
polymorphisms than, for example, RFLP, RAPD and SSR techniques (Miyashita et 
al., 1999; Beardsley et al., 2003; Terashima and Matsumoto, 2004).  
 
In addition, AFLP markers are highly reliable and reproducible due to the stringent 
reaction conditions used during primer annealing (Vos et al., 1995; Jones et al., 
1997; Miyashita et al., 1999; Gobert et al., 2002; Vanhala et al., 2004).  The 
reproducibility of RAPD, AFLP and SSR markers was tested by duplicating studies 
 23 
over time in several different labs.  RAPD markers were difficult to reproduce, 
while SSR markers (microsatelites) showed small size differences on repeat 
analyses.  AFLP markers were found to be highly reproducible, with only a single 
band difference occurring following one repetition of the experiment (Jones et al., 
1997).  
 
3.2. Examples of uses of the AFLP technique 
 
3.2.1. Measuring genetic variation 
 
Variation within and between populations used to be measured primarily by 
pedigree-based measures, or by measuring phenotypic diversity.  Molecular markers 
produce diversity levels comparable to pedigree-based methods (Barrett et al., 
1998), and higher levels than those produced by phenotypic diversity measures 
(Vanhala et al., 2004).  This is assumed to be due to the few traits that are usually 
measured during phenotypic methods, as well as the fact that many phenotypic traits 
are directly influenced by the environment.  Molecular markers are numerous and 
are influenced only indirectly by the environment, if at all (Vanhala et al., 2004).  
 
Since AFLP has the capacity to inspect a much greater number of loci for 
polymorphism than other techniques, it results in the detection of higher levels of 
genetic variation (Thomas et al., 1995; Muluvi et al., 1999; Yee et al., 1999).  In 
Vigna angularis, the mean level of molecular genetic variation detected through the 
use of AFLP was significantly higher than that detected with RAPD (Yee et al., 
1999).  AFLP analysis has been used successfully to assess the amount and pattern 
of genetic diversity in, for example, Brassica juncea (Gobert et al., 2002), Elymus 
repens, an economically important grass species (Szczepaniak et al., 2002), and 
Avicennia marina, a wide-ranging mangrove tree species in Vietnam (Giang et al., 
2003). 
 
The AFLP technique was used in garlic to see if phenotypic differences observed 
between accessions are due to genotypic differentiation or phenotypic plasticity 
(Volk et al., 2004).  It has also been used to study genetic diversity differences 
 24 
between recently founded and established plant populations (Sol? et al., 2004).  In 
Arabidopsis thaliana, nucleotide diversity was estimated from AFLP data by using a 
method discussed by Innan et al. (1999) and was used to construct a phylogenetic 
dendrogram, which was consistent with dendrograms constructed from sequencing 
and RFLP data (Miyashita et al., 1999).  It has been used to study genetic diversity 
in endangered (the wild apple) and widely utilised (Moringa oleifera) species, in 
order to assist in the development of conservation and management programmes 
(Muluvi et al., 1999; Coart et al., 2003).  
 
3.2.2. Studies between species and strains 
 
The AFLP technique has been used to assess genetic differentiation between species, 
strains, and populations.  For example, AFLP has been used to assess distinctness, 
uniformity, and stability of sugar beet varieties (de Riek et al., 2001), and to assess 
the relatedness between inbred faba bean lines (Zeid et al., 2003).  In Saponaria 
pumila, an alpine plant species, AFLP fragments were found in most individuals in 
selected groups but not in members of other groups.  This indicated that the selected 
group originated from a separate, isolated glacial refugium than that the other groups 
originated from (Tribsch et al., 2002).  A similar study was done with the alpine 
herb Rumex nivalis (Stehlik, 2002).  The AFLP technique has also been used to 
determine phenetic relationships among species in Dactylorhiza (Gobert et al., 
2002).  Variation between species of monkey flower was examined in order to 
determine the evolution and origin of hummingbird pollination (Beardsley et al., 
2003).   
 
AFLP band patterns are species-specific and may therefore be used to identify 
closely related species or strains that would be difficult to identify by traditional 
methods (Han and Ely, 2002).  For example, AFLP was used to distinguish between 
wild and cultivated gene pools in chicory (Van Cutsem et al., 2003) and adzuki bean 
(Xu-xiao et al., 2003) in order to study gene flow and strain evolution, respectively.  
Since AFLP detects variation at many loci in the genome, species identification of 
this kind is not based on the presence of one particular allele (Han and Ely, 2002). 
 
 25 
Identification of species or strains by this method also does not involve the use of 
mitochondrial DNA (mtDNA) haplotypes, which means that AFLP can be used to 
identify hybrids, an ability which is not possible with mtDNA analysis (Han and 
Ely, 2002).  Natural, interspecific hybridisation occurs at high frequency in the 
section Mentha.  This hybridisation was detectable through AFLP analysis, and 
AFLP profiles allowed classification that supported the previously established 
taxonomic classification for members of Mentha (Gobert et al., 2002). 
 
3.2.3. Other uses 
 
Genetic linkage mapping is a powerful tool for the localisation and isolation of 
genes controlling both simple and complex traits (Lu et al., 1998).  The AFLP 
technique has been used to construct genetic linkage maps in barley, melon, potato, 
Arabidopsis thaliana (Miyashita et al., 1999), rye (Bednarek et al., 2003), and maize 
(Castiglioni et al., 1999).  Linkage maps are the basis for marker-assisted selection, 
where markers are identified that are closely associated with a desirable trait, for 
example nematode resistance in peaches (Lu et al., 1998).  In tomatoes, AFLP 
analysis was used to construct a detailed linkage map, and subsequently allowed 
identification of DNA markers that are tightly linked to Cf-9, a gene that provides 
resistance to Cladosporium fulvum (Thomas et al., 1995).  This analysis has also 
been used to identify markers closely linked to Hessian fly resistance in wheat 
(Najimi et al., 2002) and powdery mildew resistance in the pea (Coyne et al., 2000).  
 
The AFLP technique has also been used for studies on organisms other than plants.  
In bacteria, AFLP analysis was used in the genus Bacillus to determine relatedness 
between two species (Ticknor et al., 2001), and was used to characterise novel 
rhizobia from agriculturally important, previously un-studied leguminous species in 
Senegal (Doignon-Bourcier et al., 2000).  In fungi, AFLP was used to study 
clustering isolates of Beauveria bassiana (de Muro et al., 2003), to investigate the 
genetic structure of the plant pathogen Phytophthora capsici (Lamour and 
Hausbeck, 2001), and to type strains of shiitake mushrooms (Terashima and 
Matsumoto, 2004).  In fish, a modified AFLP protocol involving the use of one 
 26 
restriction enzyme was used in the study of genetic variation between Morone and 
Thunnus species (Han and Ely, 2002). 
 
In summary, even though it is a relatively new technique, AFLP analysis has been 
used extensively in the study of genetic structure in many different organisms.  In 
plants, the AFLP technique has been widely used with many applications.  In this 
study, the AFLP technique will be used together with morphological studies to 
investigate genetic variation in populations of plants exposed to slimes-
 contaminated soil. 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 27 
4. RATIONALE, OJECTIVES, AND PREDICTIONS 
 
4.1. Rationale for the study 
 
There are a large number of slimes dams in South Africa.  This number can be 
expected to increase in the future, due to the fact that revenue from mining still 
contributes significantly towards the country?s economy.  The presence of slimes 
dams results in unique growth conditions being imposed on the plants growing in 
the vicinity of the dams.  It is necessary to understand the full effect that these 
growth conditions have on plants in order to be better able to plan subsequent use 
and/or rehabilitation of the slimes dams and surrounding areas.   
 
The aim of this study was to determine if plant populations show local adaptation to 
the adverse substrate conditions emanating from slimes dams, by investigating 
genetic and morphological variation between adjacent populations of selected plant 
species growing at different distances in relation to slimes dams.   
 
4.2. Objectives 
 
The broad objectives of the study were as follows: 
- Select suitable plant species to be used in the study.  These species should show 
sufficient population numbers in different areas both close to and far away from the 
dams  
- Perform sampling of leaf tissue for DNA analysis and take pertinent morphological 
measurements of plants 
- Develop an AFLP protocol and perform AFLP analysis on all individuals sampled 
- Statistically analyse the genetic and morphological data obtained.  This analysis 
fell into two categories, namely analysis according to distance from the slimes dam, 
and analysis according to estimates of soil toxicity 
- Interpret the results of the statistical analysis and discuss their biological 
significance.  Discuss whether the species studied show local adaptation to the 
adverse substrate conditions emanating from slimes dams 
 28 
- In light of the study results, make recommendations on the suitability of these 
species as candidates for persistence on slimes dams, and make suggestions for 
further study 
 
4.3. Expectations 
 
It was expected that morphological and genetic differences would be found between 
populations on and off slimes dams.  Lower genetic diversity would probably be 
noted on the slimes-inundated soil, due to relatively recent colonisation of these 
soils by plants from the surrounding areas.  Alternatively, a lack of genetic or 
morphological differentiation between areas could indicate that slimes contaminated 
soils did not have much of an adverse effect on the plants studied, or that 
colonisation occurred long ago and its affects have since been negated, or that gene 
flow levels were high enough between the areas to dilute any effect slimes 
contaminated soil may have had. 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 29 
METHODS AND MATERIALS 
 
1. SPECIES SELECTION 
 
In 2003, two slimes dams were selected for the study: the 26 year old (constructed 
1979) South Complex slimes dam (26? 54.07-54.52?S; 26?44.46-45.38?E) at 
AngloGold Ashanti Ltd?s Vaal River mine, near Orkney, North West Province, and 
the 34 year old (constructed 1971) New North Complex slimes dams (26?25.57-
 25.83?S; 27?20.99-21.60?E) at AngloGold Ashanti Ltd?s West Wits mine, 
Carletonville, North West Province.  These dams were chosen because surveys of 
vegetation (Weiersbye and Witkowski, 2002) and chemical characteristics 
(Weiersbye and Witkowski, 1998) had been performed on these and the surrounding 
dams.   
 
The selection of plant species to be used in this study was performed in January 
2003 at the Afrikander Leases dam, west of Klerksdorp.  Several species were 
sampled on and off slimes dams in order to assess plant distribution and DNA yield.  
The purpose of this was to determine the presence of species on the various areas or 
landscape units of the dams, in order that a suitable species may be chosen for study.  
Samples of young leaves were taken for DNA extraction, in order to determine the 
suitability of the selected species for study using the AFLP technique.  Samples of 
three Indigofera species (Indigofera adenoides, an identified species with grey 
leaves, and an unidentified species with sticky leaves) were obtained from the 
toepaddock and retaining wall, as well as off-site samples approximately 50 metres 
from the dam.  Samples of Asparagus laricinus and a Lippia species were obtained 
from the toepaddock.  Asparagus laricinus samples were obtained off-site, while 
off-site Lippia species samples could not be found.   
 
DNA extraction was performed on all five species as described in section 3.  AFLP 
analysis was then performed for all species using the method described in section 4.  
This initial analysis resulted in two Indigofera species being chosen as suitable for 
further study, due to criteria that will be discussed in the results section.  Indigofera 
 30 
adenoides Baker f., family Fabaceae, was chosen for sampling at the South Complex 
dam, which is dry, and relatively unvegetated.  This species could not be found at 
the New North Complex dam, which is wetter and more vegetated.  Indigofera 
zeyheri Spreng. ex Eckl. & Zeyh, family Fabaceae, was therefore chosen for 
sampling at this dam.  Slimes dam, species selection, and species characteristics are 
discussed further under the results section. 
 
2. SAMPLING 
 
Approximately 20 young leaves for DNA extraction were taken from plants growing 
on several aspects of each slimes dam, both on and off the dam.  This was done to 
provide a sufficient number of samples to allow meaningful analysis of the results.   
An aspect refers to a side of the dam, as each dam was made up of several straight 
walls placed at angles.  All plants that could be located in the toepaddock, retaining 
wall, and toepaddock wall were sampled, due to the small population sizes of the 
selected species in these areas.   
 
Off site samples were taken from beyond the toepaddock wall.  The soil in these 
areas is often contaminated through slimes being blown onto these areas, as well as 
through contaminated ground water.  These areas therefore constitute less 
contaminated areas, rather than uncontaminated areas.  Far off site samples were 
taken from sites far enough away from the dams to be considered relatively 
uncontaminated.  Off site and far off site areas showed larger population sizes of the 
selected species than toepaddock areas.  Therefore plants in these areas were 
sampled by taking approximately ten plants from each population, attempting to 
sample a range of sizes plants from large to small in each population.  
 
In order to perform morphological comparisons, measurements of each plant 
sampled were taken.  These were plant height, canopy width at the widest point, and 
a second measurement of canopy width taken at 90? to the first measurement.  Leaf 
measurements were performed by measuring lengths and widths of five mature 
leaves taken at random from each plant.  Mean leaf length and leaf width values 
were then calculated.  Resolution was to the nearest centimetre for plant 
 31 
measurements, and to the nearest 0.1cm for leaf measurements.  Sampling numbers 
are discussed under the results section.  Global positioning system (GPS) 
coordinates were obtained for approximately every fifth plant. 
 
3. SEED GERMINATION 
 
Ten seed pods were taken from ten plants, selected at random, for each species.  
These were to be germinated so that breeding system studies could be made.  The 
AFLP profile of the offspring was to be compared to that of the parent.  The degree 
of similarity between the profiles can be used to determine whether the species are 
self fertilising only, obligate out-crossers, or a combination of the two.  Seeds were 
germinated by soaking them in a solution of 0.1 mg fungicide (Benlate) per ml and 
then placing them on damp filter paper in the dark at 25?C.   
 
4. DNA EXTRACTION 
 
A sub-sample of 0.1 g of young leaf tissue, which produces the highest yield of 
DNA, was taken from each sample and ground under liquid nitrogen.  DNA was 
extracted using the Qiagen DNeasy? plant mini kit, as this has produced good 
results with plant material in the past, giving a high quality and yield of DNA from 
many different species.  DNA was eluted the first time with 100 ?l of elution buffer, 
and the second time with 50 ?l of elution buffer.  The two products were combined.  
This produced an average DNA concentration of 35 ?g/ml.  In order to determine 
the presence, and approximate the purity, of the DNA, 2 ?l of extracted DNA was 
added to 1 ?l of 6X loading dye (80% w/v sucrose, 10 mM tris-HCl, 10 mM EDTA, 
0.3% bromophenol blue) and run on a 0.8% agarose, 1X TBE (0.089 M tris, 0.089 
M boric acid, 0.002 M EDTA, pH 7.8) gel in 1X TBE.  The concentration and purity 
of a random subset of samples was also checked by spectrophotometry. 
 
 
 
 
 
 32 
5. AFLP ANALYSIS 
 
5.1. AFLP protocol 
 
The principle of AFLP analysis is discussed below, and is summarised in Figure 2.  
The DNA of interest is digested with two restriction enzymes, and adapters 
containing PCR primer sites are ligated onto the fragments.  These fragments are 
then amplified through the polymerase chain reaction (PCR).  A subset of the 
resulting fragments is then amplified further by using primers with a few selective 
nucleotides added, in order to reduce the number of fragments amplified and 
generate a fingerprint with a ?readable? number of bands (Vos et al., 1995).   
 
AFLP analysis was performed as described by Vos et al. (1995), with some changes.  
MilliQ? water (Millipore) was used throughout.  The restriction enzymes MseI (a 
frequent cutter) and ApoI (a rare cutter) were used.  Both were obtained from New 
England Biolabs.  All digestion, ligation, and PCR steps were performed in a 
thermocycler (GeneAmp? 2700, Applied Biosystems).  20 ?l ApoI digestions were 
performed containing 5 U of ApoI and 2 ?l extracted DNA (approximately 70 ng) in 
100 mM NaCl, 50 mM tris-HCl (pH 7.9), 10 mM MgCl2, and 1 mM dithiothreitol.  
Digestions were incubated at 50? C for two hours.  A 10 ?l volume of 5 U MseI in 
100 mM NaCl, 50 mM tris-HCl (pH 7.9), 10 mM MgCl2, and 1 mM dithiothreitol 
was then added to the ApoI mixture and the mixture incubated at 37? C for two 
hours.   
 
 33 
ApoI
 MseI
 ApoI adapter
 MseI adapter
 2. Restriction 
digestion into 
restriction fragments 
of varying sizes
 3. Adapter ligation
 1. DNA extraction
 4. Preselective PCR
 Preselective primer
 Selective nucleotides
 5. Selective PCR
 6. Samples run on 
denaturing 
polyacrylamide gels 
and bands visualised
 through silver 
staining
  
 
Figure 2:  Summary of the AFLP technique as developed by Vos et al. (1995) 
 
 34 
The double stranded adapter sequences used were as follows: 
 
ApoI adapter:                     5?-CTCGTAGACTGCGTACC 
                                                        CATCTGACGCATGGTTAA-5? 
 
MseI adapter:                     5?-GACGATGAGTCCTGAG 
                                                            TACTCAGGACTCAT-5? 
 
Double stranded adapter working solutions were prepared by adding the two 
adapters that make up each pair to adapter buffer (200 mM tris-HCl (pH 7.4), 2 mM 
MgCl2, and 5 mM NaCl) to a final concentration of 5 ?M for the ApoI adapter and 
50 ?M for the MseI adapter.  The mixture was then incubated at 95?C for 5 minutes, 
cooled slowly to room temperature in a thermocycler, and stored at -20?C.  
Ligations of 10 ?l, containing 5 U T4 DNA ligase (Fermentas) and 1 ?l of each 
adapter solution in 40 mM tris-HCl (pH 7.5), 10 mM MgCl2, 10 mM DTT, and 0.5 
mM ATP, were added to the digested DNA mix and incubated at 15? C overnight.  
Digested and ligated solutions were stored at -20? C until needed. 
 
The pre-selective ApoI primer contained no selective nucleotides, while the MseI 
primer contained 1 selective nucleotide.  The sequences used were as follows: 
 
ApoI primer:                 5?-CTCGTAGACTGCGTACCAATT 
MseI primer:                 5?-GACGATGAGTCCTGAGTAA 
 
Pre-selective PCRs of 50 ?l were performed containing 100 ng of each pre-selective 
primer, 0.9 U Taq polymerase (Promega), 0.25 mM of all four dNTPs, and 20 ?l of 
the digested/ligated DNA mixture, in 10 mM tris-HCl (pH 9.0), 0.1% Triton X-100, 
50 mM KCl, and 1.5 mM MgCl2.  The PCR profile used is shown in Table 2.  Pre-
 selective PCR products were diluted ten fold in MilliQ? water and used for selective 
amplification.  
 
 
 
 35 
Table 2:  PCR profile used for pre-amplification of samples. 
 
Number of Cycles Temperature Time 
1 72.0 5.0 min 
 94.0 30 sec 
40 52.0 30 sec 
 72.0 1.0 min 
1 72.0 5.0 min 
    
Various combinations of selective primers were tested in order to determine which 
set gave the most readable, reproducible, and meaningful band pattern.  Selective 
PCR primers chosen contained two selective nucleotides for the ApoI primer and 
three selective nucleotides for the MseI primer.  Only one set of primers was used as 
this was found to produce sufficient numbers of polymorphic bands to allow 
meaningful analysis.  The sequences used were as follows: 
 
ApoI primer:  5?-GACTGCGTACCAATTCC                  
MseI primer:  5?-GATGAGTCCTGAGTAACTC   
  
30 ?l selective PCRs were performed containing 15 ng ApoI selective primer, 30 ng 
MseI selective primer, 0.4 U Taq polymerase, 0.25 mM of all four dNTPs, and 3 ?l 
of the diluted pre-selective PCR product in 10 mM tris-HCl (pH 9.0), 0.1% Triton 
X-100, 50 mM KCl, and 1.5 mM MgCl2.  The PCR profile used for selective 
amplification is shown in Table 3. 
 
 
 
 
 
 
 
 
 
 36 
Table 3:  PCR profile used for selective amplification of samples.  The annealing 
temperature was reduced every cycle for the first 13 cycles. 
 
Number of cycles Temperature (?C) Time 
 94.0 30 sec 
13 65.0, reduced by 0.7 each cycle 30 sec 
 72.0 1.0 min 
 94.0 30 sec 
23 56.0 30 sec 
 72.0 1.0 min 
 
 
5.2. Denaturing polyacrylamide gel electrophoresis 
 
The gels used for fragment separation were 10% denaturing polyacrylamide gels, 
made up of 10% acrylamide, 0.25% methylene bisacrylamide, and 7.5 M urea in 1X 
TBE (0.089 M tris, 0.089 M boric acid, and 0.002 M EDTA, pH 7.8).  Gel mixes 
were made in advance by dissolving the urea in the other ingredients for several 
hours at room temperature, and storing the solution at 4? C for up to one month.  
Gels were cast by adding 300 ?l 10% ammonium persulphate solution and 50 ?l 
TEMED to 50 ml of gel solution (sufficient for two gels) and casting the gels in 18 
cm x 16 cm x 1 mm gel apparatus, adding 10 well, 0.8 cm x 1.5 cm well size combs, 
and leaving the gels to polymerise at 4? C overnight. 
 
The following day, the wells were flushed with 1X TBE to remove unpolymerised 
acrylamide.  Gels were pre-run at 1500 V for 15 minutes to remove the polymerising 
agents TEMED and ammonium persulphate, as these are very strongly charged and 
may therefore interfere with the migration of the DNA fragments.  Following the 
pre-run the wells were flushed a second time.  The entire 30 ?l selective PCR 
reaction product was run on the gel, in order to maximise the visibility of the bands.  
Reaction products were mixed with an equal volume of formamide solution (98% 
formamide, 10 mM EDTA (pH 8.0), 0.1% xylene cyanol as tracking dye) and heated 
 37 
at 90? C for 3 minutes in order to denature the DNA.  Samples were cooled quickly 
on ice before being loaded.   
 
The outer two lanes of the polyacrylamide gels were not used for samples, as they 
tend to run slightly slower due to temperature differences within the gel, creating 
bent lanes (smiling) and making accurate scoring more difficult.  A molecular 
weight marker (GeneRuler 100bp DNA ladder, Fermentas) was run in the first lane 
as a size standard.  Gels were run at 50 V for 15 minutes to allow stacking of the 
fragments, and then run at 500 V for approximately 150 minutes until the tracking 
dye was 1 cm from the bottom of the gel.  Following electrophoresis, gels were 
removed from the plates immediately and stained with silver. 
 
5.3. Silver staining  
 
Several silver staining methods were tested until one was found that allowed 
visualisation of the fainter bands.  The method used was adapted from Blum et al. 
(1987) and is shown in Table 4.  All solutions were made up with MilliQ? water as 
this produced darker and sharper bands, and fewer artefacts, than other types of 
purified water tested. 
 
Gels were visualised on the UVP BioDoc-It? system.  Bands were scored directly 
from the viewing screen, as this provided the best band intensity and clarity.  Bands 
were identified both by being compared to the molecular weight marker, and by 
being identified from a reference sheet developed that diagrammed all of the bands 
obtained, each labeled with a number.  This facilitated easier and more accurate 
scoring.  To test for reproducibility, a subset of 16 samples was randomly chosen 
and analysis was repeated, including DNA extraction, AFLP analysis, and band 
scoring.  This has been suggested to be a necessary test for reproducibility in some 
studies (Zeid et al., 2003).   
 
 
 
 
 38 
Table 4:  Silver staining protocol, adapted from Blum et al. (1987).  The original 
protocol has been shown in italics where steps have been modified.   
 
Steps Solutions Time of treatment 
1. Fix 50% methanol, 12% acetic acid 1.5 hrs to overnight (>1 hr) 
2. Wash 50% ethanol 3 x 20 minutes 
3. Pretreat 0.02% (W/V) Na2S2O3.5H20 1 minute 
4. Rinse Water 1 x 5 sec, 2 x 20 sec (3 x 20 
sec)1 
5. Impregnate 0.2% (W/V) silver nitrate, 0.075% 
(V/V) formaldehyde 
20 minutes in the dark 
6. Rinse Water 2 x 5 sec (2 x 20 sec)1 
7. Develop 6% (W/V) sodium carbonate, 
0.0008% (W/V) Na2S2O3.5H20, 
0.05% (V/V) formaldehyde 
2 hrs in the dark (>10 min 
in dark)2 
8. Wash Water Not performed (2 x 2 min)2 
9. Stop Fix solution Not performed (10 
minutes)2 
10. Wash/store 50% methanol Overnight (>20 min)3 
 
1
  The duration of rinses was decreased to retain more pre-treatment and impregnate 
solutions in the gels, thereby increasing band staining.  This process also increased 
background staining, but not to an unacceptable level. 
2
  The gels were developed for longer and the development reaction was not stopped.  
This ensured that bands developed to their maximum and gave the clearest results. 
3
  The gels were scored after storage overnight at 4? C in 50% methanol.  This shrank 
the gels, thereby concentrating the bands, making them easier to view. 
 
 
 
 
 
 
 39 
6. DATA ANALYSIS 
 
6.1. Morphological analysis 
 
6.1.1. Plant and leaf measurements  
 
Measurements taken for each plant included plant height (Ht), canopy diameter at 
the widest point (CD), canopy diameter at 90? to the first measurement (CD90?), leaf 
length (LL), and leaf width (LW).  These measurements were used to calculate 
several measures of overall plant size, leaf size, and leaf shape.  Selected variables 
that did not follow a normal distribution were converted to log10 values, in order to 
normalise the distributions of the variables and facilitate more accurate statistical 
analysis.  
 
Canopy area (ellipsoid shape) was calculated according to the formula: 
  
2
 90
 2
 0CDCDCA pi=  
 
The canopy volume (CV) measurement calculated estimates the ellipsoid volume the 
plant fills.  It was calculated according to the formula: 
 
23
 4 HtCACV =  
 
Leaf area (LA) (ellipsoid shape) was calculated according to the formula: 
 
22
 LLLWLA pi=  
 
 
 
 
 
 
 40 
The leaf length to breadth ratio was determined.  This was termed leaf shape (LS) as 
it gives an indication of the overall shape of the outline of the leaf.  The higher the 
value, the longer and thinner the leaf shape.  LS was calculated according to the 
formula: 
 
LW
 LLLS =  
 
The distributions of these measurements were plotted and compared between areas.  
Correlations between various variables were also determined and compared between 
sites. 
 
6.1.2. Statistical analysis 
 
Original measurements (Ht, CD, CD90?, LL, LW) and calculated plant 
measurements (CA, CV, LA, LS) were used for statistical analysis using the SAS 
system (SAS Institute, 2000).  Several techniques were used to compare plant 
measurements between areas.  This involved firstly comparing each measurement 
between areas, and then testing composite variables as a whole.   
 
Analysis of variance (ANOVA) was performed as a necessary precursor to Tukey 
and canonical discriminant analysis, in order to determine if differences in mean 
values for each variable are significant between groups.  The null hypothesis: 
 
H0:  ?1  =  ?2  = ? =  ?k 
 
is tested, where k is the number of groups being analysed.  Equal sample sizes are 
not necessary for ANOVA tests.  The test is performed by calculating the ratio of 
group mean squared deviations from the mean (MS) to error MS values.  If this 
value is greater than, or equal to, the critical value for the test, the null hypothesis is 
rejected and the means are shown to be unequal (Zar, 1984). 
 
 
 41 
The Tukey test allows further examination of the differences in means detected 
through ANOVA tests, through testing multiple pairs of means.  This allows a study 
of between which groups the detected mean differences are significant. 
 
It considers the null hypothesis: 
 
ABoH ?? =:  
 
versus the alternate hypothesis: 
 
ABAH ?? ?:  
 
where the subscripts denote any possible pair of groups. 
 
Pairwise mean differences are calculated, and divided by SE to produce a q value. 
 
???
 ?
 ???
 ?
 +=
 BA nn
 sSE 11
 2
 2
  
 
where s2 is the error mean square from the analysis of variance and n is the number 
of samples in groups A and B.  If the calculated q value is equal to or greater than the 
critical q value, the null hypothesis is rejected (Zar, 1984).   
 
A two way comparison T-test was used to detect differences in variable means 
between I. zeyheri groups.  This test, as in the ANOVA test, compares the null 
hypothesis:  
 
ABoH ?? =:  
 
versus the alternate hypothesis: 
 
ABAH ?? ?:  
 42 
A t value for testing the hypotheses is determined by the difference between the two 
means over the standard error of the difference between the means.  The null 
hypothesis is rejected if t is either greater than, or equal to, the critical t value, or if 
the t value is less than or equal to the negative critical t value (Zar, 1984). 
 
Canonical discriminant and discriminant analyses are techniques that allow the 
simultaneous comparison of all variables and all groups, thereby exposing which 
groups are significantly morphologically differentiated from each other.  Canonical 
discriminant analysis determines whether one set of variables, taken as a whole, is 
different from another set of variables, and if so, discovers which variable/s is/are 
responsible for the difference.  The analysis is based on the canonical correlation, 
which is the maximum correlation between linear functions of the two sets of 
variables (Blackith and Reyment, 1971; Cooley and Lohnes, 1971).  Variables that 
are shown not to account for significant variation indicate that the differences among 
group means for that variable are small in relation to the differences among 
individuals within a group (Blackith and Reyment, 1971).   
 
Canonical discriminant analysis was also used to calculate a distance matrix based 
on pair-wise squared distances between groups.  This was used to construct a 
dendrogram using the KITSCH programme in PHYLIP (Felsenstein, 2004) and 
viewed using TREEVIEW (Page, 1996).  The KITSCH programme uses a method 
of constructing rooted trees similar to the method used by the programme UPGMA.  
An evolutionary clock is assumed (expected amount of evolution in any lineage is 
proportional to elapsed time).  It is more accurate than UPGMA but is not often used 
due to it being more computationally intensive and therefore time consuming.  
 
Discriminant analysis allows the allocation of individuals to specific groups based 
on a set of variables for those individuals.  The approach involves locating a line in 
the space of the attributes of the individual plants for which the separation of the 
groups is optimised when the individual points of the different groups of plants are 
projected onto it.  A test of the statistical significance of the separation of the groups 
is then performed (Cooley and Lohnes, 1971). 
 
 43 
6.2. AFLP analysis 
 
6.2.1. Scoring of gels 
 
Any position on the gel which contained at least one band was termed a fragment.  
Monomorphic fragments are those that contained a band in each sample, while 
polymorphic fragments lacked a band in one or more samples.  AFLP profiles were 
scored by determining the presence (1) or absence (0) of bands between 
approximately 50-75 and 1500 base pairs in size.  Bands smaller than 50 base pairs 
were not scored as they were found to have no detectable variability.  This is 
presumably because of the amplification of large numbers of fragments of a very 
similar, small size.  These bands will run close to each other under electrophoresis, 
causing them to appear as one band and be scored as such.   
 
6.2.2. Statistical analysis 
 
AFLP data were analysed both within and among sample groups.  The number and 
percentage of loci that are polymorphic regardless of allele frequencies was 
determined for each group using POPGENE software v1.31 (Yeh et al., 1997).  
POPGENE software was also used to calculate Shannon?s information index (I) and 
Nei?s (1973) gene diversity statistic (h) for each group, as well as Nei?s (1973) gene 
diversity statistics for the overall population.   
 
Shannon?s information index (I) provides an estimate of the relative degree of 
genetic variation within each population.  Its advantage is that it is relatively 
insensitive to skewing of data that occurs due to the inability of AFLP analysis to 
detect heterozygotes (Jover et al., 2003). 
 
It is calculated for each locus by the formula: 
 
??= ii ppI 2log  
 
 44 
where pi is the frequency of the presence/ absence of the band at each locus.  The 
group mean diversity is estimated by averaging the I values obtained for all loci in 
that group (Lynch and Milligan, 1994). 
 
Nei?s (1973) gene diversity statistics were calculated.  These included the measure 
of gene diversity within each group (h), similar to I, the total genetic diversity of the 
entire population (HT), and the amount of total genetic diversity that is due to 
diversity within the different subpopulations (groups) (HS) (Nei, 1973).  HT indicates 
what the heterozygosity of the population would be if all subpopulations were 
pooled together and mated randomly (Hartl and Clark, 1989).  HT is calculated by 
the formula: 
  
HT = HS  + DST 
 
where DST  refers to the average gene diversity between subpopulations.  Therefore 
the gene diversity of the total population is made up of the gene diversities within 
and between subpopulations (Nei, 1973).      
 
Wright?s fixation index (FST) was measured using AFLP-SURV (Vekemans et al., 
2002).  FST measures the proportion of the total gene diversity that occurs among as 
opposed to within the groups (Vekemans et al., 2002).  It indicates the reduction in 
heterozygosity that has occurred in a group due to random genetic drift.  This value 
will equal zero if all groups are in Hardy-Weinberg equilibrium with the same allele 
frequencies (Hartl and Clark, 1989).  FST is calculated by: 
 
FST = (HT ? HS)/HT 
 
A permutation test for genetic differentiation among groups was also performed.  
The null hypothesis for this test is that there is no genetic differentiation among the 
groups.  FST values are calculated repeatedly after each random permutation of 
individuals.  The number of permutations to be performed was set to 1000, allowing 
significance testing at the 1% level.  The distribution of the set of values of FST 
obtained by permutations is then compared to the observed FST value for the groups 
 45 
being tested.  If the observed FST value is higher than the value of FST lying at the 
1% rightmost part of the distribution, the null hypothesis is rejected, and it can be 
concluded that the groups being tested are more genetically differentiated than 
random assemblages of the individuals making up those groups (Vekemans et al., 
2002).   
 
An estimate of the amount of gene flow between groups (Nm) was determined by 
POPGENE.  Nm is a measure of the number of migrants per generation and is 
represented by the formula: 
 
Nm = (1/GST -1)/4 
 
where GST is the multiple allelic form of FST (Nei, 1973; Yeh et al., 1997) 
 
Nei?s genetic distance (D) was calculated in order to construct distance matrices of 
pairwise genetic distances between groups.  The probability that a randomly chosen 
allele from each of two different groups will be identical, relative to the probability 
that two randomly chosen alleles from the same group will be identical, is 
determined.  A lack of divergence of two groups will result in a genetic distance 
between them of zero (Hartl and Clark, 1989).  Distance matrices produced during 
this analysis were used to construct a dendrogram using the PHYLIP KITSCH 
programme (Felsenstein, 2004), which was subsequently viewed in TREEVIEW 
(Page, 1996).  
 
Group allocations were performed where individual samples were assigned to a 
specific group based on an individual?s similarity to the individuals making up each 
group.  This procedure gives an indication of the degree of genetic distance between 
individuals and groups, as well as an indication of gene flow between groups.  Low 
percentages of correct assignment will be seen if high levels of gene flow occur 
between groups.   
 
Population assignment was performed through the use of the programme AFLPOP 
(Duchesne and Bernatchez, 2002).  This involves the allocation of individuals to 
 46 
populations based on a minimal log likelihood difference measure.  This value was 
set to 0 due to the close geographic proximity of the groups, as well as a lack of 
definite boundaries between the areas.  With this measure an individual is assigned 
to a population as soon as the likelihood of an individual belonging to a group is 
higher than the likelihood of it belonging to any other group (Duchesne and 
Bernatchez, 2002).   
 
The correlation between distance matrices obtained from genetic and morphological 
data was evaluated and tested using a Mantel test.  This analysis was performed 
using GenAlEx V5 (Peakall and Smouse, 2001).  The correlation rxy between 
morphological and genetic pair-wise distance matrices is determined.  A probability 
(p) is determined through random permutation of the distance matrices in order to 
test the significance of the correlation (Peakall and Smouse, 2001; Szczepaniak et 
al., 2002).  The number of permutations performed was set to 999.  
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 47 
RESULTS 
 
1. SPECIES SELECTION AND DNA EXTRACTION 
 
DNA was extracted from the five species from the three different genera sampled 
during the initial species selection process.  Genomic DNA obtained appeared to be 
of high quality and was not degraded (Figure 3).  The highest DNA yield was 
obtained for Asparagus laricinus and Indigofera adenoides.  The yield for a second 
unidentified Indigofera species was very low (lane 5).  This may be due to the 
?stickiness? of the leaves of this species, which could have prevented effective 
homogenisation of the tissues.  This species was therefore excluded from further 
analysis.   
 
           1       2      3      4       5      6 
 
 
Figure 3: An 0.8% agarose gel showing genomic DNA for selected plant species.  
The sharpest and most intense band (lane 6) indicates DNA that is both of a high 
concentration and not degraded.  Lane 1: Asparagus laricinus, Lane 2: Lippia 
species, Lane 3: Indigofera speceis (grey), Lane 4: Molecular weight marker, Lane 
5: Indigofera species (sticky), Lane 6: Indigofera adenoides.  
 
The Lippia species was also discarded due to the fact that plants of this species 
could not be found off site during the initial sampling process.  AFLP analysis was 
therefore performed on Asparagus laricinus and the two remaining Indigofera 
 48 
species.  All three were found to produce a clear band pattern with a readable 
number of bands, with Indigofera adenoides producing the largest number of bands.   
 
Asparagus laricinus was rejected as a study species due to the fact that seeds of this 
species are likely to be bird dispersed.  This would increase the rate of gene flow 
between on and off site areas dramatically, thereby reducing the chance of local 
ecotypes forming in response to slimes-associated contamination.  Initial DNA 
extraction and AFLP analysis therefore resulted in Indigofera adenoides being 
chosen as the study species.  This species could not be found during subsequent 
sampling at the New North Complex slimes dam.  Indigofera zeyheri was therefore 
chosen for sampling at this site.  Indigofera is one of the largest and most 
widespread genera in South Africa (Allen and Allen, 1981).  
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 49 
2. SAMPLING 
 
2.1. Indigofera adenoides 
 
Sampling of I. adenoides was performed in late December 2003 at the South 
Complex slimes dam, Vaal River Mine (Figures 4 and 5).   
 
 
 
Figure 4:  Layout of the areas surrounding the South Complex slimes dam (indicated 
by a red circle), including the Vaal River, other slimes dams in the area, and the 
location of the RG Williams Game Reserve.  Shaded areas indicate rock dumps and 
the town of Orkney. 
 50 
  
 
Figure 5:  The South Complex slimes dam (top left corner of the photograph) and 
surrounding areas (a), and the South Complex dam (b), at the Vaal River mine, near 
Orkney, North West province.  The indicated scales are approximate. 
 51 
Indigofera adenoides is a perennial, woody shrublet with a spreading growth habit.  
Stems and pods are both covered throughout, but not densely, in glandular hairs.  
Seed pods are many-seeded, flattened, and cylindrical (Milne-Redhead and Polhill, 
1952; van Wyk and Malan, 1998).  This species has compound leaves, each 
consisting of three to six pairs of opposite leaflets.  Small, singular flowers are 
produced, with a pink corolla and petals of 2-4 mm long (Figure 6) (van Wyk and 
Malan, 1998).  The root system is made up of a long tap root with comparatively 
fine lateral roots.  Roots are nodulated, mostly within the first 10 cm of the soil 
(Allen and Allen, 1981).  The species is likely to die back with fire and re-sprout 
from the root stock.   
 
A total of 144 I. adenoides samples were obtained, consisting of 24 from three 
aspects of the retaining wall and toepaddock wall, 54 from 6 aspects of the 
toepaddock, 22 from 6 aspects off site, 25 from far off site A, approximately 500-
 1000 metres away from the slimes dam, and 20 from far off site B (Figures 7 and 8).    
Each side of the dam is referred to as an aspect.  Far off site B is located in the RG 
Williams Game Reserve, Orkney, approximately eight kilometres away from the 
studied slimes dam and at least two kilometers away from any slimes dam.  Far off 
site A can therefore be considered to have a low level of slimes-associated 
contamination, while far off site B can be considered relatively unaffected by 
slimes-associated contamination. 
 52 
(a)
 (b)
 10 cm
 2 cm
  
 
Figure 6:  The characteristics of Indigofera adenoides, including the overall 
structure and spreading nature of the plants, which was more pronounced in 
individuals on the toepaddock (a), and a close up of the leaves and flowers (b).  The 
individual photographed is growing on the toepaddock of the South Complex slimes 
dam, Vaal River mine, near Orkney, North West province.    
 53 
Slimes dam site
 Far off site A
 Far off site B
 26
 ?
 58
 .
 71
 3?
 S
 26
 ?
 53
 .
 50
 7?
 S
 26?41.762?E 26?46.912?
 26
 ?
 58
 .
 71
 3?
 S
 26
 ?
 53
 .
 50
 7?
 S
 0m 5000m
 26
 ?
 58
 .
 71
 3?
 S
 26
 ?
 53
 .
 50
 7?
 S
 26
 ?
 58
 .
 71
 3?
 S
 26
 ?
 53
 .
 50
 7?
 S
  
 
Figure 7:  The relative locations of the three Indigofera adenoides sampling areas 
near Orkney, North West province.  Note this is a straight plot of GPS coordinates, 
and is therefore inaccurate with regard to geographic distance due to longitude 
warping.  The indicated scale is approximate.  Figure constructed in SBI, a GIS 
programme written for this study by L. A. du Pisani (2005) and obtained through 
personal correspondence.  
 
 
 54 
Slimes dam
 Retaining wall
 Toepaddock wall
 Toepaddock
 Off site
 Far off site
 Legend
 26?44.444?E 26?45.414?E
 26
 ?
 55
 .
 02
 0?
 S
 26
 ?
 53
 .
 62
 7?
 S
 26
 ?
 55
 .
 02
 0?
 S
 26
 ?
 53
 .
 62
 7?
 S
 0m 500m
 26
 ?
 55
 .
 02
 0?
 S
 26
 ?
 53
 .
 62
 7?
 S
 26
 ?
 55
 .
 02
 0?
 S
 26
 ?
 53
 .
 62
 7?
 S
  
 
Figure 8:  Locations of sampled Indigofera adenoides plants on the slimes dam and 
far off site A areas.  Each point refers to approximately three plants, and each point 
is colour-coded according to the area the plants were sampled from.  Note this is a 
straight plot of GPS coordinates, and is therefore inaccurate with regard to 
geographic distance due to longitude warping.  The indicated scale is approximate.  
Figure constructed in SBI, a GIS programme written for this study by L. A. du 
Pisani (2005) and obtained through personal correspondence.  
 55 
Indigofera adenoides plants obtained from the toepaddock were generally found 
growing in the uncolonised, slimes-inundated soils, rather than amidst other plants 
on the slightly higher, less slimes-inundated areas.  In contrast, off site samples were 
found growing amongst other vegetation.  The vegetation cover on the toepaddock 
of the South Complex dam averaged approximately 50% overall, with the slimes 
inundated areas averaging approximately 30%.  The level of overstorey shading was 
very low due to the general absence of tall shrubs and trees (Figure 9).  Differences 
in flowering time seemed to be evident between the sites.  Toepaddock plants had 
fully formed, mature seed pods and an absence of flowers, while off site plants had 
mature flowers and a few immature seed pods.  Vegetation cover at far off site B, in 
the RG Williams Game Reserve, was approximately 80%, with little overstorey 
shading (Figure 10). 
 
 
 
Figure 9:  Sampling of Indigofera adenoides occurring on the toepaddock of the 
South Complex slimes dam, Vaal river mine, near Orkney, North West Province.  
The spreading growth habit of this species can be seen, as well as its preference for 
growing in relatively uncolonised, slimes-inundated areas of the toepaddock.    
 56 
 
 
 
Figure 10:  The topography and vegetation cover of the area of the RG. Williams 
Game Reserve, Orkney, North West province, used for sampling of the far off site B 
group.   
 
 57 
2.2. Indigofera zeyheri 
 
Sampling of I. zeyheri was performed in January 2004 at the New North Complex 
slimes dam, West Wits mine, near Carletonville (Figure 11).  Indigofera zeyheri is 
taller and has a sparser crown than I. adenoides.  It is a slender, sparsely branched 
perennial herb that appears greyer in colour than I. adenoides.  The species has an 
erect growth habit and is covered throughout, but not densely, in hairs (Milne-
 Redhead and Polhill, 1952).  Leaves are compound, consisting of five to seven pairs 
of opposite leaflets.  Flowers are found on inflorescences at the end of long axes, 
and are small and pinkish to white in colour (Figure 12) (van Wyk and Malan, 
1998).  The seed pods of this species are flattened, cylindrical, and many-seeded 
(Milne-Redhead and Polhill, 1952).  The root system is made up of a long tap root 
with comparatively fine laterals.  Roots are nodulated, mostly within the first 10 cm 
of the soil (Allen and Allen, 1981).  It is likely that the species dies back during fire, 
subsequently reprouting from the root stock. 
 
A total of 98 samples were obtained for this species, 63 from 4 aspects of the 
toepaddock and 35 from 5 aspects off site (Figure 13).  Retaining wall samples 
could not be obtained as the retaining wall had been completely inundated with 
slimes.  Despite a considerable effort to find suitable populations, the species could 
not be found far enough away from the dam to allow sampling of a far off site 
population.  As was noted with I. adenoides, flowering time differences seemed to 
be apparent between the sites.  Toepaddock plants had fully formed, mature seed 
pods and an absence of flowers, while off site plants had mature flowers and a few 
immature seed pods. 
 
Vegetation on the New North Complex slimes dam is much denser than vegetation 
on the South Complex dam (approximately 70% aerial cover on the toepaddock and 
90% off site).  This could be due to the climate in the area, which appears to be 
cooler and wetter than that at the South Complex dam.  Indigofera zeyheri samples 
were not found growing on slimes inundated soils, again in contrast to I. adenoides.  
Toepaddock plants were found growing only on the highly colonised, relatively 
slimes free, higher areas formed by low banks of grassland soil.  
 58 
 
 
Figure 11:  A comparison of a map (a) and an aerial photograph (b) showing the 
New North Complex slimes dam and surrounding areas, at the West Wits mine, near 
Carletonville.  The circled area in figure (a) indicates the area where sampling of 
Indigofera zeyheri plants occurred.  Indicated scale bars are approximate. 
 
 59 
10 cm
  
 
Figure 12:  The general morphological structure of Indigofera zeyheri.  The 
individual shown was growing on the toepaddock at the New North Complex slimes 
dam, West Wits mine, near Carletonville.  
 
 60 
Legend
 Toepaddock
 Off site
 27? 20.953? E
 26
 ?
 25
 .
 90
 4?
 27? 21.665?
 26
 ?
 25
 .
 50
 0?
 ?
 26
 ?
 25
 .
 90
 4?
 26
 ?
 25
 .
 50
 0?
 ?
 0m 500m
 26
 ?
 25
 .
 90
 4?
 26
 ?
 25
 .
 50
 0?
 ?
 26
 ?
 25
 .
 90
 4?
 26
 ?
 25
 .
 50
 0?
 ?
  
 
Figure 13:  Locations of sampled Indigofera zeyheri plants in the toepaddock and off 
site areas.  Each point indicates three to ten plants, and each point is colour-coded 
according to the area the plants were sampled from.  Note this is a straight plot of 
GPS coordinates, and is therefore inaccurate with regard to geographic distance due 
to longitude warping.  The indicated scale is approximate.  Figure constructed in 
SBI, a GIS programme written for this study by L. A. du Pisani (2005) and obtained 
through personal correspondence.  
 
 
 
 
 
 
 
 
 
 
 
 
 
 61 
3. SEED GERMINATION 
 
The percentage of seeds that germinated was approximately 50%, which is 
consistent with that obtained in previous studies (Weiersbye and Witkowski, 2002).  
Unfortunately the majority (70%) of seedlings died soon after germination due to 
fungal contamination, even after treatment with fungicide.  DNA was therefore 
extracted from the remaining seedlings as soon as roots were long enough to provide 
sufficient tissue (approximately 1 cm).  This was successful, and AFLP analysis was 
performed.  The number of samples available was however not sufficient to allow 
accurate inference of breeding system from AFLP band patterns.  All that could be 
inferred was that both I. adenoides and I. zeyheri are not obligate self-fertilisers. 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 62 
4.  AFLP BAND PATTERNS 
 
4.1. Indigofera adenoides 
 
A total of 44 bands were obtained for I. adenoides samples, varying in size from 
75bp to approximately 1500bp (Figure 14).  The mean number of bands per 
individual was 30.3, and the total number of polymorphic bands was 40 (90.9% of 
the total number of bands obtained).  All bands were found to be reproducible.  
Three bands that were found at a low frequency in the far off site groups were not 
found at all in off site or toepaddock groups.  These were fragments 2, 36, and 42. 
 
1           2           3           4           5           6   7          8           9          10
 Level of first 
band scored
 Polymorphic 
band
 Last band 
scored
  
 
Figure 14:  Silver stained gel showing the typical band pattern obtained with 
Indigofera adenoides samples.  Levels of the first and last bands scored are shown, 
as well as an example of a polymorphic band.  Samples are present in lanes 2-9, lane 
10 contains a molecular weight marker.  The largest molecular marker band is 1031 
base pairs, while the smallest one visible on the gel is 100 base pairs. 
 
 63 
4.2. Indigofera zeyheri 
 
The total number of fragments obtained with I. zeyheri was 34, varying in size from 
approximately 40bp to approximately 1600bp while the mean number of fragments 
per individual was 19.1 (Figure 15).  The total number of polymorphic fragments 
was 30 (88.2% of the total number of fragments).  Three fragments were found at a 
low frequency in the off site group that were not found at all in the toepadddock 
group.  These were fragments 1, 7, and 19.   
 
First band scored
 Polymorphic band
 Last band scored
 2 3 4 5 6 7 8 9 10
  
 
Figure 15:  Silver stained gel showing the typical band pattern obtained with 
Indigofera zeyheri samples.  Levels of the first and last bands scored are shown, as 
well as an example of a polymorphic band.  Samples are present in lanes 2-9, lane 
10 contains a molecular weight marker.  The largest molecular marker band is 1031 
base pairs, while the smallest one visible on the gel is 50 base pairs. 
 
 
 
 
 64 
5. ANALYSIS OF INDIGOFERA ADENOIDES ACCORDING TO AREA 
 
Initial analysis of Indigofera adenoides data was performed by separating 
individuals into groups according to the six main areas of the slimes dam and 
surrounding areas, namely toepaddock, toepaddock wall, retaining wall, off site, far 
off site A, and far off site B.  This was done to determine if genetic and 
morphological differences could be detected between the plants that have naturally 
colonised these different areas.  Any differences detected would presumably be due 
to differences in toxicity between the areas and/or differences in local growth 
conditions, for example increased water availability in the toepaddock due to 
seepage from the moist slimes continually being added to the slimes dam. 
 
5.1. Morphological analysis 
 
Indigofera adenoides samples were divided into five groups for morphological 
analysis.  Retaining wall and toepaddock wall individuals were grouped together as 
plants from these groups appeared morphologically similar (average values for plant 
height and leaf width differed by 1.3 cm and 0.04 cm respectively).  These areas also 
have similar substrate conditions, both consisting of walls of grassland soil that have 
been placed in the toepaddock.  The groups analysed were therefore toepaddock, 
retaining/toepaddock wall, off site, far off site A, and far off site B.  Analysis of 
variance (ANOVA) showed plants from the five areas to be significantly different 
from each other, with most of the variables contributing significantly to 
differentiation between the areas (Table 5).  
 
 
 
 
 
 
 
 65 
Table 5:  Analysis of variance (ANOVA) performed for all variables (DF=4, error 
DF = 141).  This was done to determine which variables show significant 
differences between area groups.   
 
Variable F value Pr > F 
Plant height 10.47 <0.0001 
Canopy width 6.40 <0.0001 
Canopy width 90? 2.02 0.0948 
Leaf length 26.28 <0.0001 
Leaf width 71.43 <0.0001 
 
Tukey analysis emphasised the differences in plant height and leaf width that were 
shown to be significant with ANOVA (Table 6).  Differences in these variables lead 
to differences in leaf area, canopy area, and canopy volume, due to the presence of 
plant height and leaf width in the formulas used for calculating these composite 
variables.  Differences in other variables were seen, but only between far off site B 
and the other areas, showing the extent to which plants in this area are 
morphologically differentiated from those in the other areas.   
   
 
 
 
 
 
 
 
 
 
 
 
 
 
 66 
Table 6:  Canopy and leaf dimensions (mean ? standard deviation) for plants in the 
various area groups.  Superscripts with the same letter within a row indicate values 
that are not significantly different (P<0.05%; Tukey). 
 
Factor Retaining wall, 
toepaddock wall 
Toepaddock Off site Far off A Far off B 
Plant height (cm) 47.2?12.4a 33.7?11.3b 
 
39.7?12.7a 
 
28.3?6.7b 
 
27.4?7.6b 
Canopy diameter (cm) 50.1?18.6a 
 
51.2?24.6a 
 
49.3?21.3a 
 
38.0?14.8ab 
 
27.9?8.2b 
Canopy diameter90? 
(cm) 
38.4?18.9a 
 
37.8?19.5a 
 
39.1?17.7a 
 
28?12.6a 
 
18.5?5.8a 
Leaf length (cm) 1.5?0.3a 
 
1.3?0.3a 
 
1.7?0.4a 
 
1.6?0.4a 
 
1.3?0.2a 
Leaf width (cm) 0.5?0.1b 0.4?0.1b 
 
0.6?0.2a 
 
0.6?0.1a 
 
0.6?0.1a 
Leaf area (cm2) 
 
0.6?0.3ac 0.5?0.2c 0.8?0.5a 0.8?0.4ac 0.6?0.2b 
Leaf shape (cm) 
 
3.0?0.7ab 3.0?0.7a 3.3?1.0ab 2.8?0.6b 2.4?0.5ab 
Canopy area (cm2) 
 
1697?       
1341a 
1830? 
1637a 
1629?  
1500a 
963?   
770ab 
429?  
212b 
Canopy volume (cm3) 
 
82987?   
68799a 
65588?  
66369ab 
63135? 
54542a 
30254? 
28487bc 
12363?  
8457c 
Canopy volume 2 
(cm3) 
55324?   
45866a 
43725? 
44246ab 
42090? 
36361a 
20170? 
18991bc 
8241?  
5638c 
 
Following the results of this analysis, graphs were constructed showing the 
distribution of the variables shown to be significantly different between the areas, 
i.e., leaf width and plant height.  Graphs of mean values for canopy area and canopy 
volume were also constructed.  Leaf width is, on average, larger in plants from the 
areas less affected by slimes (off site and far off site areas).  Leaf width is the most 
diverse in individuals from the off site areas (Figure 16).  Plants from far off site 
areas are not only shorter on average, they also show less variation in height than 
individuals in areas affected by slimes (Figure 17).  Mean canopy area and canopy 
volume are highly variable throughout all areas (large standard deviations).  Both 
variables are significantly lower in the far off site populations (Figure 18).  
 67 
Retaining wall and Toepaddock wall
 0
 2
 4
 6
 8
 10
 0.2-
 0.29
 0.3-
 0.39
 0.4-
 0.49
 0.5-
 0.59
 0.6-
 0.69
 0.7-
 0.79
 0.8-
 0.89
 0.9-
 0.99
 1.0-
 1.09
 Leaf width (cm)
 N
 um
 be
 r 
of
  
pla
 nt
 s
 Toepaddock
 0
 5
 10
 15
 20
 25
 0.2-
 0.29
 0.3-
 0.39
 0.4-
 0.49
 0.5-
 0.59
 0.6-
 0.69
 0.7-
 0.79
 0.8-
 0.89
 0.9-
 0.99
 1.0-
 1.09
 Leaf width (cm)
 Nu
 m
 be
 r 
o
 f p
 la
 n
 ts
 Off site
 0
 1
 2
 3
 4
 5
 6
 7
 0.2-
 0.29
 0.3-
 0.39
 0.4-
 0.49
 0.5-
 0.59
 0.6-
 0.69
 0.7-
 0.79
 0.8-
 0.89
 0.9-
 0.99
 1.0-
 1.09
 Leaf width (cm)
 Nu
 m
 be
 r 
o
 f p
 la
 n
 ts
 Far off site A
 0
 1
 2
 3
 4
 5
 6
 7
 8
 0.2-
 0.29
 0.3-
 0.39
 0.4-
 0.49
 0.5-
 0.59
 0.6-
 0.69
 0.7-
 0.79
 0.8-
 0.89
 0.9-
 0.99
 1.0-
 1.09
 Leaf width (cm)
 Nu
 m
 be
 r 
o
 f p
 la
 n
 ts
 Far off site B
 0
 1
 2
 3
 4
 5
 6
 7
 8
 0.2-
 0.29
 0.3-
 0.39
 0.4-
 0.49
 0.5-
 0.59
 0.6-
 0.69
 0.7-
 0.79
 0.8-
 0.89
 0.9-
 0.99
 1.0-
 1.09
 Leaf width (cm)
 N
 u
 m
 be
 r 
of
  
pl
 an
 ts
 Figure 16:  Graphs showing the 
distribution of leaf width in the different 
areas.  Note the different y-axis scales.
 N
 um
 be
 r 
of
  
pla
 nt
 s
 Nu
 m
 be
 r 
o
 f p
 la
 n
 ts
 Nu
 m
 be
 r 
o
 f p
 la
 n
 ts
 Nu
 m
 be
 r 
o
 f p
 la
 n
 ts
 N
 u
 m
 be
 r 
of
  
pl
 an
 ts
  
 
 68 
Retaining wall and Toepaddock wall
 0
 2
 4
 6
 8
 10
 10-19 20-29 30-39 40-49 50-59 60-69 70-79
 Plant height (cm)
 N
 u
 m
 be
 r 
o
 f p
 la
 n
 ts
 Off site
 0
 1
 2
 3
 4
 5
 6
 7
 8
 9
 10-19 20-29 30-39 40-49 50-59 60-69 70-79
 Plant height (cm)
 N
 um
 be
 r 
of
  
pla
 nt
 s
 Far off site A
 0
 2
 4
 6
 8
 10
 12
 14
 10-19 20-29 30-39 40-49 50-59 60-69 70-79
 Plant height (cm)
 N
 u
 m
 be
 r 
of
  
pl
 a
 n
 ts
 Far off site B
 0
 2
 4
 6
 8
 10
 12
 14
 10-19 20-29 30-39 40-49 50-59 60-69 70-79
 Plant height (cm)
 N
 u
 m
 be
 r 
of
  
pl
 an
 ts
 Figure 17:  Graphs showing the 
distribution of plant height in the different 
areas.  Note the different y-axis scales.
 Toepaddock
 0
 5
 10
 15
 20
 25
 10-19 20-29 30-39 40-49 50-59 60-69 70-79
 Plant height (cm)
 N
 u
 m
 be
 r 
o
 f p
 la
 n
 ts
 N
 u
 m
 be
 r 
o
 f p
 la
 n
 ts
 N
 um
 be
 r 
of
  
pla
 nt
 s
 N
 u
 m
 be
 r 
of
  
pl
 a
 n
 ts
 N
 u
 m
 be
 r 
of
  
pl
 an
 ts
 N
 u
 m
 be
 r 
o
 f p
 la
 n
 ts
  
 
 69 
Canopy Area
 0
 500
 1000
 1500
 2000
 2500
 3000
 3500
 4000
 Retaining/
 toepaddock
 wall
 Toepaddock Off site Far off site A Far off site B
 Area
 C
 a
 n
 o
 py
  
a
 re
 a
  
(cm
 2 )
 Canopy volume 
0
 20000
 40000
 60000
 80000
 100000
 120000
 Retaining/
 Toepaddock
 wall
 Toepaddock Off site Far off site A Far off site B
 Area
 Ca
 n
 o
 py
  
v
 o
 lu
 m
 e 
(cm
 3 )
 C
 a
 n
 o
 py
  
a
 re
 a
  
(cm
 2 )
 Ca
 n
 o
 py
  
v
 o
 lu
 m
 e 
(cm
 3 )
  
 
Figure 18:  A comparison of canopy area (mean ? standard deviation) and canopy 
volume (mean ? standard deviation) in plants from the different areas. 
 
The most notable differences in correlation coefficients were found in far off site B 
canopy area and leaf area, and canopy volume and leaf area correlations.  These 
differences are due to the significantly lower canopy volume and canopy area values 
in plants from this group, when correlated with a relatively constant leaf area 
measurement (Table 7).   
 
 70 
Table 7:  A comparison of correlation coefficients between various variables in 
plants from the different areas.  The asterix (*) indicates a p value of less than 
0.0001, showing the two variables to be highly significantly correlated.   
 
Correlation 
between 
Overall Retaining and 
toepaddock wall 
Toepaddock Off site Far off 
site A 
Far off 
site B 
Height and leaf 
length 
0.340* 0.285 0.165 0.146 0.648 -0.211 
Height and leaf 
width 
0.272 0.0667 0.161 
 
0.134 0.185 
 
0.200 
Height and  
canopy area 
0.374* 0.319 0.260 0.0695 0.645 0.231 
Height and leaf 
area 
0.306 0.234 0.150 0.0999 0.561 0.110 
Leaf length and 
leaf width 
0.597* 0.226 0.230 0.543 0.115 0.204 
Canopy area 
and leaf area 
0.744* 0.798* 0.774* 0.773* 0.705* 0.236 
Canopy volume 
and leaf area 
0.722* 0.759* 0.722* 0.738* 0.718* 0.242 
 
Discriminant analysis show toepaddock and far off site B plants to classify most 
consistently to the correct area group, while plants from the other groups show only 
a low percentage of correct classifications (Table 8).  This indicates that plants in 
these two area groups are morphologically distinct from plants in all other areas 
groups. 
 
 
 
 
 
 
 
 
 71 
Table 8:  Classification of plants from each area group into the different areas, 
expressed as a percentage of the total number of individuals in each area group, 
obtained through discriminant analysis.  The total number of plants that classifies to 
each area group is shown, expressed as a percentage of the total number of plants 
sampled.      
 
    Area group classified to   
Area group of origin Sample 
size 
R/TW 
(%) 
TP 
(%) 
OS 
(%) 
FOSA 
(%) 
FOSB 
(%) 
Retaining wall and 
toepaddock wall (R/TW) 
24 40.91 45.45 13.64 0.00 0.00 
Toepaddock (TP) 54 7.14 78.57 8.93 5.36 0.00 
Off site (OS) 24 8.70 39.13 43.48 8.70 0.00 
Far off site A (FOSA) 25 0.00 40.00 20.00 40.00 0.00 
Far off site B (FOSB) 19 0.00 20.00 0.00 0.00 80.00 
Total  10.27 52.74 15.75 10.27 10.96 
 
Canonical discriminant analysis was used to produce a distance matrix of pair-wise 
squared distances between the various area groups.  The dendrogram constructed 
from that data shows plants from far off site B to be highly morphologically distant 
from plants in the other groups.  Plants in toepaddock/retaining wall and off site 
groups are the most morphologically similar to each other (Figure 19). 
 
 
 
 
 
 
 
 
 
 
 
 72 
 
 
1
 Far off site B
 Toepaddock
 Far off site
 Retaining wall and 
toepaddock wall
 Off site
 1
 A
  
 
Figure 19:  Dendrogram based on pair-wise squared distances of morphological 
data, obtained through canonical discriminant analysis.  The scale bar indicates 1 
distance unit.   
 
The diversity of each variable within groups of plants from the different areas was 
determined by calculating coefficients of variation (Table 9), in order to allow the 
comparison of morphological variation and genetic variation.  Toepaddock and far 
off site A area groups contain the most variance between individuals.  Leaf length, 
leaf width, and leaf shape show the largest differences in correlation coefficients 
between area groups. 
 
 
 
 
 
 73 
 
Table 9:  A comparison of coefficients of variation for each variable between groups 
of plants from the different areas.  The highest coefficient of variation for each 
variable is shown in bold.   
 
Variable Retaining/ 
toepaddock wall 
Toepaddock Off site Far off 
site A 
Far off 
site B 
Sample size 24 54 24 25 19 
Height 7.33 11.79 8.54 8.16 7.93 
Canopy width 37.69 47.47 43.17 38.83 29.28 
Canopy width 90 48.95 51.47 45.29 45.07 31.30 
Leaf length 32.90 39.30 28.68 41.62 60.80 
Leaf width 61.03 106.08 50.13 63.29 37.54 
Canopy area 11.37 16.09 10.93 16.22 11.35 
Canopy volume1 8.56 12.39 7.98 12.70 8.41 
Canopy volume2 8.89 12.89 8.29 13.24 8.80 
Leaf area 51.58 56.61 57.26 51.61 29.09 
Leaf shape 45.63 47.86 37.36 75.30 28.91 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 74 
5.2. Genetic analysis 
 
5.2.1. Initial grouping 
 
Initial organisation of groups for genetic analysis involved grouping retaining wall 
and toepaddock wall plants together.  Far off site A and B were also grouped 
together.  This was done to provide a broad overview of genetic structure between 
different substrate types.  Genetic diversity is lower in plants from areas that have a 
higher level of slimes contamination, i.e., toepaddock area groups (toepaddock, 
retaining wall, and toepaddock wall).   Plants from these areas also have a lower 
percentage of loci polymorphic (Table 10).  The majority of total population 
variation is due to diversity within groups, indicating a low level of genetic 
differentiation between groups.  Estimated gene flow levels are high (Table 11). 
 
Table 10:  Genetic structure of plants from the different areas in terms of sample 
size, percentage of loci polymorphic, Shannon?s information index (I) (mean ? 
standard deviation), and Nei?s (1973) gene diversity (h) (mean ? standard deviation).   
 
Area Sample 
size 
Percentage of 
loci polymorphic 
I h 
Retaining wall/ 
toepaddock wall 
24 61.34 0.335 ? 0.293 0.226 ? 0.206 
Toepaddock 54 65.9 0.383 ? 0.294 0.261 ? 0.207 
Off site 24 75.0 0.443 ? 0.276 0.304 ? 0.197 
Far off site 44 84.1 0.446 ? 0.244 0.300 ? 0.175 
 
 
 
 
 
 
 
 75 
Table 11:  Nei?s gene diversity statistics (mean ? standard deviation) for subdivided 
populations, showing the genetic structure of the overall population.  An estimate of 
gene flow (Nm) is also shown.   
 
Statistic Total genetic 
diversity (HT) 
Diversity within 
groups (HS) 
Gene flow between 
groups (Nm) 
Value 0.288 ? 0.031 0.273 ? 0.030 8.975 
 
A permutation test for genetic differentiation among area groups was performed for 
all groups.  The FST value is 0.0211 ? 0.408 (mean ? standard error).  Permutation 
testing on this FST value showed the null hypothesis to be rejected at the 1% level, 
indicating that the area groups are more genetically diverse than a random 
assemblage of the individuals.  Permutation testing of FST values between each pair 
of area groups emphasises the diverse nature of the plants in the far off site groups, 
with these groups being significantly differentiated from each of the other groups 
(Table 12).   
 
Table 12:  Permutation test for genetic differentiation among groups of plants from 
different areas, based on pairwise FST values between each pair of groups.  Numbers 
in italics indicate rejection of the null hypothesis at the 5% level, while numbers in 
bold type indicate rejection of the null hypothesis at the 1% level.  If the null 
hypothesis is rejected, it indicates the two populations are more genetically 
differentiated than random assemblages of the individuals at the significance level 
indicated.   
 
 Retaining wall/ toepaddock wall Toepaddock Off site 
Toepaddock 0.005 /  
Off site 0.016 0.013 / 
Far off site 0.029 0.024 0.037 
 
 
 
 76 
Nei?s genetic distances (D) between plants from the area groups were calculated and 
used to construct a dendrogram.  Plants in the far off site groups are genetically 
distant from those in the other groups, while plants in the toepaddock and 
toepaddock/ retaining wall groups are the most genetically similar to each other 
(Figure 20). 
 
 
0.001
 Far off site
 Off site
 Toepaddock
 Toepaddock wall/ 
Retaining wall
  
 
Figure 20:  Dendrogram based on pair-wise Nei?s genetic distances (D) between 
plant groups from different areas.  The scale bar indicates 0.001 genetic distance 
units.   
 
Plants were assigned to area groups based on genetic similarity (Table 13).  
Accurate classification is higher in the off site and far off site groups than in the on 
site groups.  Probability (P) values are however lower in the off site groups, 
indicating that classification is less certain in those areas.  Individuals were assigned 
as soon as the likelihood of being assigned to one group was higher than that for any 
other group, as assignment only when likelihood values were 10 times higher 
 77 
produced a high percentage (75%) of nil assignments.  This indicates the groups are 
quite similar, as would be expected considering their close geographic proximity. 
 
Table 13:  Percentage of plants from each area group that assign to the different 
groups, based on AFLPOP analysis.  Values shown indicate the percentage of 
individuals from each area group that are shown to assign to each of the groups.  
Individuals are assigned to groups as soon as the likelihood of belonging to that 
group is higher than the likelihood of belonging to any other group.  Probability (P) 
values are shown for each group indicating the likelihood that assignment was 
correct.   
 
                  Area of origin    
Area group assigned to (%) R/TW TP OS FOS 
Sample size 24 54 24 44 
Retaining/ Toepaddock Wall (R/TW) 66.7 20.4 12.5 4.6 
Toepaddock (TP) 8.3 61.1 4.2 11.4 
Off site (OS) 4.2 13 75 9.1 
Far off site (FOS) 20.9 5.6 8.3 75 
Average P value 0.70 0.61 0.55 0.52 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 78 
5.2.2. Subgroups 
 
Following initial genetic analysis, groups of plants were subdivided further in order 
to provide information on genetic differences between plants due to finer scale 
differences in growth conditions.  Toepaddock wall plants were separated from 
retaining wall plants and the far off site plant group was separated into far off site A 
for the plants sampled several hundred metres from the dam, and far off site B for 
the plants sampled from the site several kilometres away from the dam.   
 
As was seen in the initial area groups, genetic diversity is lower in the toepaddock, 
toepaddock wall, and retaining wall groups than in off site and far off site groups.  
Lowest values are seen in the retaining wall group.  Interestingly, the two far off site 
groups show different diversities, with far off site A showing a higher diversity than 
far off site B (Table 14).   
 
Table 14:  Genetic structure of plants from the different areas in terms of sample 
size, percentage of loci polymorphic, Shannon?s information index (I) (mean ? 
standard deviation), and Nei?s (1973) gene diversity (h) (mean ? standard deviation).  
Note toepaddock and offsite group values are the same as in Table 10, and are 
reproduced here for comparison purposes.   
 
Population Sample 
size 
Percentage of 
loci polymorphic 
h I 
Retaining wall 9 47.7 0.178 ? 0.205 0.264 ? 0.295 
Toepaddock wall 15 54.6 0.215 ? 0.215 0.315 ? 0.307 
Toepaddock 54 65.9 0.264 ? 0.207 0.385 ? 0.295 
Off site 24 75.0 0.304 ? 0.197 0.443 ? 0.276 
Far off site A 25 77.3 0.251 ? 0.175 0.383 ? 0.251 
Far off site B 19 72.7 0.304 ? 0.203 0.440 ? 0.285 
 
 
 
 79 
The majority of total variation of the entire population is due to diversity between 
plants within the areas, indicating a low genetic differentiation between plants from 
different areas.  Estimated gene flow is of a moderate level, lower than that 
estimated for the initial area groups (Table 15). 
 
Table 15:  Nei?s gene diversity statistics (mean ? standard deviation) for subdivided 
populations, showing the genetic structure of the overall population.  An estimate of 
gene flow (Nm) is also shown.  
 
Statistic Total genetic 
diversity (HT) 
Diversity within 
groups (HS) 
Gene flow between 
groups (Nm) 
Value 0.281 ? 0.031 0.253 ? 0.027 4.448 
 
The overall FST for all groups is 0.0369 ? 0.1830.  Permutation testing for genetic 
differentiation among groups showed this value to be significant, resulting in a 
rejection of the hypothesis at the 1% level.  Therefore, although differentiation 
between plants from different areas is low, they are still significantly different from 
each other.  Permutation testing between each pair of groups showed far off site 
plants to be genetically differentiated from those in most other areas, as was noted in 
the initial area groupings.  The retaining wall group is not significantly differentiated 
from the toepaddock wall group, supporting morphological analysis where these two 
groups were grouped together (Table 16). 
 
 
 
 
 
 
 
 
 
 
 
 80 
Table 16:  Permutation test for genetic differentiation among groups of plants from 
different areas, based on pairwise FST values between each pair of groups.  Numbers 
in italics indicate rejection of the null hypothesis at the 5% level, while numbers in 
bold type indicate rejection of the null hypothesis at the 1% level.  If the null 
hypothesis is rejected, it means the two populations are more genetically 
differentiated than random assemblages of the individuals.   
 
 Retaining 
wall 
Toepaddock 
wall 
Toepaddock Off site Far off 
site A 
Toepaddock wall  0.0434 /    
Toepaddock 0.0176 0.0110 /   
Off site 0.0212 0.0262 0.0119 /  
Far off site A 0.1029 0.0277 0.0450 0.0678 / 
Far off site B 0.0460 0.0348 0.0223 0.0200 0.0495 
 
As for the initial area grouping, a dendrogram based on Nei?s genetic distances (D) 
shows plants from the far off site areas to be genetically distant from plants in the 
other area groups.  Interestingly, the retaining wall group is not shown to group with 
the toepaddock wall group, even though these groups are morphologically similar 
and were not significantly different from each other during FST permutation testing 
(Figure 21).  
 
 
 
 
 
 81 
0.001
 Far off site A
 Far off site B
 Retaining wall
 Off site
 Toepaddock
 Toepaddock wall
  
 
Figure 21:  Dendrogram based on pair-wise Nei?s genetic distances (D) between 
plant groups from different areas.  The scale bar indicates 0.001 genetic distance 
units.   
 
A genetic distance matrix was constructed showing distances between groups as 
shown in Figure 21, except that retaining wall and toepaddock wall plants were 
grouped together.  This was in order that a Mantel test could be performed 
comparing this distance matrix to the one used to construct the dendrogram of 
morphological distance (Figure 19).  This test showed a slight, non-significant 
(P=0.65), negative correlation (rxy=-0.456) between the two distance matrices.  
Group assignment showed a higher percentage of accurate classifications in plants 
from the retaining wall, toepaddock wall, and far off site A areas.  Accurate 
classification was lowest in plants from the toepaddock group (Table 17). 
 
 
 82 
Table 17:  Percentage of plants from each area that assign to the different area 
groups, based on relatedness values between individuals.  Values shown indicate the 
percentage of individuals from each area that is shown to assign to the various 
groups.  Individuals are assigned to groups as soon as the likelihood of belonging to 
that group is higher than the likelihood of belonging to any other group.   
 
 Population of origin 
Population assigned to (%) RW TW TP OS FOSA FOSB 
Sample size 9 15 54 24 25 19 
Retaining wall (RW) 77.8 0 16.7 12.5 0 5.3 
Toepaddock wall (TW) 0 86.7 16.7 4.2 4 5.3 
Toepaddock (TP) 0 0 40.7 8.3 0 10.5 
Off site (OS) 22.2 0 11.1 66.7 0 5.3 
Far off site A (FOSA) 0 13.3 5.6 4.2 84 10.5 
Far off site B (FOSB) 0 0 9.2 4.2 12 63.2 
Average P value 0.68 0.70 0.62 0.55 0.58 0.53 
Percentage P values less 
than 0.3 
0 6.6 9.3 20.8 20 21.1 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 83 
6. ANALYSIS OF INDIGOFERA ZEYHERI ACCORDING TO AREA  
 
6.1. Morphological analysis 
 
Indigofera zeyheri plants were divided into two groups ? toepaddock and off site ? 
as no far off site or retaining wall plants could be obtained.  To begin, T-tests were 
performed on each variable to determine which variables contribute significantly 
towards differences between plants from the different areas (Table 18).  The 
variables shown to be significantly different between plants from the different areas, 
canopy area and canopy volume, were represented graphically.  Canopy width 
variables were not represented in a graph as these variables make up the canopy area 
measurement.  In contrast to Indigofera adenoides, measurements for both canopy 
area and canopy volume are significantly lower in plants growing on the toepaddock 
than in plants in the other areas.  All measurements have a very high standard 
deviation, indicating high variability in the morphology of plants within the different 
areas (Figure 22).  No large differences in correlation coefficients were observed for 
pairs of variables between the two areas (Table 19). 
 
Table 18:  T-test showing which variables are significantly different between plants 
from the different areas (DF=96). 
  
Variable T value Pr > t 
Plant Height 1.55 0.1234 
Canopy Width 4.23 <0.0001 
Canopy Width 90? 4.08 <0.0001 
Canopy Area 4.33 <0.0001 
Canopy Volume  4.09 <0.0001 
Leaf Area -0.60 0.5486 
Leaf Shape -2.58 0.0115 
Leaf Length -1.34 0.1839 
Leaf Width 0.36 0.7230 
 
 
 84 
 
Canopy area 
-1000
 0
 1000
 2000
 3000
 4000
 5000
 6000
 7000
 Toepaddock Off site
 Area
 A
 re
 a
  
(cm
 2 )
 Canopy volume 
-100000
 0
 100000
 200000
 300000
 400000
 500000
 Toepaddock Off site
 Area
 Ca
 n
 o
 py
  
v
 o
 lu
 m
 e 
(cm
 3 )
 A
 re
 a
  
(cm
 2 )
 Ca
 n
 o
 py
  
v
 o
 lu
 m
 e 
(cm
 3 )
  
 
Figure 22:  A comparison of canopy area (mean ? standard deviation) and canopy 
volume (mean ? standard deviation) between plants from the two area groups.   
 
 
 
 
 
 
 85 
Table 19:  A comparison of correlation coefficients between various variables in 
plants from different areas.  The asterix (*) indicates a p value of less than 0.0001, 
showing the two variables to be highly correlated.   
 
Correlation between Overall Toepaddock Off site 
Height and leaf length 0.340* 0.165 0.146 
Height and leaf width 0.272 0.161 0.134 
Height and  canopy area 0.374* 0.260 0.0695 
Height and leaf area 0.306 0.150 0.0999 
Leaf length and leaf width 0.597* 0.230 0.543 
Canopy area and leaf area -0.161 -0.139 -0.186 
Canopy volume and leaf area -0.145 -0.0953 -0.241 
 
Coefficients of variation show plants from the toepaddock to be more 
morphologically diverse than plants from the off site areas (Table 20).  Discriminant 
analysis and canonical discriminant analysis could not be performed for Indigofera 
zeyheri as only two groups were available for comparison. 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 86 
Table 20:  A comparison of coefficients of variation for each variable in plants from 
the two area groups.  The higher coefficient for each variable is shown in bold type.   
 
Variable Off site Toepaddock 
Sample size 35 63 
Height 5.16 7.78 
Canopy Width 42.61 57.58 
Canopy Width 90 54.29 68.72 
Leaf Length 19.79 21.42 
Leaf Width 22.42 18.85 
Canopy Area 12.43 21.58 
Canopy Volume 8.81 14.89 
Canopy Volume 2 9.12 15.46 
Leaf area 33.13 32.58 
Leaf shape 12.48 13.25 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 87 
6.2. Genetic analysis 
 
Genetic analysis was performed to compare the genetic structure of the toepaddock 
group of plants with that of the off site group of plants.  A subset of I. zeyheri plants 
were used for AFLP analysis, due to time constraints.  The toepaddock group has a 
lower genetic diversity than the off site group, although the difference between the 
groups is smaller than was seen with I. adenoides (Table 21).  Total population 
diversity is mostly due to diversity within groups, indicating that diversity between 
the two groups of plants is low.  Estimated gene flow levels are very high (Table 
22). 
   
Table 21:  Genetic structure of the groups of plants from the two areas in terms of 
sample sizes, percentage of loci polymorphic, Shannon?s information index (I) 
(mean ? standard deviation), and Nei?s (1973) gene diversity (h) (mean ? standard 
deviation).   
 
Population Sample 
size 
Percentage of 
loci polymorphic 
h I 
Toepaddock 26 73.5 0.2531 ? 0.1941 0.3800 ? 0.2711 
Off site 16 82.4 0.2542 ? 0.1839 0.3872 ? 0.2538 
 
Table 22:  Nei?s gene diversity statistics (mean ? standard deviation) for subdivided 
populations, showing the genetic structure of the overall population.  An estimate of 
gene flow (Nm) is also shown. 
 
Statistic Total genetic 
diversity (HT) 
Diversity within 
groups (HS) 
Gene flow between 
groups (Nm) 
Value 0.2642 ? 0.0303 0.2536 ? 0.0290 12.0201 
 
The FST value for the overall population is 0.0067 ? 0.0318.  Permutation testing 
showed this FST value to not be significant.  The null hypothesis was therefore not 
rejected, indicating that the groups are not more genetically differentiated than a 
random assemblage of the individuals.  
 88 
7. ANALYSIS OF INDIGOFERA ADENOIDES ACCORDING TO TOXICITY 
 
The analysis performed in the previous chapter involved the relation of the plants to 
the slimes dams.  Following this analysis, plants were separated into groups 
according to toxicity information obtained from AngloGold Ashanti Ltd. technical 
data (Labuschagne, 2004), in order to determine if plants show genetic or 
morphological differences that correspond to differing toxicity levels.  Indigofera 
adenoides plants were separated into six toxicity groups based on total dissolved 
solids (TDS) levels in groundwater, measured from boreholes in the area of the 
slimes dam, which have been shown to be correlated with soil toxicity (Figure 23).   
 
Total dissolved solids in this data set are made up mostly of sulphates and chlorides.  
Increased TDS levels increase the toxicity of the groundwater to plants due to 
increased osmotic stress, a lower pH, as well as due to the toxicity of the individual 
salts and mobile elements, and the interactions between them (Weiersbye et al., 
2002).  Zone 2 can therefore be considered to have the least toxic groundwater for 
the purposes of this study, while zone 7 can be considered to have the most toxic 
groundwater.  The depth to groundwater is important as it determines the extent to 
which the toxicity of the groundwater will impact on plants growing on the surface.  
A consistently high depth to groundwater is evident throughout the area surrounding 
the South Complex slimes dam (14-19 mbgl), with slightly shallower levels being 
evident in higher toxicity zones (Figure 24).  
 
 
 
 
 
 
 
 
 
 
 
 89 
 
200
 600
 1000
 1400
 1800
 2200
 2600
 3000
 3400
 3800
 TDS (mg/l)
 2 3 4 5 6 7
 1000m0m
  
 
Figure 23:  Zones of levels of total dissolved solids (TDS) surrounding the South 
Complex slimes dam.  Adapted from AngloGold Ashanti Ltd. technical data.  Plants 
were divided into the indicated zones 2 to 7. 
 
WL (mbgl)
 0
 1
 2
 3
 4
 5
 6
 7
 8
 9
 10
 11
 12
 13
 14
 15
 16
 17
 18
 19
 20
 21
 22
 14m
 bgl
 0m 1000m
  
 
 90 
Figure 24:  Depth to groundwater in metres below ground level (mbgl) for the 
toxicity zones studied.  Adapted from AngloGold Ashanti Ltd. technical data.   
7.1. Genetic analysis using plants from all areas 
 
Initial analysis involved separating plants from the retaining wall, toepaddock wall, 
toepaddock, and off site areas into the six zones (zones 2 to 7) shown in Figure 23.  
This was done to determine if a significant genetic or morphological difference 
could be detected between the toxicity zones, regardless of the changes in local 
growth conditions caused by proximity of the plants to the slimes dam.   
 
Genetic diversity is the highest in the lowest toxicity group (group 2), is fairly 
constant in groups 3, 4, and 5, decreases in group 6, then increases sharply in the 
highest toxicity group (group 7).  The percentage of loci polymorphic decreases 
from group 2 to 6, then increases in group 7 (Table 23).  The majority of total 
genetic diversity is due to diversity within toxicity groups, indicating a low level of 
genetic differentiation between the different toxicity groups of plants.  Estimated 
gene flow levels are moderate (Table 24). 
 
Table 23:  Genetic structure of the plants in the toxicity groups in terms of sample 
size, percentage of loci polymorphic, Shannon?s information index (I) (mean ? 
standard deviation), and Nei?s (1973) gene diversity (h) (mean ? standard deviation). 
 
Toxicity 
group 
Sample size Percentage of 
loci polymorphic 
h I 
Group 7 15 65.91 0.2703 ? 0.2137 0.3919 ? 0.3018 
Group 6 10 47.73 0.1948 ? 0.2187 0.2829 ? 0.3109 
Group 5 10 56.82 0.2362 ? 0.2182 0.3422 ? 0.3115 
Group 4 17 61.36 0.2292 ? 0.2041 0.3398 ? 0.2918 
Group 3 20 63.64 0.2266 ? 0.2018 0.3378 ? 0.2879 
Group 2 30 68.18 0.2590 ? 0.2014 0.3820 ? 0.2864 
 
 
 91 
 
 
Table 24:  Nei?s gene diversity statistics (mean ? standard deviation) for subdivided 
populations, showing the genetic structure of the overall population.  An estimate of 
gene flow (Nm) is also shown.   
 
Statistic Total genetic 
diversity (HT) 
Diversity within 
groups (HS) 
Gene flow between 
groups (Nm) 
Value 0.2666 ? 0.0377 0.2360 ? 0.0309 3.8547 
 
The overall FST is 0.0320 ? 0.2110 (mean ? standard error).  Permutation testing 
showed this value to be significant, resulting in a rejection of the null hypothesis at 
the 1% level.  The groups of plants are therefore more genetically diverse than a 
random assemblage of the individuals.  Permutation testing of FST  values for each 
pair of groups shows plants in the highest toxicity group (group 7) to be the most 
genetically differentiated from plants in the other groups (Table 25).  FST data are 
supported by a dendrogram constructed from Nei?s genetic distances, which shows 
plants in the highest toxicity groups (group 7 and 6) to be the most genetically 
distant from plants in the other groups (Figure 25).   
 
Table 25:  Permutation test for genetic differentiation among toxicity groups, based 
on pairwise FST values between each pair of groups.  Numbers in italics indicate 
rejection of the null hypothesis at the 5% level, while numbers in bold type indicate 
rejection of the null hypothesis at the 1% level.  If the null hypothesis is rejected, it 
means the two populations are more genetically differentiated than random 
assemblages of the individuals at the significance level indicated. 
 
 Group7 Group6  Group5 Group4 Group3 
Group6 0.0585 /    
Group5 0.0345 0.0269 /   
Group4 0.0405 0.0284 0.0071 /  
Group3 0.0491 0.0335 0.0231 0.0114 / 
 92 
Group2 0.0248 0.0257 0.0546 0.0360 0.0163 
 
 
 
Group 5
 Group 7
 Group 6
 Group 2
 Group 3
 Group 4
 0.001
  
 
Figure 25:  Dendrogram based on pair-wise Nei?s genetic distances (D) between 
plants in the different toxicity groups.  The scale bar indicates 0.001 genetic distance 
units.   
 
 
 
 
 
 
 
 
 
 93 
 
 
 
7.2. Morphological analysis using plants from all areas 
 
Initial morphological analysis was performed with the same groups used in the 
genetic analysis section 7.1.  Toxicity groups of plants are not significantly 
morphologically different from each other at the P<0.001 level.  The two measures 
of canopy diameter, together with leaf shape, are the closest to being significantly 
different between the groups.  Consequently, canopy area is significantly different 
between the groups at the P<0.05 level (Table 26).  Canopy diameter, and 
consequently canopy area, is the only variable shown to be significantly different 
between plants in the different toxicity groups following Tukey analysis (Table 27). 
 
Table 26:  Analysis of variance (ANOVA) for all variables. DF=5, error DF=96.  P 
values indicate whether the variable accounts significantly towards variation 
between the different toxicity groups of plants.  
 
 F P > F 
Plant height 0.34 0.8861 
Canopy diameter 2.31 0.0500 
Canopy diameter 90? 2.36 0.0456 
Leaf length 1.38 0.2388 
Leaf width 1.59 0.1714 
Leaf area 1.65 0.1553 
Leaf shape 2.14 0.0676 
Canopy area 2.83 0.0198 
Canopy volume 2.10 0.0724 
 
 
 
 
 
 94 
 
Table 27:  Plant and leaf dimensions (mean ? standard deviation) for plants in each 
of the toxicity groups.  Superscripts of the same letter within a row indicate values 
that are not significantly different (P<0.05%; Tukey).  
 
Factor Group7 Group6 Group5 Group4 Group3 Group2 
Plant 
height (cm) 
37.8?13.0a 43.7?14.0a 
 
38.1?14.0a 
 
37.5?11.6a 
 
37.2?14.8a 38.2?12.5a 
Canopy 
diameter 
(cm) 
47.9?17.3a 
 
41.0?13.9a 
 
44.2?21.1a 
 
66.2?24.4a 
 
50.0?27.0a 48.8?21.9a  
Canopy 
diameter 
90? (cm) 
34.4?17.0ab 
 
27.5?12.9a 
 
34.8?18.4ab 
 
50.4?23.8b 
 
37.3?20.3ab 38.9?16.0ab 
Leaf length 
(cm) 
1.4?0.4a 
 
1.4?0.2a 
 
1.2?0.2a 
 
1.5?0.4a 
 
1.3?0.4a 1.5? 0.4a 
Leaf width 
(cm) 
0.5?0.2a 0.5?0.1a 
 
0.4?0.1a 
 
0.6?0.2a 
 
0.5?0.2a 0.5? 0.1a 
Leaf area 
(cm) 
0.6?0.4a 0.6?0.2a 0.4?0.1a 0.7?0.5a 0.6?0.3a 0.5? 0.3a 
Leaf shape  
 
3.1?0.9a 2.8?0.3a 2.9?0.7a 2.8?0.5a 3.0?0.9a 3.4? 0.9a 
Canopy 
area (cm2) 
1464? 
1305ab 
985?   
796a 
1443?  
1170ab 
2941?  
2103b 
1799?  
1533ab 
1664? 
1454ab 
Canopy 
volume 
(cm3) 
61442?  
66560a 
40017?  
28757a 
56634? 
52776a 
113552? 
87756a 
73674?  
68541a 
65238? 
60077a 
Canopy 
volume 2 
(cm3) 
40962?  
44374a 
26678? 
19172a 
37756? 
35184a 
75701? 
58504a 
49116?  
45694a 
43492? 
40051a 
 
Several differences in correlation coefficients of plant measurements can be noted 
between toxicity groups, but these differences do not seem to be located in any 
particular group, nor follow any particular pattern (Table 28).  This separation of 
plants into toxicity groups was not morphologically analysed further as it was 
apparent these groups of plants are not significantly differentiated from each other.  
Genetic analysis of these groups did not show dramatic differences, either (Section 
 95 
7.1).  This is thought to be due to local conditions that differ between off site and 
toepaddock areas, for example the toepaddock has increased water levels due to 
seepage from the slimes.  Diversity of morphological characteristics of plants within 
toxicity groups would therefore be high, resulting in little differentiation between 
groups.  Toxicity groups were therefore redesigned. 
 
Table 28:  A comparison of correlation coefficients between various variables in 
plants in the different toxicity groups.  The asterix (*) indicates a p value of less than 
0.0001, showing the two variables to be highly correlated.   
 
Correlation 
between 
Group 7 Group 6 Group 5 Group 4 Group 3 Group 2 
Height and leaf 
length 
-0.122 -0.107 0.467 0.438 0.437 0.085 
Height and leaf 
width 
-0.092 0.091 0.043 
 
0.546 -0.231 
 
-0.213 
Height and  
canopy diameter 
0.403 -0.297 -0.057 0.314 0.335 0.280 
Height and 
canopy 
diameter90 
0.523 -0.317 0.215 0.100 0.599 -0.001 
Canopy diameter 
and diameter 90 
0.795 0.780 0.860 0.748 0.827* 0.643 
Leaf length and 
leaf width 
0.712 0.812 0.373 0.838* 0.602 0.523 
Canopy area and 
leaf area 
0.141 0.186 -0.445 -0.314 0.043 0.488 
Canopy area and 
canopy volume 1 
0.968* 0.793 0.857 0.909* 0.900* 0.927* 
Canopy volume 
and leaf area 
-0.015 0.220 -0.302 -0.167 0.036 0.413 
 
 
 96 
 
7.3. Genetic analysis using plants from toepaddock areas 
 
The previous analysis containing plants from all areas surrounding the slimes dam 
did not indicate significant differentiation between plants from different toxicity 
groups.  Off site plants were therefore removed from the analysis to see if a 
significant difference exists between toxicity groups containing individuals from the 
toepaddock areas (toepaddock, toepaddock wall, and retaining wall).  Groups 6 and 
7 were combined due to the small sample sizes of the individual groups, to create 
group 6/7.   
 
Genetic diversity and the percentage of loci polymorphic are similar in all groups, 
with variation in values seeming not to follow any particular pattern (Table 29).  
Diversity between plants from different groups is very low, as the majority of total 
genetic diversity is due to diversity within groups.  Estimated gene flow levels are 
moderately high (Table 30).   
 
Table 29:  Genetic structure of the toxicity groups in terms of sample size, 
percentage of loci polymorphic, Nei?s (1973) gene diversity (h) (mean ? standard 
deviation), and Shannon?s information index (I) (mean ? standard deviation).  
 
Toxicity 
group 
Sample size Percentage of 
loci polymorphic 
h I 
Group 6/7 13 52.27 0.2215 ? 0.2253 0.3187 ? 0.3193 
Group 5 10 56.82 0.2362 ? 0.2182 0.3422 ? 0.3115 
Group 4 13 54.55 0.2099 ? 0.2114 0.3088 ? 0.3015 
Group 3 20 63.64 0.2266 ? 0.2018 0.3378 ? 0.2879 
Group 2 22 59.09 0.2295 ? 0.2134 0.3362 ? 0.3033 
 
 
 
 97 
Table 30:  Nei?s gene diversity statistics (mean ? standard deviation) for subdivided 
populations, indicating the genetic structure of the overall population.  An estimate 
of gene flow (Nm) is also shown. 
 
Statistic Total genetic 
diversity (HT) 
Diversity within 
groups (HS) 
Gene flow between 
groups (Nm) 
Value 0.2487 ? 0.0413 0.2247 ? 0.0352 4.6833 
 
The overall FST value is 0.0175 ? 0.3832.  Permutation testing showed this value to 
be significant, and the null hypothesis was rejected at the 1% level.  Therefore 
although differentiation between plants from different toxicity groups is low, they 
are still significantly differentiated from each other.  Permutation testing between 
each pair of area groups shows plants from different groups to be highly genetically 
similar to each other (Table 31). 
 
Table 31:  Permutation test for genetic differentiation between plants from different 
toxicity groups, based on pairwise FST values between each pair of groups.  Numbers 
in italics indicate rejection of the null hypothesis at the 5% level.  If the null 
hypothesis is rejected, it indicates the two populations are more genetically 
differentiated than random assemblages of the individuals at the significance level 
indicated. 
 
 Group7/6  Group5 Group4 Group3 
Group5 0.0258 /   
Group4 0.0042 0.0020 / 
 
Group3 0.0408 0.0231 -0.0056 / 
Group2 0.0054 0.0528 0.0066 0.0119 
 
These analyses show the different toxicity groups of plants to not be significantly 
differentiated from each other.  Further genetic analysis was therefore not performed 
on these groups.  Morphological analysis was also not performed.  Toxicity groups 
were therefore redesigned further.   
 
 98 
7.4. Genetic analysis using plants from the toepaddock only  
 
Analysis was performed on toxicity groups containing toepaddock individuals only.  
This was done to remove the effect of local growth condition changes caused by 
proximity of the plants to the slimes dam.  Two different sets of groups were 
analysed.  One set contained groups 6 and 7 grouped together (group 6/7) while the 
other contained groups 5, 6, and 7 grouped together (group 5/6/7).  This was done 
due to the small size of group 6/7.   
 
Diversity is lower in group 6/7, although this could be due to the small sample size 
of this group.  Diversity of group 5/6/7 is very high.  Diversity in the lower toxicity 
groups (groups 2, 3, and 4) decreases slightly with increasing toxicity (Table 32).  
When calculated for the set containing group 6/7, differentiation between groups is 
higher, and gene flow lower, than that calculated for the set containing group 5/6/7 
(Table 33). 
 
Table 32:  Genetic structure of toxicity groups in terms of sample sizes, percentage 
of loci polymorphic, Shannon?s information index (I) (mean ? standard deviation), 
and Nei?s (1973) gene diversity (h) (mean ? standard deviation).  Two separate sets 
of groups are shown: groups 6 and 7 combined, or groups 5, 6, and 7 combined.   
 
Toxicity 
group 
Sample 
size 
Percentage of loci 
polymorphic 
h I 
Group 6/7 6 40.91 0.1578 ? 0.2083 0.2314 ? 0.2965 
Group 5 10 56.82 0.2362 ? 0.2182 0.3422 ? 0.3115 
Group 4 12 52.27 0.1988 ? 0.2106 0.2931 ? 0.3004 
Group 3 18 61.36 0.2191 ? 0.2054 0.3256 ? 0.2926 
Group 2 17 56.82 0.2206 ? 0.2146 0.3232 ? 0.3046 
Group 5/6/7 16 61.36 0.2572 ? 0.2141 0.3729 ? 0.3056 
 
 
 
 99 
Table 33:  Nei?s gene diversity statistics (mean ? standard deviation) for subdivided 
populations, showing the genetic structure of the overall population.  An estimate of 
gene flow (Nm) is also shown.  Values are shown for the two different sets of 
toxicity groups. 
 
Statistic Total genetic 
diversity (HT) 
Diversity within 
groups (HS) 
Gene flow between 
groups (Nm) 
Value 6/7 0.2480 ? 0.0406 0.2065 ? 0.0304 2.4859 
Value 5/6/7 0.2457 ? 0.0405 0.2239 ? 0.0347 5.1499 
 
Overall FST is 0.0556 ? 0.5497 for the set containing group 6/7 and 0.0183 ? 0.3825 
for the set containing group 5/6/7.  Permutation testing showed the null hypothesis 
to be rejected at the 1% level for both sets of groups, indicating the groups are more 
genetically diverse than a random assemblage of the individuals.  Permutation 
testing of FST values was then performed between each pair of toxicity groups in 
each set of groups individually.  Group 6/7 is significantly genetically differentiated 
from all other groups.  Group 5/6/7 is significantly genetically differentiated from 
most other groups.  The remaining groups are genetically similar to each other 
(Tables 34 and 35).  A dendrogram constructed from Nei?s genetic distances (D) 
shows group 6/7 and group 5/6/7 to be genetically distant from the remaining groups 
in their respective sets (Figure 26). 
 
 
 
 
 
 
 
 
 
 
 
 
 100 
Table 34:  Permutation test for genetic differentiation among toxicity groups of the 
set containing group 6/7, based on pairwise FST values between each pair of groups.  
Values in italics indicate rejection of the null hypothesis at the 5% level, while 
numbers in italics indicate rejection of the null hypothesis at the 1% level.  If the 
null hypothesis is rejected, it indicates the two populations are more genetically 
differentiated than random assemblages of the individuals.   
 
 Group 6/7  Group5 Group4 Group3 
Group5 0.0981 /   
Group4 0.0858 0.0070 /  
Group3 0.1439 0.0214 -0.0016 / 
Group2 0.1078 0.0511 0.0127 0.0016 
 
Table 35:  Permutation test for genetic differentiation among toxicity groups of the 
set containing group 5/6/7, based on pairwise FST values between each pair of 
groups.  Values in italics indicate rejection of the null hypothesis at the 5% level, 
while numbers in italics indicate rejection of the null hypothesis at the 1% level.  If 
the null hypothesis is rejected, it indicates the two populations are more genetically 
differentiated than random assemblages of the individuals.   
 
 Group5/6/7 Group4 Group3 
Group4 0.0109 / 
 
Group3 0.0347 -0.0016 / 
Group2 0.0451 0.0127 0.0016 
 
 
 
 101 
(a)
 (b)
 0.01
 0.001
 Group6/7
 Group5
 Group2
 Group3
 Group4
 Group5/6/7
 Group2
 Group3
 Group4
  
 
Figure 26:  Dendrograms based on Nei?s genetic distances (D) showing the genetic 
distance between plants in different toxicity groups.  Two sets of groups are shown: 
the set containing group 6/7 (a) and the set containing group 5/6/7 (b).  The scale bar 
indicates the shown number of genetic distance units.   
 
 102 
Assignment of plants to the correct toxicity group is high in all groups, and for both 
sets of groups.  The P value is the highest in group 6/7, indicating assignment is the 
most certain in that group.  Individuals are not assigned incorrectly to group 5/6/7 
with a high frequency (Tables 36 and 37). 
 
Table 36:  Percentage of plants from each toxicity group in the set containing group 
6/7 that assign to the different toxicity groups.  Values shown indicate the 
percentage of plants from each group that assigns to the various groups.  Plants are 
assigned to groups as soon as the likelihood of belonging to that group is higher than 
the likelihood of belonging to any other group.  Probability (P) values are also 
shown indicating the average probability of correct classification for each group. 
 
                                        Group of origin 
Group assigned to (%) Group 6/7 Group 5 Group 4 Group 3 Group 2 
Sample size 6 10 12 18 17 
Group 6/7 66.7 0 0 0 5.9 
Group 5 16.7 80.0 0 0 0 
Group 4 16.7 20.0 75.0 5.6 23.5 
Group 3 0 0 8.3 77.8 5.9 
Group 2 0 0 16.7 16.7 64.7 
Average P value 0.87 0.55 0.57 0.57 0.59 
 
 
 
 
 
 
 
 
 
 
 
 
 103 
Table 37:  Percentage of plants from each toxicity group in the set containing group 
5/6/7 that assign to the different toxicity groups.  Values shown indicate the 
percentage of plants from each group that assigns to the various groups.  Plants are 
assigned to groups as soon as the likelihood of belonging to that group is higher than 
the likelihood of belonging to any other group.  Probability (P) values are also 
shown indicating the average probability of correct classification for each group. 
 
                                        Group of origin 
Group assigned to (%) Group 5/6/7 Group 4 Group 3 Group 2 
Sample size 16 12 18 17 
Group 5/6/7 75.0 8.3 0 0 
Group 4 18.8 66.7 5.6 23.5 
Group 3 0 8.3 77.8 5.9 
Group 2 6.3 16.7 16.7 70.6 
Average P value 0.52 0.60 0.57 0.60 
 
.   
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 104 
7.5. Morphological analysis using plants from the toepaddock only 
 
Morphological analysis was performed with groups 6/7, 5, 4, 3, and 2, as these 
groups were shown to be the most genetically differentiated from each other.  No 
variables are significantly different at the P<0.001 level between plants in different 
toxicity groups.  The two measures of canopy diameter, and consequently canopy 
area and canopy volume, are significantly different between plants in different 
toxicity groups at the P<0.05 level (Table 38).   
 
Table 38:  Analysis of variance (ANOVA) for all variables.  DF=4, error DF=58.  P 
values indicate whether the variable accounts significantly towards variation 
between the different toxicity groups of plants. 
 
 F P > F 
Plant height 0.31 0.8697 
Canopy diameter 2.83 0.0324 
Canopy diameter 90? 4.19 0.0047 
Leaf length 0.56 0.6941 
Leaf width 0.78 0.5429 
Leaf area 0.74 0.5656 
Leaf shape 1.03 0.3983 
Canopy area 5.33 0.0010 
Canopy volume 4.12 0.0053 
 
Tukey results support ANOVA analysis, showing plants in group 4 to be over twice 
the canopy area and canopy volume of those plants in other toxicity groups (Table 
39).  Following the results of these analyses, graphs were constructed showing the 
distribution of canopy area in the various toxicity groups (Figure 27).  Canopy area 
is highest in group 4, and lowest in group 6/7. 
 
 
 
 105 
Table 39:  Plant and leaf dimensions (mean ? standard deviation) for plants in the 
various toxicity groups.  Superscripts of the same letter within a row indicate values 
that are not significantly different (P<0.05%; Tukey).  
 
Factor Group7/6 Group5 Group4 Group3 Group2 
Plant height (cm) 30.8?7.4a 38.1?13.2a 
 
37.5?12.1a 
 
35.4?14.5a 35.0?12.0a 
Canopy diameter 
(cm) 
54.0?8.3ab 
 
44.2?20.0ab 
 
71.2?24.3a 
 
48.6?28.0ab 44.7?21.4b  
Canopy diameter 
90? (cm) 
34.0?13.8ab 
 
34.8?17.4b 
 
58.2?24.0a 
 
34.7?19.8b 33.5?11.9b 
Leaf length (cm) 1.4?0.4a 
 
1.2?0.2a 
 
1.4?0.4a 
 
1.3?0.3a 1.4? 0.3a 
Leaf width (cm) 0.4?0.1a 0.4?0.1a 
 
0.5?0.2a 
 
0.4?0.2a 0.4? 0.1a 
Leaf area (cm) 0.5?0.2a 0.4?0.1a 0.6?0.4a 0.5?0.3a 0.5? 0.2a 
Leaf shape 3.4?0.9a 2.9?0.7a 2.8?0.5a 3.0?1.0a 3.2? 0.7a 
Canopy area 
(cm2) 
1514?  
818b 
1443?  
1110b 
3609?  
2155a 
1672?  
1557b 
1276? 
1002b 
Canopy volume 
(cm3) 
47706?  
29803b 
56634? 
50068b 
138846? 
91048a 
64310?  
63763b 
50704? 
56209b 
Canopy volume 
2 (cm3) 
31804?  
19869b 
37756? 
33378b 
92564? 
60699a 
42873?  
42509b 
33802? 
37473b 
 
 
 106 
Group 6/7
 0
 0.5
 1
 1.5
 2
 2.5
 3
 3.5
 0-600 601-
 1200
 1201-
 1800
 1801-
 2400
 2401-
 3000
 3001-
 3600
 3601-
 4200
 4201-
 4800
 4801-
 5400
 5401-
 ?
 Canopy area (cm2)
 N
 u
 m
 be
 r 
o
 f p
 la
 n
 ts
 Group 4 
0
 0.5
 1
 1.5
 2
 2.5
 3
 3.5
 0-600 601-
 1200
 1201-
 1800
 1801-
 2400
 2401-
 3000
 3001-
 3600
 3601-
 4200
 4201-
 4800
 4801-
 5400
 5401-
 ?
 Canopy area (cm2)
 N
 u
 m
 be
 r 
o
 f p
 la
 n
 ts
 Group 5
 0
 0.5
 1
 1.5
 2
 2.5
 3
 3.5
 0-600 601-
 1200
 1201-
 1800
 1801-
 2400
 2401-
 3000
 3001-
 3600
 3601-
 4200
 4201-
 4800
 4801-
 5400
 5401-
 ?
 Canopy area (cm2)
 N
 u
 m
 be
 r 
o
 f p
 lan
 ts
 Group 3
 0
 1
 2
 3
 4
 5
 6
 7
 0-600 601-
 1200
 1201-
 1800
 1801-
 2400
 2401-
 3000
 3001-
 3600
 3601-
 4200
 4201-
 4800
 4801-
 5400
 5401-
 ?
 Canopy area (cm2)
 N
 u
 m
 be
 r 
o
 f p
 la
 n
 ts
 Group 2
 0
 1
 2
 3
 4
 5
 6
 7
 0-600 601-
 1200
 1201-
 1800
 1801-
 2400
 2401-
 3000
 3001-
 3600
 3601-
 4200
 4201-
 4800
 4801-
 5400
 5401-
 ?
 Canopy area (cm2)
 N
 u
 m
 be
 r 
o
 f p
 la
 n
 ts Figure 27: The distribution of canopy 
area in the different toxicity groups.
 N
 u
 m
 be
 r 
o
 f p
 la
 n
 ts
 N
 u
 m
 be
 r 
o
 f p
 la
 n
 ts
 N
 u
 m
 be
 r 
o
 f p
 lan
 ts
 N
 u
 m
 be
 r 
o
 f p
 la
 n
 ts
 N
 u
 m
 be
 r 
o
 f p
 la
 n
 ts
  
 
The majority of differences in correlation coefficients are seen in plants from the 
highest (group 6/7) and lowest (group 2 and group 3) toxicity groups, and in 
correlations containing canopy area or associated variables, further indicating the 
extent that canopy area varies between plants from different toxicity groups (Table 
40).  A higher percentage of accurate classifications is seen in plants from higher 
toxicity groups (groups 4, 5, and 6/7) than in plants from lower toxicity groups 
(groups 2 and 3) (Table 41). 
 107 
Table 40:  A comparison of correlation coefficients between variables in plants from 
different toxicity groups.  The asterix (*) indicates a p value of less than 0.0001, 
showing the two variables to be highly correlated.   
 
Correlation between Group 7/6 Group 5 Group 4 Group 3 Group 2 
Height and leaf length 0.542 0.467 0.448 0.334 -0.170 
Height and leaf width         0.601 0.043 0.526 -0.437 -0.427 
Height and  canopy 
diameter 
0.149 -0.057 0.258 0.297 0.504 
Height and canopy 
diameter90 
0.185 0.215 0.167 0.530 -0.045 
Canopy diameter and 
diameter 90 
0.951 0.860 0.846 0.841* 0.532 
Leaf length and leaf 
width 
0.698 0.373 0.785 0.505 0.621 
Canopy area and leaf 
area 
0.712 -0.445 -0.151 -0.093 -0.007 
Canopy area and 
canopy volume 1 
0.846 0.857 0.885 0.908* 0.970* 
Canopy volume and 
leaf area 
0.921 -0.302 -0.018 -0.204 -0.046 
 
 
 
 
 
 
 
 
 
 
 
 108 
Table 41:  Classification of plants from each toxicity group into different groups, 
expressed as a percentage of the total number of plants in each group, obtained 
through discriminant analysis.  The total number of plants that classifies to each 
toxicity group is shown, expressed as a percentage of the total number of plants.  
 
From area Group2 
(%) 
Group3 
(%) 
Group4 
(%) 
Group5 
(%) 
Group6/7 
(%) 
Sample size 17 18 12 10 6 
Group 2 29.41 17.65 11.76 11.76 29.41 
Group 3 22.22 5.56 22.22 27.78 22.22 
Group 4 0.00 0.00 75.00 16.67 8.33 
Group 5 0.00 20.00 10.00 60.00 10.00 
Group 6/7 0.00 16.67 33.33 0.00 50.00 
Total 14.29 11.11 28.57 23.81 22.22 
 
A dendrogram constructed from pair-wise squared distances between groups of 
plants, obtained through canonical discriminant analysis, shows plants in groups 4 
and 6/7 to be the most morphologically distant from plants in the other groups 
(Figure 28).  A Mantel test was performed to compare the distance matrix used to 
construct the dendrogram above with the one used to construct the dendrogram of 
genetic distance between the same groups, shown in Figure 26.  The test showed a 
slight, significant (P = 0.940), negative correlation (Rxy = -0.490) between the two 
distance matrices.   
 
 
 109 
0.1
 Group 4
 Group 6/7
 Group 2
 Group 5
 Group 3
  
 
Figure 28:  Dendrogram based on pair-wise squared distances of morphological 
data, obtained through canonical discriminant analysis.  The scale bar indicates 0.1 
distance units.   
 
The diversity of each variable within the different toxicity groups of plants was 
determined by calculating coefficients of variation, in order to allow a comparison 
of morphological variation and genetic variation (Table 42).  The most variable 
plants are generally found in the least toxic areas, while the least variable individuals 
do not seem to follow a particular trend in toxicity. 
 
 
 
 
 
 110 
Table 42:  A comparison of coefficients of variation for each variable in plants from 
different toxicity groups.  The highest coefficient for each variable is shown in bold, 
with the lowest coefficient shown in italics.   
 
Variable Group7/6 Group5 Group4 Group3 Group2 
Height 6.69 9.75 9.67 10.60 16.87 
Canopy width 15.45 47.67 34.09 57.68 47.94 
Canopy width 90 40.71 52.74 41.25 56.99 35.60 
Leaf length 97.38 100.17 102.01 138.41 82.40 
Leaf width 36.53 28.62 39.87 36.50 26.29 
Canopy area 54.06 81.10 59.71 93.14 78.48 
Canopy volume1 62.47 93.19 65.58 99.15 110.86 
Leaf area 52.48 33.79 64.19 54.83 46.34 
Leaf shape 27.30 24.86 19.67 31.87 20.07 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 111 
8. ANALYSIS OF INDIGOFERA ZEYHERI ACCORDING TO TOXICITY 
 
Indigofera zeyheri plants were separated into two toxicity groups based on 
groundwater sulphate levels, measured from boreholes in the area of the slimes dam, 
which have been shown to be correlated with toxicity (Figure 29).  Higher sulphate 
levels increase the toxicity of the groundwater to plants due to higher levels of 
osmotic stress.  High sulphate levels also indicate a low pH of the groundwater, 
which leads to a higher concentration of heavy metals in the groundwater through 
increasing the solubility of these metals (Weiersbye et al., 2002).  The depth to 
groundwater is important as it determines the extent to which the toxicity of the 
groundwater will impact on plants growing on the surface.  A moderate depth to 
groundwater is evident throughout the area surrounding the New North Complex 
slimes dam (4-14 mbgl), with slightly shallower levels being evident in higher 
toxicity zones (Figure 30).  
 
MBH2
 MBH3
 MBH41
 MBH44
 0
 100
 200
 300
 400
 500
 600
 700
 800
 900
 1000
 1100
 Sulphate
 (mg/l)
 0m 1000m
  
 
Figure 29:  Zones of sulphate (mg/l) surrounding the New North Complex slimes 
dam.  Individuals were separated into two toxicity groups.  The low toxicity group 
contained individuals from zones of 200 ? 500 mg sulphate/l, while the high toxicity 
group contained individuals from zones of 600-800 mg sulphate/l. 
 112 
 
MBH2
 MBH3
 MBH41
 MBH44
 0
 40
 36
 32
 28
 24
 20
 16
 12
 8
 4
 MBH6
 WL (mbgl)
 0m 1000m
  
 
Figure 30:  Zones of groundwater level in metres below ground level (mbgl) for the 
two toxicity zones.  Individuals sampled were found in zones of 4-14 mbgl. 
 
8.1. Genetic analysis using plants from all areas 
 
Initial analysis of Indigofera zeyheri according to toxicity levels involved separating 
plants into two toxicity groups as described in Figure 29.  Toepaddock and off site 
samples were included in this analysis, in order to determine if a significant 
difference exists between the two groups regardless of local growth conditions 
caused by proximity to the slimes dam.  Genetic diversity and the percentage of loci 
polymorphic are higher in plants from the higher toxicity group (Table 43).  Total 
genetic diversity is mostly due to diversity within toxicity groups, indicating that 
differentiation between plants from different toxicity groups is low.  Estimated gene 
flow levels are high (Table 44).   
 
 
 
 113 
Table 43:  Genetic structure of plants from the two toxicity groups in terms of 
sample size, percentage of loci polymorphic, Shannon?s diversity index (I) (mean ? 
standard deviation), and Nei?s (1973) gene diversity (h) (mean ? standard deviation). 
 
Area Sample 
number 
Percentage of loci 
polymorphic  
h I 
High toxicity 27 79.41 0.2614 ? 0.1961 0.3918 ? 0.2701 
Low toxicity 15 67.65 0.2159 ? 0.1953 0.3272 ? 0.2772 
 
Table 44:  Nei?s gene diversity statistics (mean ? standard deviation) for subdivided 
populations, showing the genetic structure of the overall population.  An estimate of 
gene flow (Nm) is also shown.  
 
Statistic Total genetic 
diversity (HT) 
Diversity within 
groups (HS) 
Gene flow between 
groups (Nm) 
Value 0.2553 ? 0.0278 0.2386 ? 0.0265 7.1662 
 
The FST  value is 0.0249 ? 0.00242 (mean ? standard error).  A permutation test for 
genetic differentiation among toxicity groups showed this FST value to not be 
significant.  The null hypothesis is therefore not rejected, indicating that the groups 
are not more genetically diverse than a random assemblage of the individuals.  
 
 
 
 
 
 
 
 
 
 
 
 114 
8.2. Morphological analysis using plants from all areas 
 
Initial morphological analysis involved comparing morphology of off site and 
toepaddock plants that were separated into two toxicity groups.  No variable came 
close to being significantly different between the areas at the P<0.05 level (Table 
45).  These groups of plants are therefore not significantly differentiated from each 
other.  Therefore further morphological analysis was not performed on these groups. 
 
Table 45:  T-test showing which variables contribute significantly to differences 
between the toxicity groups (DF = 40).  P values indicate which variables are 
significantly different between the areas.  
 
Variable T value Pr > t 
Plant Height 1.39 0.1731 
Canopy Width 0.70 0.4854 
Canopy Width 90? 0.79 0.4370 
Leaf Length 0.35 0.7246 
Leaf Width 0.25 0.8009 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 115 
8.3. Genetic analysis using plants from the toepaddock only 
 
Plants from the toepaddock only were separated into two toxicity groups.  This was 
done to remove the effect of local variations in growth conditions caused by 
proximity to the slimes dam.  As was seen in the initial toxicity groups, genetic 
diversity and the percentage of loci polymorphic are higher in plants from the group 
with higher toxicity (Table 46).  Total genetic diversity is mostly due to diversity 
within toxicity groups, indicating that diversity between plants from different 
toxicity groups is low.  Estimated gene flow levels are high, although lower than 
those estimated for the previous toxicity groups containing off site individuals 
(Table 47). 
 
Table 46:  Genetic structure of plants from the two toxicity groups in terms of 
sample sizes, percentage of loci polymorphic, Shannon?s information index (I) 
(mean ? standard deviation), and Nei?s (1973) gene diversity (h) (mean ? standard 
deviation).  
 
Area Sample 
size 
Percentage of 
loci polymorphic 
h I 
High toxicity 19 73.53 0.2504 ? 0.2045 0.3730 ? 0.2832 
Low toxicity 7 50.00 0.1721 ? 0.1985 0.2591 ? 0.2852 
 
Table 47:  Nei?s gene diversity statistics (mean ? standard deviation) for subdivided 
populations, showing the genetic structure of the overall population.  An estimate of 
gene flow (Nm) is also shown. 
  
Statistic Total genetic 
diversity (HT) 
Diversity within 
groups (HS) 
Gene flow between 
groups (Nm) 
Value 0.2343 ? 0.0303 0.2113 ? 0.0267 4.5879 
 
The FST  value for the overall population is 0.0289 ? 0.0142 (mean ? standard error).  
A permutation test for genetic differentiation showed this FST value to not be 
 116 
significant.  The null hypothesis was therefore not rejected, indicating that plants 
from different toxcity groups are not more genetically diverse than a random 
assemblage of the individuals.  
 
8.4. Morphological analysis using plants from the toepaddock only  
 
Morphological analysis was performed on the groups used in section 8.3.  
Differences in plant height are statistically significant at the P<0.05 level (Table 48).  
Graphs were created showing the distribution of plant height in the two different 
toxicity groups (Figure 31).  This shows that plants in the high toxicity group have 
both a higher average height and a larger variation in height.   
  
Table 48:  T-test showing which variables contribute significantly to differences 
between the toxicity groups (DF = 24).  P values indicate whether the variable 
accounts for significant differentiation between the groups.   
 
Variable T value Pr > t 
Plant Height 2.06 0.0502 
Canopy Width 0.64 0.5274 
Canopy Width 90? 0.18 0.8611 
Leaf Length 0.12 0.9092 
Leaf Width 0.00 0.9983 
Canopy area 0.60 0.5538 
Canopy volume 0.84 0.4102 
Leaf area 0.56 0.5780 
Leaf shape 0.21 0.8316 
             
 
 117 
Low toxicity
 0
 0.5
 1
 1.5
 2
 2.5
 3
 3.5
 31-40 41-50 51-60 61-70 71-80 81-90 91-
 100
 101-
 110
 111-
 120
 121-
 130
 Plant height (cm)
 N
 u
 m
 be
 r 
o
 f p
 la
 n
 ts
 High toxicity
 0
 1
 2
 3
 4
 5
 6
 7
 8
 31-40 41-50 51-60 61-70 71-80 81-90 91-
 100
 101-
 110
 111-
 120
 121-
 130
 Plant height (cm)
 N
 u
 m
 be
 r 
o
 f p
 la
 n
 ts
 N
 u
 m
 be
 r 
o
 f p
 la
 n
 ts
 N
 u
 m
 be
 r 
o
 f p
 la
 n
 ts
  
 
Figure 31:  The distribution of plant height in the two toxicity groups.  Note the 
different y-axis scales.  
 
Differences in correlation coefficients are apparent in many pairs of variables.  
These differences are particularly notable in correlations involving plant height, 
emphasising the difference in this variable between plants in the different toxicity 
groups.  A marked difference is also seen in the correlation of leaf length and leaf 
width, with this correlation being significant in plants from the high toxicity group, 
but not in plants from the low toxicity group. 
  
 
 118 
Table 49:  A comparison of correlation coefficients between various variables in 
plants from the different toxicity groups.  The asterix (*) indicates a p value of less 
than 0.0001, showing the two variables to be highly correlated.   
 
Correlation between High toxicity Low toxicity 
Height and leaf length 0.518 0.126 
Height and leaf width          0.387 -0.576 
Height and  canopy diameter 0.474 0.918 
Height and canopy diameter90 0.390 0.530 
Canopy diameter and diameter 90 0.768 0.350 
Leaf length and leaf width 0.904* 0.366 
Canopy area and leaf area -0.424 0.091 
Canopy volume and leaf area -0.391 0.205 
 
The diversity of each variable in plants within the different toxicity groups was 
determined by calculating coefficients of variation, in order to allow a comparison 
of morphological variation and genetic variation.  Higher morphological variation 
was found in the higher toxicity areas, except in the case of plant height, the only 
variable shown to account for variation between the areas, and leaf shape, both of 
which show very similar coefficients of variation in the two groups (Table 50). 
 
 
 
 
 
 
 
 
 
 
 
 
 119 
Table 50:  A comparison of coefficients of variation for each variable in plants from 
the different toxicity groups.  The higher coefficient for each variable is shown in 
bold.   
 
Variable High toxicity Low toxicity 
Sample size 19 7 
Height 7.22 7.76 
Canopy Width 55.48 45.10 
Canopy Width 90 70.77 65.38 
Leaf Length 30.41 16.13 
Leaf Width 22.36 9.71 
Canopy Area 135.08 66.00 
Canopy Volume 156.24 68.60 
Leaf area 40.41 21.79 
Leaf shape 15.76 15.96 
                                    
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 120 
DISCUSSION AND CONCLUSIONS 
 
1. Genetic differentiation between areas 
 
Relatively lower genetic diversity was observed in plants more closely associated 
with the slimes dams, on the toepaddock and associated areas of the toepaddock wall 
and retaining wall, when compared to plants further away from the dam, i.e., off site 
and far off site areas.  This difference was particularly notable when toepaddock and 
associated areas were compared to off site areas.  These groups were sampled in 
close proximity to each other, and the sizes of the areas sampled were similar.  
Therefore the only difference between the areas was proximity to the slimes dam, 
with an increase in slimes-contaminated runoff and other pollution existing in the 
toepaddock and associated areas.  This difference was found in both Indigofera 
adenoides and Indigofera zeyheri, although the difference was smaller in I. zeyheri.  
Lower diversity was noted in all measures of genetic diversity, namely the 
percentage of loci polymorphic, Shannon?s information index (I), and Nei?s (1973) 
gene diversity statistic (h). 
  
It cannot be determined whether this difference in diversity is due to colonisation, 
persistence, or natural selection.  A lower genetic diversity in the toepaddock and 
associated areas, together with the presence of some bands in the far off site but not 
toepaddock samples, provides support for colonisation of the dams occurring by a 
small number of individuals from the surrounding areas.  Colonisation of toxic areas 
results in several members of the founding population dying shortly after 
colonisation, which would produce a loss of genetic variation in the colonising 
population, and the presence of unique alleles off the colonised area (Bush and 
Barrett, 1993).   
 
In addition, changes in diversity could have involved persistence of plants from 
before the initial disturbance occurred, with individuals pre-adapted to the slimes-
 contaminated environment surviving, while other genotypes died out over time.  It is 
unlikely that all plants sampled were persisters from before construction of the 
slimes dams, as many of them appeared very small and were not significantly 
 121 
wooded, indicating they could not be over 25 years old.  Retaining and toepaddock 
wall plants were not likely to be persisters, as these walls were created when the 
slimes dams were constructed and the height of the plants makes it unlikely they 
grew through from below the wall.  These plants may, however, have resprouted 
from root stocks that were included intact when the retaining wall and toepaddock 
wall were created from soil from the surrounding areas. 
 
The slimes dams studied are both over 25 years old.  Both I. adenoides and I. 
zeyheri are perennial plants that presumably produce seeds once a year.  The 
generation time is the time taken for newly established seeds to become reproductive 
adults, and it is not known how long this would take.  However, it is likely that there 
had been several generations of plants between establishment of the dams and the 
time of the study.  Gene flow levels were estimated to be moderate to high between 
areas.  Therefore with these levels of gene flow occurring over so many years, why 
did differences in genetic variation still exist between areas?  And why were unique 
alleles present in far off site populations that were not found on site?  It could be that 
seedling mortality was so high that only a few plants from each seed production 
event survived, or natural selection could have played a role, resulting in only 
particular genotypes being able to survive on the toepaddock and associated areas.   
 
Alternatively, genetic differences may have been maintained due to the observed 
difference in flowering time between the areas.  Flowering occurred earlier in the 
season on the toepaddock and associated areas, presumably due to increased 
available water in these areas, due to continued deposition of moist slimes on the top 
of the dam.  Increased amounts of available water could lead to more rapid growth 
and maturation of plants in toepaddock areas, restricting pollination between the 
areas and causing an isolation of toepaddock populations.  Gene flow between areas 
would therefore have been due mostly to the movement of seeds rather than the 
movement of pollen.  Lower genetic diversity in the toepaddock and associated 
areas therefore originated and was maintained through either a founder effect 
following colonisation, natural selection, flowering time differences, or a 
combination of these factors. 
 
 122 
In I. adenoides, the initial grouping of plants (toepaddock, toepaddock wall/retaining 
wall, off site, and far off site A and B) had an overall genetic diversity (HT) only 
marginally larger than the diversity within groups (HS).  This indicated a very low 
genetic differentiation between groups, indicating that significant gene flow 
occurred between the areas.  This was supported by the measures of FST and Nm, 
which indicated high levels of differentiation within compared to between groups, 
and a large amount of estimated gene flow, respectively.  Significant levels of gene 
flow between slimes dam-associated and off site groups were expected, as these 
areas are in extremely close proximity to each other and the studied species were 
shown to be not entirely self-fertilising following seedling analyses.  Overall FST 
values indicated that the groups of plants were more genetically differentiated than a 
random assemblage of the individuals, but this appeared to be mostly due to the far 
off site group, which in pairwise FST analysis was shown to be significantly 
differentiated from the other groups.  This is supported by the dendrogram based on 
measurements of Nei?s genetic distance, which showed far off site plants to be 
distant from plants from other areas.   
 
Following subdivision of I. adenoides groups (retaining wall, toepaddock wall, 
toepaddock, off site, far off site A, and far off site B), a more defined population 
structure became apparent.  Toepaddock wall plants were significantly genetically 
differentiated from retaining wall plants, despite the morphological similarities of 
the members of these groups, and the fact that they grow in what appears to be very 
similar substrates on the dam.  HS and Nm values were reduced with this analysis, 
indicating less gene flow between areas than was seen with the initial set of groups.  
These groups of plants therefore showed a fair degree of differentiation from each 
other.  This is supported by pairwise FST analysis, which showed the two far off site 
groups to be differentiated from all other groups, and from each other.  Toepaddock 
plants seemed to be most similar to toepaddock wall and off site plants.   
 
The sample sizes in the retaining and toepaddock wall groups were relatively small.  
This may have led to inaccurate indications of diversity in these groups.  Plants from 
these areas had a lower genetic diversity than those from the toepaddock, even 
though the retaining wall and toepaddock wall would be expected to have less 
 123 
contamination by slimes due to the raised ground and the presence of grassland soil 
in these areas.  The degree of contamination would be even lower in those regions of 
the toepaddock wall that were further away from the slimes dam.  Genetic diversity 
was low in plants from far off site A, probably due to the small area that this group 
was sampled from, when compared to the entire circumference of the dam used for 
sampling of toepaddock and off site groups, as shown on the sampling map (Figure 
6).  Far off site B plants were sampled from an even smaller area than those from far 
off site A, and yet plants from this area showed a high genetic diversity.  This is 
interesting as far off site B was the only group that can be considered relatively 
unaffected by slimes-associated contamination.   
 
In the overall population analysis of I. zeyheri, overall genetic diversity (HT) was 
marginally higher than the diversity within groups of plants from each area (HS).  
Estimated gene flow between areas was very high.  These values indicate very little 
genetic differentiation between groups of plants due to a high degree of gene flow 
between areas.  This is supported by FST analysis, which indicated that plants from 
the different areas were not more genetically differentiated than a random 
assemblage of the individuals.   
 
The smaller difference in genetic diversity between plants from the two areas 
observed with this species may therefore be due to high levels of gene flow which 
reduced any effect colonisation and natural selection may have had.  Alternatively, 
any plants on the toepaddock that had persisted from before the dam was established 
may not have been exposed to the disturbance for a long enough time to result in 
marked genetic changes.  Younger plants that germinated in the slimes may be more 
genetically differentiated.  The New North Complex dam is older (34 years old at 
the time of the study) than the South Complex dam (26 years old).  This may have 
led to a smaller difference in diversity between the two areas at the New North 
Complex dam, as the plants at this site had more generations in which gene flow 
could lead to an elimination of the low diversity caused by colonisation.  It would be 
necessary to perform a second study several years in the future to determine if this 
were the case. 
 
 124 
2. Differences in morphology between areas 
 
Morphological analysis of both species showed a large degree of differentiation 
between the areas.  Morphological differences may be environmentally caused or 
may be genetically inherited and a fundamental part of the population (Antonovics 
et al, 1971).  Morphological analysis may therefore indicate genetic changes in the 
selected genome ? the physiological manifestation of traits that have been selected 
for.  Alternatively, it could indicate phenotypic plasticity without a genetic basis.  It 
is impossible to know whether the phenotypic differences observed were based on 
genetic variation between plants from the different areas, or whether they constituted 
phenotypic plasticity without a genetic basis.  It is also impossible to determine 
whether the plasticity observed conferred fitness advantages on plants in the 
different areas or not.  Some of the traits appear to have the potential to provide an 
advantage to the plants, as will be discussed, but this is not necessarily the case.  
Morphological variation may have been due to inherited differences present in a 
small colonising population that provide no actual benefit to the plant.  
 
Differences in morphology were found between plants growing on different areas of 
the slimes dams.  Indigofera adenoides plants growing on the toepaddock grew in a 
low, spreading fashion when they achieved a certain (adult) size, and were found 
growing only in sparsely colonised, slimes inundated areas.  Indigofera zeyheri 
plants were found growing amongst other plants in highly colonised areas, and did 
not exhibit the spreading habit of I. adenoides.  Differences in canopy area were 
observed between the areas in both species.  Plant height differences were also 
observed in both species, although these differences were not statistically significant 
in I. zeyheri.  Leaf width differences were observed in I. adenoides only.  In contrast 
to genetic differences between the areas, morphological differences did not 
generally follow the same pattern in both species. 
 
Plants adapted to local environments have been shown to have a decreased 
competitive ability (Antonovics et al., 1971).  Therefore the absence of I. adenoides 
plants in highly colonised regions of the toepaddock may be an indication that 
adaptation and a subsequent reduction in competitive ability have occurred.  A 
 125 
spreading growth habit is also a possible indication of adaptation (Antonovics et al., 
1971).  The growth characteristics of I. zeyheri may therefore suggest that this 
species is less adapted to the toepaddock environment than I. adenoides, or that 
conditions at the New North Complex dam are less detrimental to plant growth than 
those at the South Complex dam.   
 
In I. adenoides, canopy area was found to be relatively constant in the toepaddock 
wall, retaining wall, toepaddock, and off site plants, while being significantly lower 
in the far off site plants.  This trend was followed with respect to canopy volume.  It 
would be expected that plants would show a greater size on toepaddock and 
retaining wall areas due to decreased competition allowing increased growth.  The 
noted trend of a more spreading growth on toepaddock soils would also lead to 
larger measurements of canopy area in these individuals.  This does not however 
explain the larger size in off site areas, which has higher levels of vegetation cover 
and therefore increased competition.   
 
A possible reason for the observed differences could therefore be that toepaddock, 
retaining wall, and off site plants were of larger size when compared to far off site 
plants due to increased water availability as water from newly added slimes would 
seep into these areas.  Toepaddock and retaining wall plants should therefore have 
been larger than off site plants, due to reduced competition.  These plants may, 
however, have been restricted in their growth due to the toxic component of an 
increase in slimes-contamination in these areas.   
 
This suggested hypothesis seems to be supported by canopy area and canopy volume 
measurements for I. zeyheri.  This species showed a marked reduction in canopy 
area and canopy volume in toepaddock plants when compared to off site plants.  
Since plants on both areas appeared to grow in sites that showed the same levels of 
colonisation, the levels of competition can be assumed to have been fairly equal and 
can be excluded from the comparison.  Water availability levels should also have 
been similar between the areas.  The difference in plant size may therefore have 
been entirely due to a difference in toxicity between the areas, with toepaddock 
plants being restricted in their growth.   
 126 
Plant height data can also be explained using the same hypothesis.  In I. adenoides, 
toepaddock area plants were shorter than plants in the retaining wall and off site 
areas, which are presumed to have lower toxicity due to the greater proportion of 
grassland soil present in these areas.  Toepaddock plants were however taller than 
far off site plants, presumably due to less competition and increased levels of water.  
Far off site areas have more competition due to increased colonisation, and may 
have decreased levels of available water.  In I. zeyheri, plant height was lower in off 
site plants, although not significantly.  This may indicate that soils at the New North 
Complex slimes dam are less toxic than those at the South Complex slimes dam.  It 
may also be an indication of the preferred growth conditions of the two species.  
Areas of the toepaddock are either inundated with slimes that has spilled out of the 
dam, or not inundated.  Indigofera adenoides appeared to prefer slimes-indundated 
areas of the dam, while I. zeyheri appeared to prefer non-inundated areas.  Therefore 
the local environment favoured by I. adenoides may be more toxic than that 
favoured by I. zeyheri.  
 
Leaves of many species are known to be highly plastic in response to environmental 
conditions (Witkowski and Lamont, 1991).  In I. adenoides, significant differences 
were found in leaf width between areas, with width tending to be larger in areas with 
a decrease in slimes-associated contamination.  Leaf size has been shown to 
decrease with increasing altitude, decreasing mean annual temperature, decreasing 
mean annual rainfall, and lower soil fertility (Givnish, 1987; McDonald et al., 2003).  
Few studies have been done on leaf width and what variables may affect it.  Within 
habitat variation in leaf size or shape may indicate adaptation to variations in 
ecological conditions within that habitat (Givnish, 1987).   
 
In I. adenoides, leaves were shown to be significantly narrower in retaining wall and 
toepaddock plants.  Altitude and rainfall are no different between the areas.  
Temperature may be slightly higher on the slimes dam than off due to less 
vegetation and therefore increased levels of radiation reaching the plants and 
reflecting off the slimes (white in colour) back at the plants.  Plants growing on 
slimes dams may therefore have had narrower leaves in order to prevent water loss 
within this increased temperature regime.  This hypothesis is supported by leaf 
 127 
width measurements in I. zeyheri, which were not siginificantly different in plants 
from different areas.  The New North Complex slimes dam has more vegetation 
cover, and appears to have higher rainfall levels and lower average temperatures.  
Temperature and radiation on and off the dam should therefore be similar. 
 
Dendrograms, ANOVA, and Tukey analyses showed far off site plants, especially 
far off site B samples, to be morphologically differentiated from the plants in the 
other areas.  This indicates that morphology of plants growing in the relative 
absence of slimes-contaminated soil is very different from that of plants growing in 
the presence of slimes-contaminated soil.  The differentiation of far off site B plants 
was not surprising as these individuals were sampled from a site seven kilometers 
away from the other individuals.  It is unlikely that gene flow occurs frequently 
between the two sites.  Similar classification patterns were found with I. zeyheri 
individuals, with toepaddock classifying to itself, but off site classifying equally to 
toepaddock and off site.  Classification analysis suggests that toepaddock plants 
consist of a subset of off site plants.  This is expected since colonisation of slimes-
 contaminated soil was assumed to have occurred by plants from the surrounding 
areas.  
 
3. Genetic and morphological differentiation 
 
Genetic and morphological data did not follow similar patterns in most instances.  In 
I. adenoides, dendrograms for genetic data showed plants from far off site A and B 
to be the most differentiated from plants in the other groups, while dendrograms for 
morphological data showed plants from far off site B to be highly differentiated 
from plants in the other groups, while far off site A plants were not.  Morphological 
differentiation patterns probably indicate the differences in micro-climate and biotic 
habitat conditions between the greater slimes dam area, and the nature reserve that 
was the site of sampling for far off site B.   
 
Morphological variation was highest in plants from the toepaddock for most of the 
single measurements, and highest in plants from far off site A for most of the 
composite variables.  This is in contrast to genetic variation, which was lowest in the 
 128 
toepaddock and associated areas, and in far off site A.  This pattern was repeated in 
I. zeyheri, with toepaddock plants showing the highest morphological variation and 
the lowest genetic variation.  High morphological variation is likely to be an 
indication of adaptation to different conditions within the toepaddock, for example, 
exposed or not, and slimes-inundated or not.   
   
Indigofera adenoides classification analyses gave consistent results with genetic and 
morphological data.  Classification based on morphological data gave the highest 
percentage of correct classifications in toepaddock and far off site B plants, while 
genetic classification gave the highest percentage of correct classifications in 
toepaddock and far off site A plants.  Far off site A plants were therefore genetically 
differentiated from plants in the other groups, but not morphologically 
differentiated, presumably due to the similar climate in far off site A and the slimes 
dam areas.  Toepaddock plants were shown to be most similar to themselves.  This 
indicates the morphological and genetic differentiation of these plants from those in 
the other areas, due to the unique environmental conditions in the toepaddock. 
 
It would seem that I. adenoides has adapted to toepaddock conditions more 
successfully than I. zeyheri.  This was shown by I. adenoides plants growing only in 
sparsely colonised areas of the toepaddock, indicating a reduced competitive ability 
which may be indicative of adaptation (Hangelbroek et al, 2003), as well as by the 
prostrate, spreading growth habit seen on the toepaddock, which has been suggested 
to be an adaptation to low nutrient availability and exposed conditions (Antonovics 
et al., 1971).  It cannot be determined, however, whether these differences are 
heritable differences, phenotypic plasticity, or a combination of the two. 
 
In summary, high degrees of phenotypic plasticity existed between the different 
areas.  It does not seem likely that all of this variation can be accounted for by 
genetic diversity, given the high levels of gene flow between areas and the 
comparatively low differences in genetic diversity.  There is therefore a degree of 
non-genetic phenotypic plasticity, which may or may not confer an advantage to 
growth on slimes-contaminated soil.  Morphological and genetic analyses indicated 
less differentiation between areas, and higher levels of gene flow, in I. zeyheri.  This 
 129 
may indicate less toxic conditions at the New North Complex slimes dam, or it may 
indicate a reduced ability to grow on toxic soils by this species when compared to I. 
adenoides.   
 
4. Analysis according to groundwater toxicity 
 
Analysis according to groundwater toxicity indicated little effect of groundwater 
toxicity on genetic or morphological diversity.  Toxicity group analysis was 
performed based on data on the contamination of groundwater in the areas studied.  
I. adenoides plants were divided into six groups, with group 2 having the lowest 
groundwater toxicity, and group 7 having the highest.  Initial analysis was 
performed containing all individuals, i.e., those from the toepaddock, retaining wall, 
off site and toepaddock wall.  This was shown to result in no or very small 
morphological differences between the sites.   
 
This was thought to be due to differences in growth conditions between the various 
areas.  For example, the toepaddock has more available water than the other areas, 
due to seepage from the slimes dams, as well as less competition due to sparser 
vegetation cover.  In contrast, off site areas have higher vegetation cover and 
therefore a higher organic matter content in the soil than the toepaddock and 
associated areas (Witkowski and Weiersbye, 1998).  These area-specific differences 
would result in area-specific growth patterns, as was seen in morphological analysis 
of the area groups discussed above.  These differences would result in a high within-
 group variation in morphological traits, which would result in any toxicity-related 
differences in morphology being undetectable.  
 
Toxicity analysis was therefore performed with toepaddock plants only, in order to 
determine if a detectable difference could be noted between toxicity groups once the 
masking effect of proximity to the slimes dam was removed.  In morphological 
analysis of I. adenoides, groups 6 and 7 were grouped together, due to the low 
sample sizes of the individual groups.  Plants in group 4, from a moderate toxicity 
area, were found to have over twice the canopy area and canopy volume of plants 
from the other groups.  Plants from this group classified accurately with the highest 
 130 
percentage (75%), and were shown to be morphologically distant from the other 
groups on a dendrogram. Groups 2 and 3 were found to contain the most 
morphologically variable plants. 
 
These results are confusing and seem to conflict with the results obtained from 
grouping plants by area.  In area classification, toepaddock plants, which are 
presumed to grow on the highest toxicity, showed the highest morphological 
variability.  This trend was followed in toxicity group analysis of I. zeyheri, where 
plants in the higher toxicity groups were found to have higher variation of most of 
the variables, especially canopy area and canopy volume, which were twice as 
variable in the higher toxicity group.  I. zeyheri individuals were also found to be 
taller, and have increased variability in height, in the higher toxicity areas.  This 
pattern was found in the area groups in this species.   
 
A hypothesis has been developed to explain the results found in I. zeyheri.  
Evapotranspiration can cause contaminated groundwater to be drawn to the surface 
if the groundwater depth is shallow.  Groundwater levels in the New North Complex 
area were, however, no higher than 4 metres below ground level.  It is unlikely that 
the root systems of these plants would reach these levels, as they appear to be 
relatively shallow-rooted.  Plants dug up were mature and obviously several years 
old, and had root systems only reaching approximately 60 cm deep.  Seepage from 
the dams would also reduce evapotranspiration of the ground water due to it 
increasing the water level in the soil.   
 
Indigofera zeyheri higher toxicity plants were found growing mostly in between two 
slimes dams, as was shown in the sulphate zone diagram (Figure 29).  This area 
could have higher available water due to seepage from dams on either side, and also 
may be slightly protected from wind.  This would result in both taller plants, and in a 
larger population size, which would result in higher genetic diversity.  This is 
supported by genetic data, which showed a marked increase in genetic variation in 
the higher toxicity group, which could be due to a larger population size.  The 
influence of the contamination level in the groundwater may therefore be 
 131 
insignificant when compared to direct contamination by slimes runoff from the 
dams.   
 
This hypothesis is not as easy to apply to results for I. adenoides.  Samples were 
taken from all aspects of the South Complex dam, and ground water toxicity zones 
span the dam completely, making it unlikely that aspect would be responsible for 
differences in morphology.  The groundwater levels in this area are very deep ? over 
14 metres below ground level.  The contamination levels of the ground water are 
therefore unlikely to have influenced the plants growing on the toepaddock.   
 
It can be postulated that the differences in sample size influenced variation in I. 
adenoides.  It was noted that fewer plants were obtained in the higher toxicity 
groups, indicating the presence of fewer plants in these areas.  This could reduce the 
variation of the plants in these areas due to a decreased gene pool within which gene 
flow can occur.  This hypothesis is supported by genetic data, which showed low 
genetic variation in group 6/7.  Genetic variation followed a pattern of decreasing 
slightly with increasing ground water toxicity up to group 4, then increasing 
dramatically in group 5, or group 5/6/7, and then decreasing dramatically to group 
6/7, if measured.  Low diversity in group 6/7 could simply be due to the smaller 
sample size of this group, while high diversity in group 5/6/7 could be due to the 
large area this group was sampled from.  Gene flow probably occurred from the 
surrounding areas, so plants taken from a large area would have had a larger 
population to obtain genes from.  This would increase the diversity of these groups.   
 
FST values were significant for both populations, indicating that although genetic 
differentiation between groups was low, it is still statistically significant.  Groups 
6/7 and 5/6/7 were shown to be significantly different from the other groups, which 
was emphasised by dendrograms.  It is possible that more gene flow from outside 
sources occurred in these areas.  Fewer plants were found in these areas, indicating 
the population size may be small, which would result in the plants receiving more 
genes from outside areas.   
 
 132 
Aerial photographs showed another slimes dam a short distance to the east of the 
South Complex slimes dam, close to the higher groundwater toxicity zones (groups 
6 and 7).  This dam could influence the surface toxicity of the soil in these areas due 
to slimes runoff and slimes dust blown over the area, which could account for the 
smaller population sizes noted.  Plants in group 4 could therefore have been the 
largest in terms of canopy area because of increased water availability due to 
seepage from this dam, and showed a higher morphological diversity due to higher 
toxicity in this area, as was noted in the area groups.  The lack of genetic diversity 
and sample size in groups 5, 6, and 7 may be because toxicity reached a critical level 
in these areas due to the presence of two slimes dams, so that sufficient growth and 
population size may not have been maintained.  Plants were the shortest in this area, 
which may indicate stunting due to high toxicity and exposed growing conditions 
due to a lack of colonisation.   
 
Any differences in morphology and genetics due to levels of toxicity in groundwater 
therefore seem to have been masked by differences in surface toxicity due to 
pollution from slimes dams, or other climatic, biotic, or edaphic factors.  It would be 
necessary to repeat the study using data on surface soil toxicity in order to determine 
the effect local changes in toxicity have on morphological and genetic variation in 
these species.  
 
5. Applications and further study 
 
Slimes-contaminated soils had both a morphological and genetic effect on the 
Indigofera species studied.  This appears to indicate some degree of adaptation to 
slimes-contaminated soil, either through phenotypic plasticity or adaptive genetic 
differences.  The species studied therefore seem suitable as candidates for vegetation 
of slimes dams as well as the revegetation of the surrounding polluted soils. 
 
The lower diversity observed in the toepaddock and associated areas should not have 
a detrimental effect on populations used for vegetating, as diversity was still 
comparatively high in these groups when compared to other plant populations 
studied (Jacquemyn et al., 2004; Sol? et al., 2004).  It would be important to 
 133 
maintain the diversity that exists, however.  This can be done by using a large 
population size to vegetate dams, in order to avoid further reductions in diversity 
through genetic drift due to small population sizes.  Vegetation should be performed 
by taking plants that have naturally colonised the toepaddock and associated areas 
and using them, and their seeds, as starting populations for vegetation.  This would 
ensure that any specialised genotypes that may exist in the toepaddock-adapted 
populations were included in initial vegetation populations. 
 
Further studies that would be useful in determining the suitability of these species as 
candidates for vegetation would include the germination of seeds obtained from 
toepaddock plants and the analysis of subsequent progeny under controlled, non-
 toxic conditions to determine if the morphological differences observed are due to 
plasticity or heritable differences.  Further long-term genetic diversity studies could 
also be performed to determine if diversity of individuals on the toepaddock 
increases or decreases further with time.  Complete breeding system studies could 
also be performed.  
 
The lack of I. zeyheri individuals far away from the dam could indicate this species 
only grows in the sheltered, higher water availability areas around the dams.  
Alternatively, individuals of this species may exist far away from the dams, but 
could not be located.  Indigofera zeyheri needs to be studied further in order to gain 
a more in depth picture of this species? response to slimes-associated contamination.  
This analysis should include comparing slimes dam populations to a far off site 
population if one can be located.   
 
Further study would be necessary to determine the effect of local changes in soil 
toxicity on these species.  If a loss of genetic diversity occurs and persists with 
increasing toxicity, the overall adaptive ability of the population would be reduced.  
Soil toxicity data are needed in order to perform a more accurate study of 
morphological and genetic variation in response to soil toxicity.     
 
 
 
 134 
6. Conclusions 
 
1. Lower genetic diversity was observed in those areas around the dams with 
higher levels of slimes-associated contamination 
2. Reduced morphological variation was observed in those areas around the 
dams with higher levels of slimes-associated contamination 
3. Morphological differences were observed between the areas, some of which 
appear to provide some benefit to the plants 
4. The above differences were all observed in both species 
5. Some degree of adaptation to slimes-contaminated soil therefore seems to 
have occurred, with this being more pronounced in I. adenoides, although it 
cannot be determined whether this is purely phenotypic, or a combination of 
phenotypic and genetic 
6. These species therefore seem suitable as candidates for vegetation of slimes 
dams, although further work must be done to determine the effect of surface 
toxicity on these species 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 135 
REFERENCES  
 
Allen, O. N. and Allen, E. K. (1981). The Leguminosae ? a source book of 
characteristics, uses, and nodulation. Macmillan, London, 812pp. 
 
Antonovics, J., Bradshaw, A. D., and Turner, R. G. (1971). Heavy metal tolerance in 
plants. Advances in Ecological Research 7: 1-85. 
 
Bagley, M. J., Franson, S. E., Christ, S. A., Waits, E. R., and Toth, G. P. (2002). 
Genetic diversity as an indicator of ecosystem condition and sustainability: Utility 
for regional assessments of stream condition in the eastern united states. US 
environmental protection agency, Cincinatti, 154pp. 
 
Barrett, B. A., Kidwell, K. K., and Fox, P. N. (1998). Comparison of AFLP and 
pedigree-based genetic diversity assessment methods using wheat cultivars from the 
Pacific northwest. Crop Science 38: 1271-1278. 
 
Beardsley, P. M., Yen, A., and Olmstead, R. G. (2003). AFLP phylogeny of 
Mimulus section Erythranthe and the evolution of hummingbird pollination.  
Evolution 57(6): 1397-1410. 
 
Bednarek, P. T., Lewandowska, R., Golas, T., and Pa?nik, M. (2003). The 
chromosomal location of rye AFLP bands. Cellular and Molecular Biology Letters  
8: 955-961. 
 
Blackith, R. E. and Reyment, R. A. (1971). Multivariate morphometrics. Academic 
press, London, 412pp. 
 
Blum, H., Beier, H., and Gross, H. J. (1987). Improved silver staining of plant 
proteins, RNA, and DNA in polyacrylamide gels. Electrophoresis 8: 93-99. 
 
Bradshaw, A. D. (1995). Alternative end points for reclamation. In: Cairns, J. (Ed.).  
Rehabilitating damaged ecosystems. CRC press, Boca Raton, FL, 69-86. 
 136 
 
Bush, E. J. and Barrett, S. C. H. (1993). Genetics of mine invasions by Deschampsia 
cespitosa (Poaceae). Canadian Journal of Botany 71: 1336-1348. 
 
Campbell, D. R., Rochefort, L., and Lavoie, C. (2003). Determining the immigration 
potential of plants colonising disturbed environments: the case of milled peatlands in 
Quebec. Journal of Applied Ecology 40:78-91.  
 
Castiglioni, P., Ajmone-Marsan, P, van Wijk, R., and Motto, M. (1999). AFLP 
markers in a molecular linkage map of maize: codominant scoring and linkage group 
distribution. Theoretical Applied Genetics 99: 425-431. 
 
Clausen, J., Keck, D. D., and Hiesey, W. M. (1948). Experimental studies on the 
nature of species. III: Environmental responses of climatic races of Achillea. 
Carnegie institute, Washington D.C., 128pp. 
 
Clausen, J. (1951). Stages in the evolution of plant species. Cornell University Press, 
New York, 214pp. 
 
Clegg, M. T. (1997). Plant genetic diversity and the struggle to measure selection. 
Journal of Heredity 88(1): 1-7. 
 
Coart, E., Vekemans, X., Smulders, M. J. M., Wagner, I., Van Huylenbroeck, J., 
Van Bockstaele, E., and Rold?n-Ruiz, I. (2003). Genetic variation in the endangered 
wild apple (Malus sylvestris (L.) Mill.) in Belgium as revealed by amplified 
fragment length polymorphism and microsatellite markers. Molecular Ecology 12: 
845-857. 
 
Cooke, J. A. and Johnson, M. S. (2002). Ecological restoration of land with 
particular reference to the mining of metals and industrial minerals: A review of 
theory and practice. Environmental Reviews 10: 41-71. 
 
 137 
Cooley, W. W. and Lohnes, P. R. (1971). Multivariate data analysis. John Wiley and 
Sons, New York, 364pp. 
 
Coyne, C. J., Pilet, M. L., Deniot, G., Baranger, A., Prioul, S., and McClendon, M. 
T. (2000). Selected AFLP primer pairs for fine mapping in pea. Pisum Genetics 32: 
27-30. 
 
de Muro, M. A., Mehta, S., and Moore, D. (2003). The use of amplified fragment 
length polymorphism for molecular analysis of Beauveria bassiana isolates from 
Kenya and other countries, and their correlation with host and geographical origin.  
FEMS Microbiology Letters 229: 249-257. 
 
de Riek, J., Calsyn, E., Everaert, I., Van Bockstaele, E.  (2001). AFLP based 
alternatives for the assessment of distinctness, uniformity, and stability of sugar beet 
varieties. Theoretical Applied Genetics 103: 1254-1265. 
 
Doignon-Bourcier, F., Willems, A., Coopman, R., Laguerre, G., Gillis, M., and de 
Lajudie, P. (2000). Genotypic characterization of Bradyrhizobium strains nodulating 
small Senegalese legumes by 16S-23S rRNA intergenic gene spacers and amplified 
fragment length polymorphism fingerprint analyses. Applied and Environmental 
Microbiology 66(9): 3987-3997.  
  
Duchesne, P., and Bernatchez, L. (2002). AFLPOP: A computer program for 
simulated and real population allocation based on AFLP data. Molecular Ecology 
Notes 3: 380-383. 
 
Eckhart, V. M., Geber, M. A., and McGuire, C. M. (2004). Experimental studies of 
adaptation in Clarkia xantiana. I: Sources of trait variation across a subspecies 
border. Evolution 58(1): 59-70. 
 
Ernande, B. and Dieckmann, U. (2004). The evolution of phenotypic plasticity in 
spatially structured environments: implications of intraspecific competition, 
 138 
plasticity costs, and environmental characteristics. Journal of Evolutionary Biology  
17: 613-628. 
 
Faugeron, S., Valero, M., Destombe, C., Martinez, E. A., and Correa, J. A. (2001). 
Hierarchical spatial structure and discriminant analysis of genetic diversity in the red 
alga Mazzaella laminarioides (Gigartinales, Rhodophyta). Journal of Phycology 37: 
705-716. 
 
Felsenstein, J. (2004). PHYLIP (Phylogeny Inference Package) version 3.6.  
Distributed by the author. Department of Genome Sciences, University of 
Washington, Seattle, USA. 
 
Fiedler, P. L. (1992). Cladistic test of the adaptational hypothesis for serpentine 
endemism. In: Baker, A. J. M., Proctor, J., and Reeves, R. D. (Ed.). The Vegetation 
of Ultramafic (Serpentine) Soils. Intercept, UK, 421-434. 
 
Giang, L. H., Hong, P. N., Tuan, M. S., and Harada, K. (2003). Genetic variation of 
Avicennia marina (Forsk.) Vierh. (Avicenniaceae) in Vietnam revealed by 
microsatellite and AFLP markers. Genes and Genetic Systems 78: 399-407. 
 
Gibson, J. P. and Pollard, A. J. (1988). Zinc tolerance in Panicum virgatum L. 
(switchgrass) from the Picher Mine area. Proceedings of the Oklahoma Academy of 
Science 68: 45-49. 
 
Givnish, T. J. (1987). Comparative studies of leaf form: assessing the relative roles 
of selective pressures and phylogenetic constraints. New Phytologist 106: 131-160. 
 
Gobert, V., Moja, S., Colson, M., and Taberlet, P. (2002). Hybridisation in the 
section Mentha (Lamiaceae) inferred from AFLP markers. American Journal of 
Botany 89(12): 2017-2023. 
 
 139 
Hamrick, J. L. (1989). Isoenzymes and the analysis of genetic structure in plant 
populations.  In: Soltis, D. E. and Soltis, P. S. (Ed.). Isozymes in plant biology. 
Chapman and Hall, UK, 87-105. 
 
Han, K. and Ely, B. (2002). Use of AFLP analyses to assess genetic variation in 
Morone and Thunnus species. Marine Biotechnology 4: 141-145. 
 
Hangelbroek, H. H., Santamaria, L., and De Boer, T. (2003). Local adaptation of the 
pondweed Potamogeton pectinatus to contrasting substrate types mediated by 
changes in propagule provisioning. Journal of Ecology 91: 1081-1092. 
 
Hartl, D. L. and Clark, A. G. (1989). Principles of population genetics (2nd edition).  
Sinauer associates, Mass, 682pp. 
 
Haynes, R. J. and Graham, M. H. (2004). Soil biology and biochemistry ? a new 
direction for South African soil science? South African Journal of Plant Soil 21(5): 
330-344. 
 
He, T., Krauss, S. L., Lamont, B. B., Miller, B. P., and Enright, N. J. (2004). Long 
distance seed dispersal in a metapopulation of Banksia hookeriana inferred from a 
population allocation analysis of AFLP data. Molecular Ecology 13(5): 1099-1109. 
 
Hillis, D. M., Moritz, C., and Mable, B. K. (Ed.). (1996). Molecular systematics (2nd 
edition). Sinauer associates, Massachusetts, 655pp. 
 
Innan, H., Terauchi, R., Kahl, G., and Tajima, F. (1999). A method for estimating 
nucleotide diversity from AFLP data. Genetics 151: 1157-1164. 
 
Jacquemyn, H., Honnay, O., Galbusera, P., and Rold?n-Ruiz, I. (2004). Genetic 
structure of the forest herb Primula elatior in a changing landscape. Molecular 
Ecology 13: 211-219. 
 
 140 
Jones, C. J., Edwards, K. J., Castaglione, S., Winfield, M. O., Sala, F., van de Wiel, 
C., Bredemeijer, G., Vosman, B., Matthes, M., Daly, A., Brettschneider, R., Bettini, 
P., Buiatti, M., Maestri, E., Malcevschi, A., Marmiroli, N., Aert, R., Volckaert, G., 
Rueda, J., Linacero, R., Vazquez, A., and Karp, A. (1997). Reproducibility testing of 
RAPD, AFLP and SSR markers in plants by a network of European laboratories.  
Molecular Breeding 3: 381-390. 
 
Joshi, J., Schmid, B., Caldeira, M. C., Dimitrakoppolos, P. G., Good, J., Harris, R., 
Hector, A., Huss-Danell, K., Jumpponen, A., Minns, A., Mulder, C. P. H., Pereira, J. 
S., Prinz, A.,  Scherer-Lorenzen, M., Siamantziouras, A. S. D., Terry, A. C., 
Troumbis, A. Y., and Lawton, J. H. (2001). Local adaptation enhances performance 
of common plant species. Ecology Letters 4: 536-544. 
 
Jover, M. A., Castillo-Agudo, L., Garcia-Carrascosa, M., and Segura, J. (2003). 
Random amplified polymorphic DNA assessment of diversity in Western 
Mediterranean populations of the seagrass Posidonia oceanica. American Journal of 
Botany 90(3): 364-369. 
 
Kittelson, P. M. and Maron, J. L. (2001). Fine-scale genetically based differentiation 
of life-history traits in the perennial shrub Lupinus arboreus. Evolution 55(12): 
2429-2438. 
 
Knoll, C. (1998). Rehabilitation with vetiver. African mining 2(2): 43.  
 
Knight, T. M. and Miller, T. E. (2004). Local adaptation within a population of 
Hydrocotyle bonariensis. Evolutionary Ecology Research 6: 103-114. 
 
Lamour, K. H. and Hausbeck, M. K. (2001). Investigating the spatiotemporal 
genetic structure of Phytophthora capsici in Michigan. Phytopathology 91(10): 973-
 980. 
 
Lesica, P. and Allendorf, F. W. (1999). Ecological genetics and the restoration of 
plant communities: Mix or match? Restoration ecology 7(1): 42-50. 
 141 
 
Leteinturier, B., Baker, A. J. M., and Malaisse, F. (1999). Early stages of natural 
revegetation of metalliferous mine workings in South Central Africa: a preliminary 
survey. Biotechnology, Agronomy, Society, and Environment 3(1): 28-41. 
 
Lopes, I., Baird, D. J., and Ribeiro, R. (2004). Genetic determination of tolerance to 
lethal and sublethal copper concentrations in field populations of Daphnia 
longispina. Archives of Environmental Contamination and Toxicology 46: 43-51. 
 
Lu, Z., Sosinski, B., Reighard, G. L., Baird, W. V., and Abbott, A. G. (1998). 
Construction of a genetic linkage map and identification of AFLP markers for 
resistance to root-knot nematodes in peach rootstocks. Genome 41: 199-207. 
 
Lynch, M. and Milligan, B. G. (1994). Analysis of population genetic structure with 
RAPD markers. Molecular Ecology 3: 91-99. 
 
Macnair, M. R. (1983). The genetic control of copper tolerance in the yellow 
monkey flower, Mimulus guttatus. Heredity 50(3): 283-293. 
 
Macnair, M. R. (1992). Preliminary studies on the genetics and evolution of the 
serpentine endemic Mimulus nudatus Curran. In: Baker, A. J. M., Proctor, J., and 
Reeves, R. D. (Ed.). The Vegetation of Ultramafic (Serpentine) Soils. Intercept, 
Great Britain, 409-419.  
 
Macnair, M. R., Cumbes, Q. J., and Smith, S. (1993). The heritability and 
distribution of variation in degree of copper tolerance in Mimulus guttatus on a 
copper mine at Copperopolis, California. Heredity 71: 445-455.  
 
McDonald, P. G., Fonseca, C. R., Overton, J. M., and Westoby, M. (2003). Leaf-size 
divergence along rainfall and soil-nutrient gradients: is the method of size reduction 
common among clades? Functional Ecology 17: 50-57. 
 
 142 
Mengoni, A., Barabesi, C., Gonnelli, C., Galardi, F., Gabbrielli, R., and 
Bazzicalupo, M. (2001). Genetic diversity of heavy metal-tolerant populations in 
Silene paradoxa L. (Caryophyllaceae): a chloroplast microsatellite analysis. 
Molecular Ecology 10: 1909-1916. 
 
Milne-Redhead, E. and Polhill, R. M. (Ed.). (1952). Flora of tropical East Africa.  
Kew publishing, UK, 893pp. 
 
Miyashita, N. T., Kawabe, A., and Innan, H. (1999). DNA variation in the wild plant 
Arabidopsis thaliana revealed by amplified fragment length polymorphism analysis.  
Genetics 152: 1723-1731. 
 
Morris, J. B. (1999). Legume genetic resources with novel ?value added? industrial 
and pharmaceutical use. In: Janick, J. (Ed.). Perspectives on new crops and new 
uses. ASHS Press, Alexandria, VA, 196-201. 
 
Muluvi, G. M., Sprent, J. I., Soranzo, N., Provan, J., Odee, D., Folkard, G., 
McNicol, J. W., and Powell, W. (1999). Amplified fragment length polymorphism 
(AFLP) analysis of genetic variation in Moringa oleifera Lam. Molecular ecology  
8: 463-470. 
 
Najimi, B., Boukhatem, N., Jaafari, S. E., Jlib?ne, M., Paul, R., and Jacquemin, J. 
(2002). Amplified fragment length polymorphism (AFLP) analysis of markers 
associated with H5 and H22 Hessian fly resistance genes in bread wheat.  
Biotechnology, Agronomy, Society, and Environment 6(2): 79-85. 
 
Nei, M. (1973). Analysis of gene diversity in subdivided populations. Proceedings 
of the National Academy of Science 70(12): 3321-3323. 
 
Page, R. D. M. (1996). TREEVIEW: An application to display phylogenetic trees on 
personal computers. Computer Applications in the Biosciences 12: 357-358. 
 
 143 
Peakall, R. and Smouse, P. E. (2001). GenAlEx: Genetic Analysis in Excel.  
Population genetic software for teaching and research. Australian National 
University, Canberra, Australia. www.anu.edu.au/BoZo/GenAlEx/ 
 
Price, T. D., Qvarnstr?m, A., and Irwin, D. E. (2003). The role of phenotypic 
plasticity in driving genetic evolution. Proceedings of the Royal Society of London 
270: 1433-1440. 
 
Prus-Glowacki, W. and Godzik, St. (1991). Changes induced by zinc smelter 
pollution in the genetic structure of pine (Pinus sylvestris L.) seedling populations.  
Silvae genetica 40: 184-188. 
 
Prus-Glowacki, W. and Godzik, St. (1995). Genetic structure of Picea abies trees 
tolerant and sensitive to industrial pollution. Silvae Genetica 44: 62-65 
 
Prus-Glowacki, W., Wojnicka-Poltorak, A., Oleksyn, J., and Reich, P. B. (1999).  
Industrial pollutants tend to increase genetic diversity: Evidence from field-grown 
European Scots pine populations. Water, Air, and Soil pollution 116: 395-402. 
 
Reed, D. H., Lowe, E. H., Briscoe, D. A., and Frankham, R. (2003). Fitness and 
adaptation in a novel environment: effect of inbreeding, prior environment, and 
lineage. Evolution 57(8): 1822-1828. 
 
Reznick, D. N. and Ghalambor, C. K. (2001). The population ecology of 
contemporary adaptations: what empirical studies reveal about the conditions that 
promote adaptive evolution. Genetica 112-113: 183-198. 
 
Rossum, F. V., Bonnin, I., F?nart, S., Pauwels, M., Petit, D. and Saumitou-Laprade, 
P. (2004). Spatial genetic structure within a metallicolous population of Arabidopsis 
halleri, a clonal, self-incompatible and heavy-metal-tolerant species. Molecular 
Ecology 13: 2959-2967. 
 
 144 
Rout, G. R. and Das, P. (2003). Effect of metal toxicity on plant growth and 
metabolism: I. Zinc. Agronomie 23: 3-11. 
 
SAS Volume 8. (2000). SAS Institute, Cary, N. Carolina, USA. 
 
Schoeman, J. L. and van Deventer, P. W. (2004). Soils and the environment: the past 
25 years. South African Journal of Plant Soil 21(5): 369-387. 
 
Shaw, J., Antonovics, J., and Anderson, L. E. (1987). Inter- and intraspecific 
variation of mosses in tolerance to copper and zinc. Evolution 41(6): 1312-1325. 
 
Shaw, A. J. (1994). Adaptation to metals in widespread and endemic plants.  
Environmental Health Perspectives 102(12): 105-108. 
 
Silander, J. A. and Antonovics, J. (1979). The genetic basis of the ecological 
amplitude of Spartins patens. I: Morphometric and physiological traits. Evolution 
33(4): 1114-1127. 
 
Sledge, M. K., Bouton, J. H., Dall?Agnoll, M., Parrott, W. A., and Kochert, G.  
(2002). Identification and confirmation of aluminium tolerance QTL in diploid 
Medicago sativa subsp. coerulea. Crop Science 42: 1121-1128. 
 
Sn?ll, T., Fogelqvist, J., Ribeiro, P. J., and Lascoux, M. (2004). Spatial genetic 
structure in two congeneric epiphytes with different dispersal strategies analysed by 
three different methods. Molecular ecology 13: 2109-2119.   
 
Sol?, M., Durka, W., Eber, S., and Brandl, R. (2004). Genotypic and genetic 
diversity of the common weed Cirsium arvense (Asteraceae). International Journal 
of Plant Science 165(3): 437-444. 
 
Stehlik, I. (2002). Glacial history of the alpine herb Rumex nivalis (Polygonaceae): a 
comparison of common phylogeographic methods with nested clade analysis. 
American Journal of Botany 89(12): 2007-2016. 
 145 
 
Storz, J. F., Ramakrishnan, U., and Alberts, S. C. (2001). Determinants of effective 
population size for loci with different modes of inheritance. Heredity 92(6): 497-
 502. 
 
Szczepaniak, M., Cie?lak, E., and Bednarek, P. T. (2002). Morphological and AFLP 
variation of Elymus repens (L.) gould (Poaceae). Cellular and Molecular Biology 
Letters 7: 547-558. 
 
Terashima, K. and Matsumoto, T. (2004). Strain typing of shiitake (Lentinula 
edodes) cultivars by AFLP analysis, focusing on a heat-dried fruiting body.  
Mycoscience 45: 79-82. 
 
Thomas, C. M., Vos, P., Zabeau, M., Jones, D. A., Norcott, K. A., Chadwick, B. P., 
and Jones, J. D. G. (1995). Identification of amplified restriction fragment 
polymorphism (AFLP) markers tightly linked to the tomato Cf-9 gene for resistance 
to Cladosporium fulvum. Plant Journal 8(5): 785-794. 
 
Ticknor, L. O., Kolst?, A., Hill, K. K., Keim, P., Laker, M. T., Tonks, M., Jackson, 
P. J. (2001). Fluorescent amplified fragment length polymorphism analysis of 
Norwegian Bacillus cereus and Bacillus thuringiensis soil isolates. Applied and 
Environmental Microbiology 67(10): 4863-4873.   
 
Tribsch, A., Schonswetter, P., and Stuessy, T. F. (2002). Saponaria pumila 
(Caryophyllaceae) and the ice age in the European alps. American Journal of Botany 
89(12): 2024-2033. 
 
Valladares, F., Balaguer, L., Martinez-Ferri, E., Perez-Corona, E., and Manrique, E.  
(2002). Plasticity, instability, and calanization: Is the phenotypic variation in 
seedlings of sclerophyll oaks consistent with the environmental unpredictability of 
Mediterranean ecosystems? New Phytologist 156: 457-467. 
 
 146 
Van Cutsem, P., du Jardin, P., Boutte, C., Beauwens, T., Jacqmin, S., and 
Vekemans, X. (2003). Distinction between cultivated and wild chicory gene pools 
using AFLP markers. Theoretical Applied Genetics 107: 713-718. 
 
Vanhala, T. K., van Rijn, C. P. E., Buntjer, J., Stam, P., Nevo, E., Poorter, H., and 
van Eeuwijk, F. A. (2004). Environmental., phenotypic, and genetic variation of 
wild barley (Hordeum spontaneum) from Israel. Euphytica 137: 297-309. 
  
Van Wyk, B. and Malan, S. (1998). Field guide to the wild flowers of the highveld.  
Struik publishers, Cape Town, 352pp. 
 
Vekemans, X., Beauwens, T., Lemaire, M., and Roldan-Ruiz, I. (2002). Data from 
amplified fragment length polymorphism (AFLP) markers show indication of size 
homoplasy and of a relationship between degree of homoplasy and fragment size.  
Molecular ecology 11: 139-151. 
 
Vellend, M. and Waterway, M. J. (1999). Geographic patterns in the genetic 
diversity of a northern sedge, Carex rariflora. Canadian Journal of Botany 77: 269-
 278. 
 
Volk, G. M., Henk, A. D., and Richards, C. M. (2004). Genetic diversity among 
U.S. garlic clones as detected using AFLP methods. Journal of the American Society 
for Horticultural Science 129(4): 559-569. 
 
Vos, P., Hogers, R., Bleeker, M., Reijans, M., van de Lee, T., Hornes, M., Frijters, 
A., Pot, J., Peleman, J., Kuiper, M., and Zaeau, M. (1995). AFLP: a new technique 
for DNA fingerprinting. Nucleic Acids Research 23 (21): 4407-4414. 
 
Weiersbye, I. M. and Witkowski, E. T. F. (2002). Seed fate and practical 
germination methods for 46 perennial species that colonise gold mine tailings and 
acid mine drainage-polluted soils in the grassland biome. In: Seydack, A. H. W., 
Vorster, T., Vermeulen, W. J., and van der Merwe, I. J. (Ed.). Multiple use 
management of natural forests and woodlands: policy refinements and scientific 
 147 
progress. Natural forests and Savanna Woodlands Symposium III, KNP. Department 
of Water Affairs and Forestry indigenous forest management, Pretoria, 221-255. 
 
Weiersbye, I. M., Witkowski, E. T. F., Dye, T., Vivier, J. J. K., van Rensburg, H. J., 
Herbert, M., Amis, J., Parsons, J., Reicharet, M., Fourie, L., Holmwood, R., van 
Wyk, A. (2002).  The containment of pollution from tailings dams; the sustainable 
vegetation of slimes dams project and the mine woodlands project. Invited review 
and discussion document. WITS_GOVT_20/07/02V2 to the Directorate of 
Community Forestry, Department of Water Affairs and Forestry, and the Directorate 
of Mine Rehabilitation, Department of Minerals and Energy, 71pp. 
 
Weiersbye, I. M., Witkowski, E. T. F., and Reichardt, M. (2005). The flora of gold 
and uranium slimes dams and polluted soils of South Africa?s deep-level mines. 
Bothalia In press. 
 
Weir, B. S. (1996). Genetic data analysis II. Sinauer associates, MA, 445pp. 
 
Westerbergh, A. and Saura, A. (1992). Serpentine and the population structure of 
Silene dioica (L.) Clairr. (Caryophyllaceae). Evolution 46(5): 1537-1548. 
 
Williamson, S. D., Robinson, E. R., and Balkwill, K. (1997). Evolution of two 
serpentine endemic taxa in Mpumalanga. South African Journal of Botany 63(6): 
507-513. 
 
Wimp, G. M., Young, W. P, Woolbright, S. A., Martinsen, G. D., Keim, P., and 
Whitman, T. G. (2004). Conserving plant genetic diversity for dependent animal 
communities. Ecology Letters 7: 776-780. 
 
Witkowski, E. T. F. and Lamont, B. B. (1991). Leaf specific mass confounds leaf 
density and thickness. Oecologia 88: 486-493. 
 
 148 
Witkowski, E. T. F. and Weiersbye, I. M. (1998). Establishment of plants on gold 
slimes dams: characterisation of the slimes and adjacent soils. Plant Ecology and 
Conservation Series No. 6. (Report to the Anglo-American Corporation). 
 
Wu, L. and Antonovics, J. (1975). Zinc and copper uptake by Agrostis stolonifera, 
tolerant to both zinc and copper. New Phytologist 75: 231-237. 
 
Xu-xiao, Z., Vaughan, D., Tomooka, N., Kaga, A., Xin-wang, W., Jian-ping, G., and 
Shu-min, W. (2003). Preliminary study on geographical distribution and 
evolutionary relationships between cultivated and wild adzuki bean (Vigna 
angularis var. angularis and var. nipponensis) by AFLP analysis. Agricultural 
Sciences in China 2(4): 368-376. 
 
Yee, E., Kidwell, K. K., Sills, G. R., and Lumpkin, T. A. (1999). Diversity among 
selected Vigna angularis (azuki) accessions on the basis of RAPD and AFLP 
markers. Crop Science 39: 268-275. 
 
Yeh, F. C., Yang, R. C., Boyle, T., Ye, Z. H., and Mao, J. X. (1997). POPGENE, the 
user friendly shareware for population genetic analysis.  Molecular biology and 
biotechnology centre, University of Alberta, Edmonton, Alberta, Canada. 
 
Zar, J. H. (1984). Biostatistical analysis (second edition).  Prentice-Hall, New 
Jersey, 718pp. 
 
Zeid, M., Sch?n, C., and Link, W. (2003). Genetic diversity in recent elite faba bean 
lines using AFLP markers. Theoretical Applied Genetics 107: 1304-1314. 
 
 
 
 
 
 
 
 149 
APPENDIX A - Raw morphological data
 1. Indigofera adenoides
 Ht CW CW90 LL1 LW1 LL2 LW2 LL3 LW3 LL4 LW4 LL5 LW5
 RW1 74 38 65 1.8 0.2 1.3 0.2 1.5 0.3 1.5 0.4 1.2 0.2
 RW2 53 32 27 1.2 0.4 1.5 0.4 1.8 0.4 1 0.5 1 0.5
 RW3 51 74 38 1.5 0.3 1.4 0.4 1 0.4 0.7 0.2 1.5 0.2
 RW4 48 86 83 0.8 0.2 1 0.4 0.8 0.4 0.7 0.3 1.2 0.5
 RW5 27 28 17 1.3 0.6 1.2 0.6 0.8 0.3 1 0.5 0.9 0.2
 RW6 30 21 28 1.9 0.8 1.3 0.5 1.4 0.7 1 0.5 0.8 0.2
 RW7 54 79 48 1.8 0.7 1.3 0.5 1.3 0.6 1.4 0.7 1.8 0.5
 RW8 41 69 38 1.8 0.8 3 1.5 2.6 1.1 1.9 0.6 2.1 0.6
 RW9 40 71 69 1.4 0.5 1.8 0.6 2.2 0.6 1.3 0.4 1.3 0.5
 TPW1 46 44 19 1.5 0.5 1.4 0.6 1.5 0.4 1.2 0.5 1.3 0.5
 TPW2 38 41 23 1.5 0.6 1.4 0.7 1.6 0.4 0.9 0.6 1.6 0.4
 TPW3 47 23 20 1.5 0.7 1.6 0.5 1 0.5 1.3 0.4 1.5 0.4
 TPW4 57 32 13 1.5 0.4 1.3 0.5 1.4 0.5 1 0.3 1.2 0.8
 TPW5 42 45 18 1.8 0.5 1.8 0.6 1.8 0.8 1.5 0.8 1 0.6
 TPW6 73 39 36 1.6 0.6 1.5 0.5 1 0.4 1 0.3 1 0.3
 TPW7 45 69 48 1.4 0.3 1.6 0.4 1.3 0.3 0.8 0.5 1.4 0.4
 TPW8 56 60 42 1.7 0.5 2.1 0.6 1.3 0.4 1.9 0.6 1.4 0.4
 TPW9 53 50 25 1.7 0.4 1.5 0.4 1.3 0.5 1.6 0.4 1.5 0.4
 TPW10 32 38 26 1.3 0.5 0.9 0.2 1.1 0.4 1.2 0.4 1.1 0.5
 TPW11 49 43 37 1.2 0.5 1.7 0.5 1.8 0.6 1.6 0.5 2.4 1
 TPW12 59 69 64 1.6 0.5 1.9 0.6 1.3 0.6 1.9 0.7 1.8 0.5
 TPW13 46 55 56 1.8 0.6 2.3 0.7 2.4 0.8 2.4 0.9 1.6 0.6
 TPW14 26 47 43 1.8 0.6 1.7 0.6 1.9 0.9 1.8 0.4 1.8 0.4
 TP1 44 61 45 2 0.6 1.8 0.6 1.8 0.6 1.5 0.5 1.8 0.6
 TP2 23 65 52 1.4 0.6 1.5 0.4 1.2 0.4 1.6 0.4 1 0.5
 TP3 66 53 48 1.5 0.5 1.5 0.4 1.4 0.4 1.3 0.3 1.3 0.6
 TP4 51 13 20 1.2 0.4 1.3 0.6 1.4 0.6 1.3 0.4 1.1 0.4
 TP5 34 20 15 1.1 0.6 1.4 0.4 1.5 0.5 1.2 0.5 1.4 0.5
 TP6 42 52 26 1.9 0.8 1.5 0.7 1.1 0.4 2 0.6 1.2 0.4
 TP7 27 34 23 0.9 0.3 1 0.4 1.1 0.3 0.9 0.4 1.2 0.5
 TP8 32 100 87 1.6 0.4 1.6 0.6 1.4 0.5 1.1 0.3 1.3 0.4
 TP9 57 66 49 1.8 0.5 1.4 0.4 1.1 0.5 1 0.6 1.3 0.5
 TP10 45 98 84 1.9 0.7 1.7 0.4 1.9 0.5 1.8 0.4 2.1 0.6
 TP11 34 84 66 1.3 0.5 1.4 0.5 1.4 0.5 1.6 0.5 1.9 0.6
 TP12 40 27 21 1.6 0.4 1.8 0.4 1.8 0.4 1.6 0.3 1.6 0.3
 TP13 48 86 66 1.3 0.4 1.5 0.5 1 0.5 1 0.3 1.1 0.3
 TP14 24 63 45 1.1 0.3 0.9 0.3 1 0.3 1.3 0.4 1.1 0.5
 TP15 31 88 51 1.2 0.5 1.4 0.5 1.8 0.6 1.4 0.6 1.1 0.5
 TP16 23 16 9 1.1 0.6 1.2 0.5 1.7 0.4 1.6 0.5 1.4 0.7
 TP17 38 42 29 1.8 0.9 2.1 0.7 1.8 0.6 1.8 0.5 1.8 0.5
 TP18 21 10 12 1 0.5 1.3 0.7 0.7 0.3 0.8 0.6 0.7 0.3
 TP19 28 52 26 0.7 0.3 1 0.4 1 0.3 0.9 0.4 1 0.3
  
 150 
TP20 24 12 8 1.3 0.5 1 0.6 1.1 0.6 0.9 0.3 0.9 0.6
 TP21 32 103 69 1 0.3 0.9 0.3 0.6 0.3 0.8 0.2 0.8 0.2
 TP22 21 10 8 0.8 0.4 0.7 0.4 0.8 0.4 0.7 0.4 0.6 0.3
 TP23 29 47 26 1 0.3 1 0.3 0.6 0.2 1 0.3 0.9 0.3
 TP24 26 42 29 0.9 0.3 1.2 0.4 1 0.4 1.1 0.4 1 0.3
 TP25 38 67 36 1.2 0.4 1.8 0.6 1.8 0.6 1.1 0.4 1 0.2
 TP26 23 47 35 0.7 0.4 0.7 0.5 1.2 0.5 1 0.3 0.7 0.3
 TP27 55 107 49 1.2 0.4 1.2 0.3 1.3 0.5 1.3 0.4 0.9 0.4
 TP28 31 55 45 1.5 0.6 2.1 0.8 1.9 0.5 1.2 0.4 1.3 0.5
 TP29 31 20 26 1.4 0.5 1.6 0.5 1 0.3 1.2 0.4 1.1 0.4
 TP30 36 32 15 1.8 0.5 1.6 0.5 1.1 0.3 1.8 0.3 1.8 0.3
 TP31 35 53 38 1.5 0.3 1.4 0.3 1.5 0.4 1.2 0.2 1.5 0.4
 TP32 43 50 26 1.2 1 2.1 0.7 2.4 0.6 2 0.5 1.6 0.5
 TP33 4 18 42 1.2 1 2.1 0.7 2.4 0.6 2 0.5 1.6 0.5
 TP34 33 32 27 1.2 0.3 1.4 0.4 1.2 0.4 1.4 0.4 1.3 0.4
 TP35 44 45 27 1.1 0.3 1.2 0.3 1.3 0.3 1.2 0.2 1 0.4
 TP36 31 30 19 1.1 0.2 1.5 0.4 1.5 0.4 1.5 0.4 1.5 0.4
 TP37 49 59 64 1.2 0.4 1.8 0.6 1.5 0.4 1.8 0.5 1.2 0.4
 TP38 32 37 38 1.1 0.4 1.1 0.5 1.2 0.4 1 0.6 1.1 0.4
 TP39 52 45 26 1.6 0.3 1.8 0.5 1.9 0.4 1.1 0.4 1.3 0.4
 TP40 32 20 28 1.1 0.3 0.9 0.3 0.9 0.4 0.8 0.3 0.7 0.3
 TP41 49 58 40 1.3 0.5 1.5 0.5 1.4 0.4 1.2 0.4 1.3 0.3
 TP42 24 55 47 1.1 0.5 1.8 1 2.5 1.2 1.6 0.9 1.5 0.6
 TP43 28 63 42 1.7 0.5 1.5 0.5 1.4 0.4 0.8 0.4 1 0.4
 TP44 24 62 41 1.8 0.6 1.4 0.5 1.4 0.5 1.8 0.5 1.5 0.4
 TP45 24 79 84 0.7 0.4 1 0.4 0.8 0.4 0.8 0.5 0.9 0.4
 TP46 18 73 45 0.9 0.2 1.3 0.4 1.1 0.2 1 0.4 1.1 0.2
 TP47 50 62 60 1 0.4 0.9 0.2 1 0.2 1 0.2 1.1 0.3
 TP48 28 23 10 1.1 0.6 0.9 0.5 1.1 0.4 0.9 0.4 1.5 0.5
 TP49 24 46 38 0.9 0.4 0.9 0.6 0.9 0.5 0.8 0.4 0.9 0.5
 TP50 35 61 58 1.4 0.4 1.3 0.5 1 0.5 1.1 0.4 1.3 0.4
 TP51 24 78 50 1.1 0.3 1.4 0.3 1.4 0.3 1.4 0.4 1.1 0.2
 TP52 25 47 18 1.4 0.3 1 0.2 1.1 0.2 0.9 0.2 1.3 0.2
 TP53 30 47 20 1.6 0.5 1.2 0.5 1.3 0.5 1.1 0.2 0.8 0.3
 TP54 30 46 29 1 0.4 0.8 0.3 1 0.4 0.9 0.4 0.8 0.2
 TP55 33 58 29 40 0.5 1.7 0.5 2 0.5 1.7 0.5 2.1 0.4
 OS1 27 25 18 2.1 0.5 2.1 0.5 1.6 0.7 1.6 0.6 1.8 0.6
 OS2 31 38 36 1.7 0.6 1.9 0.6 2 0.5 2 0.6 1.2 0.5
 OS3 30 17 16 0.8 0.3 1.1 0.3 1.1 0.3 0.9 0.4 1 0.4
 OS4 45 48 38 1.3 0.4 1.6 0.4 1.6 0.4 2.1 0.8 1.8 0.6
 OS5 26 39 20 1.4 0.4 1.4 0.7 1.4 0.6 1.2 0.5 1.2 0.4
 OS6 28 61 46 1.9 0.9 2.2 1 2.5 1.1 2.7 1.1 2.2 1
 OS7 38 37 30 1.7 0.5 1.5 0.5 1.3 0.3 0.6 0.3 1.1 0.3
 OS8 29 23 19 1.6 0.4 1.4 0.5 1.1 0.3 1.4 0.4 1.1 0.5
 OS9 28 42 28 1.3 0.5 1.7 0.4 1.4 0.4 1.1 0.4 1.3 0.5
 OS10 42 36 28 1.4 0.5 1.5 0.5 1.5 0.5 1.2 0.5 1 0.6
 OS11 36 107 83 1.2 1 3.5 0.6 2.7 0.9 2.1 0.8 2 0.6
 OS12 22 60 43 1.9 0.5 1.7 0.4 1.7 0.5 1.3 0.2 1.7 0.4
 OS13 35 47 34 1.3 0.2 1.8 0.3 1.1 0.4 0.8 0.2 1.6 0.3
  
 
 151 
OS14 63 15 48 2.4 0.5 2.4 0.5 2.4 0.4 2.4 0.4 2.1 0.3
 OS15 35 50 56 2.4 0.4 1.6 0.4 1.6 0.4 1.7 0.4 1.9 0.3
 OS16 55 55 49 1.8 0.5 2.4 0.3 1.8 0.3 1.4 0.4 2.2 0.5
 OS17 27 76 79 1.5 0.8 1.6 0.8 1.2 0.5 1.4 0.4 1.4 0.5
 OS18 35 52 21 1.8 0.6 2.3 1 1.9 0.6 1.9 0.6 1.9 0.7
 OS19 53 93 31 1.9 0.7 2.8 1.4 2.8 0.8 1.9 0.8 1.8 0.9
 OS20 45 37 36 2 1.3 1.9 0.9 2.2 0.8 2.2 0.9 1.5 1
 OS21 41 45 26 1.5 0.5 1.3 0.5 1.2 0.6 1.6 0.5 1.3 0.6
 OS22 45 55 59 2.1 0.9 2.2 0.8 2.2 0.8 1.9 0.6 1.9 0.6
 OS23 67 55 28 1.3 0.3 1.3 0.4 1.2 0.5 1.2 0.2 1 0.5
 OS24 60 43 47 1.4 0.3 1.4 0.3 1.3 0.3 1.3 0.3 1.2 0.2
 FOSA1 27 29 24 1 0.4 1.1 0.6 1.2 0.5 1.9 0.4 0.9 0.5
 FOSA2 36 42 32 1.8 0.6 1.6 0.6 1.6 0.8 1.1 0.6 1.9 0.6
 FOSA3 35 52 38 1.1 0.3 2.5 0.6 2.2 0.6 1.8 0.4 1.5 0.3
 FOSA4 31 28 23 1.8 0.6 1.6 0.5 1.7 0.6 1.6 0.3 1.9 0.4
 FOSA5 32 34 14 1.4 0.7 1.6 0.5 1.5 0.6 1.6 0.5 1.7 0.4
 FOSA6 33 40 34 1.4 0.4 1.5 0.3 1.4 0.4 0.9 0.3 1.3 0.4
 FOSA7 31 33 15 1.5 0.6 1.9 0.6 1.6 0.5 1.7 0.6 2 0.6
 FOSA8 36 34 31 1.2 0.5 2 0.9 2 0.8 1.8 0.5 1.9 0.6
 FOSA9 36 68 56 1.3 0.6 1.1 0.5 1.2 0.5 0.8 0.4 0.8 0.5
 FOSA10 35 34 26 2.8 0.7 2.4 1 2.6 0.6 2.9 0.8 2.6 0.7
 FOSA11 17 6 3 1.7 0.5 1.2 0.6 1.8 0.5 1.7 0.5 0.8 0.5
 FOSA12 34 73 48 1.4 0.6 1.7 0.7 1.7 0.9 1.6 0.6 1.1 0.7
 FOSA13 23 41 28 1 0.5 1.1 0.5 1.2 0.5 1.4 0.4 0.9 0.4
 FOSA14 21 32 15 1.5 0.6 1.4 0.7 1.4 0.8 1.3 0.7 1.2 0.5
 FOSA15 24 40 31 0.9 0.5 0.8 0.4 0.9 0.5 1.3 0.8 0.9 0.4
 FOSA16 24 41 36 1.9 0.8 1.4 0.5 1.7 0.4 1.9 0.9 2.1 0.8
 FOSA17 25 42 35 2.3 0.6 2.1 1.1 2.2 1 1.8 1 2.3 1
 FOSA18 31 33 30 0.9 0.4 1.1 0.5 1.2 0.6 1.3 0.4 1.2 0.4
 FOSA19 13 24 16 1.6 0.6 1.6 0.6 1.2 0.5 1.9 0.5 1.4 0.5
 FOSA20 21 34 19 1.8 0.6 2.6 0.9 2.9 0.9 3 1.1 1.2 0.4
 FOSA21 21 26 16 1.4 0.6 1 0.4 1.2 0.4 1.2 0.4 1.5 0.9
 FOSA22 29 28 26 1.1 0.5 0.9 0.3 1.1 0.5 0.9 0.5 1.2 0.4
 FOSA23 35 47 45 3.1 1 1.6 0.5 1.9 0.5 2.1 0.7 1.7 0.6
 FOSA24 36 66 45 2.2 0.8 2.5 0.8 2.2 0.5 2.2 0.6 2.1 0.6
 FOSB1 25 29 14 2 0.8 1.5 0.5 1.9 0.6 1.9 0.6 1.6 0.5
 FOSB2 16 17 13 1.6 0.7 1.8 0.5 1.6 0.5 2.1 0.6 1.7 0.4
 FOSB3 19 29 14 1.7 0.8 1.7 0.6 1.4 0.4 1.8 0.9 1.9 0.6
 FOSB4 45 34 17 1.8 1.3 1.5 1.2 1.2 1 1.3 1 1 0.6
 FOSB5 23 26 18 1.3 0.6 1.5 0.5 1.2 0.5 1.3 0.5 1 0.5
 FOSB6 28 39 25 2.2 1.1 1.8 0.6 1.4 0.3 1.6 0.4 1.2 0.4
 FOSB7 23 29 12 1.3 0.5 1.4 0.7 1.5 0.9 1.7 0.8 1.4 0.6
 FOSB8 47 35 28 1.8 0.8 1.3 0.6 1.5 0.6 1.3 0.8 1.5 0.6
 FOSB9 26 29 28 1.5 0.6 1 0.4 1.4 0.5 1.1 0.4 1 0.4
 FOSB10 24 20 19 1.8 0.8 1.5 0.6 1.3 0.5 1.4 0.3 1.6 0.6
 FOSB11 28 32 19 1.5 0.7 1.6 0.4 1.8 0.5 1.9 0.5 1.6 0.5
 FOSB12 30 33 22 1.7 0.5 1.8 0.6 1.5 0.4 1.4 0.5 1.3 0.4
 FOSB13 27 39 21 1.1 0.4 1.4 0.4 1.2 0.5 1.1 0.4 1.2 0.5
 FOSB14 25 24 16 1.4 0.6 1.3 0.4 1.9 0.5 1.4 0.3 1.4 0.4
 FOSB15 31 23 18 1.5 0.5 1.2 0.4 1.6 0.3 1.5 0.5 1.5 0.5
 FOSB16 24 30 23 1.5 0.4 1.6 0.6 1.6 0.5 1.5 0.4 1.8 0.4
 FOSB17 31 40 23 2.5 0.8 2 0.8 1.8 0.6 1.3 0.4 1.2 0.4
 FOSB18 18 26 23 1.5 0.6 1.6 0.8 1.4 0.6 1.2 0.6 1.5 0.6
 FOSB19 29 11 11 1 0.4 1.8 0.6 1.2 0.3 1.5 0.5 1.6 0.5
  
 
 
 152 
2. Indigofera zeyheri 
 
Ht CW CW90 LL1 LW1 LL2 LW2 LL3 LW3 LL4 LW4 LL5 LW5
 TP1 67 37 50 2.5 2.1 1.8 1.6 3.2 3 2.2 2.4 2 1.9
 TP2 79 55 32 2.8 2.1 2.7 2.1 2.5 3 2.2 2.7 2.3 2.7
 TP3 84 47 31 3.1 2.8 2.1 2 3.2 2.8 1.8 1.5 2.1 2
 TP4 89 55 24 3.2 2.8 2.4 2.1 2.8 2.5 2.5 2.5 2.8 2.7
 TP5 88 48 19 3.1 2 3.3 3 2.6 2 2.4 2.1 2.2 1.7
 TP6 57 25 15 2.1 2.8 2.7 2.7 2.4 2.5 1.9 1.9 2.3 2.5
 TP7 85 37 59 3.8 2.5 3.5 2.5 4 2.7 3.1 2.3 3.9 2.8
 TP8 36 8 6 3.1 3 3.3 3 2.3 2.3 3.1 3 2.9 2.7
 TP9 96 56 37 2.4 2.1 2.7 2 3.3 3 2.5 2.3 3.5 3.4
 TP10 100 59 45 3 3 2 2 3.7 3.6 3.2 3.2 3 3.1
 TP11 50 14 11 1.7 1.7 1.9 1.5 1.9 1.9 2 2 2.3 1.8
 TP12 75 16 12 3 1.9 2.9 2.5 3 2 2.8 2.2 2.8 2.3
 TP13 104 12 32 3.5 2.5 3.5 2.4 3.7 2.9 3.1 2.7 2.8 2.4
 TP14 64 18 25 2.1 2 2.4 2.3 2 1.5 2.7 2.7 2.5 2.5
 TP15 76 73 23 2.4 2.4 2.4 2.2 2.5 2.3 2.6 2.5 2.3 1.9
 TP16 128 102 102 2.2 2.2 2.2 2 1.9 2.1 2 1.8 1.1 1.3
 TP17 105 69 55 2.3 2.4 2.6 2.6 1.2 1.4 1.3 1.1 1.9 2
 TP18 111 67 59 2.9 3 2.8 2.8 2.5 2.1 2.3 2.4 2.6 2.5
 TP19 104 65 28 1.8 1.8 1.7 1.7 1.7 1.8 1.8 1.7 1.7 2
 TP20 79 41 25 2.5 3 2.4 2.3 2.5 2.6 2.6 2.6 2 2.1
 TP21 118 36 18 3.9 3 3.4 2.6 2.7 2.7 3.8 3.6 3.2 2.9
 TP22 34 15 12 1.8 2.2 1.4 1.2 1.5 1.8 1.1 1.3 1 1.2
 TP23 56 11 13 3.1 2.9 3.3 3.1 2.6 2.5 2.6 2.8 3 3.2
 TP24 110 32 17 3.5 1.9 1 1.8 4.5 2.9 4.5 2.7 4 2.6
 TP25 89 49 45 3.1 2.8 3.3 3.2 2.6 2.7 2.6 2.6 3 3
 TP26 48 2 4 2 1.9 2.4 1.9 2.4 1.6 2.4 1.8 1.9 1.7
 TP27 106 29 20 1.9 2.2 3.1 2.5 2.5 3.3 2.9 3.4 2.6 3.3
 TP28 112 24 15 2.9 2.5 3.1 2.4 2.8 2.4 2.8 2.1 3 2.8
 TP29 59 10 9 3.1 3.1 2.8 2.6 3.1 3 2.9 3.2 3.2 2.9
 TP30 110 18 12 3.5 2.9 3.5 2.5 3.3 2.3 2.8 2.3 2.7 1.7
 TP31 87 58 33 2.9 2.9 2.8 2.6 1.7 1.8 2.9 2.8 2.3 2.4
 TP32 110 18 17 3.4 3.2 4.1 3.2 3.6 3 3.9 2.8 4.4 2.8
 TP33 103 48 20 3.1 3.2 4.1 3 3.8 3.6 3.9 2.5 3.9 4
 TP34 94 31 17 3.6 2.4 3.7 2.9 3.5 3.1 3.5 2.9 3.7 2.6
 TP35 123 41 14 4 3.5 4.2 3.6 3.6 2.7 3.8 2 3.5 2.6
 TP36 120 49 29 3.6 2.2 3.9 2.3 4.2 3 2.8 2.5 4 2.4
 TP37 107 38 20 3.5 2 3.3 2.6 3.7 2.4 3.8 2.8 3.5 3
 TP38 92 41 37 2.5 1.9 2.1 2 2.4 2.1 2.5 2.3 2.3 2.1
 TP39 112 82 55 1.9 2.2 2.2 2.1 2.5 2.3 1.9 1.8 1.2 1
 TP40 132 66 63 2.3 1.9 2.7 2 2 1.8 1.5 1.4 3.2 2.6
 TP41 62 8 5 3.1 2.8 3.1 2.8 3.2 3 3.1 3.1 3.2 3.2
 TP42 132 102 99 3 2.5 2.1 1.9 1.8 1.6 1.3 1.2 1.8 1.9
 TP43 105 35 29 3.6 3.5 3.5 3.2 3.7 3.8 1.7 1.8 1.5 1.8
 TP44 93 42 26 3.2 3 2.1 1.7 2 2 1.8 1.7 2 1.9
 TP45 62 8 5 1.5 1.7 2.3 2 2.8 2.4 3 2.8 1.7 2.1
 TP46 86 79 49 2.5 2 2.5 2.1 1.8 1.8 3.1 2.3 3.5 2.8
 TP47 96 58 54 3.5 3.3 2 2 1.7 2 1.7 2 1.9 2.2
 TP48 71 40 35 2.6 2 2.6 2.2 1.8 1.4 3.4 2.6 3.3 3
 TP49 108 68 59 3.7 4.1 3.4 3.3 2.5 3.2 3.5 3.8 2.8 2.7
 TP50 72 43 55 3.2 2.9 4 3.5 2.6 2.6 3.8 3.1 3.5 3.1
 TP51 131 67 72 2.2 2.4 2 1.6 2 2.3 2.5 2.8 3.2 3.2
  
 153 
TP52 53 30 23 2.8 2.2 2.6 2.4 2.2 2.6 2.9 2.6 2.4 2.2
 TP53 94 43 20 1.8 2 2.4 2.3 2.8 3.1 2.6 2.9 3.2 3.5
 TP54 55 41 20 2.8 2.5 3.2 3 2.9 3.4 2.5 2.7 3.8 3.4
 TP55 91 33 16 3 2.4 2.8 2.2 2.6 2.6 3.2 2.8 2.9 2.9
 TP56 59 39 28 3.5 2.8 3.7 3 2.6 2.3 3.7 3.3 3.6 3.4
 TP57 86 44 36 1.4 1.6 3 2.9 1.5 1.4 2.6 2.5 3.2 3.6
 TP58 115 119 90 3.6 2.4 2.5 2.1 2 2 2 2 3.6 3.2
 TP59 117 54 62 2.1 2.2 2.1 2.1 2.2 2.2 2.8 2.4 2.4 2.8
 TP60 179 21 18 3.3 3.5 3.5 3 3.2 3.4 3.2 3 1.5 1.4
 TP61 33 17 17 2.2 2.5 2.8 2.7 3 3.3 3.4 3.1 2.6 3.1
 TP62 97 53 49 3.3 2.7 2.7 2.6 3 2.7 3.2 2.8 3.4 2.8
 TP63 60 30 24 3 1.8 3.5 3.4 3.5 2.8 1.4 2.3 3 2.5
 OS1 76.5 37 19.5 3.1 3.5 2.3 2.9 3.2 3.3 2.7 2.7 2.3 2.4
 OS2 109 54 30 2.6 2.2 2.1 2 3.5 3.6 2.1 2.4 2.7 2.6
 OS3 125 110 100 1.9 1.8 1.8 1.3 1.8 1.5 1.9 1.4 1.5 1.4
 OS4 72 46 20 2.8 2.8 3.2 3.1 1.8 2.2 3.4 3.1 3.1 3
 OS5 81 49 29 3.7 3.2 2.5 2.3 3.6 4.2 2.4 2.4 3.3 3.7
 OS6 86 84 45 3.6 3.3 2.1 2.4 3.1 3.2 3.2 2.7 1.7 1.9
 OS7 124 37 20 2.1 2.2 2.1 2.5 2.1 1.7 3.6 3 3.8 3.3
 OS8 64 16 10 3.2 3.1 3.2 2.9 3.2 2.8 2.7 3 2.3 2.4
 OS9 98 46 55 3 2.8 2 1.8 1.8 1.5 1.7 1.5 1.7 1.7
 OS10 84 75 51 2.9 3.4 3.5 2.9 3.4 3.6 2.7 2.6 1.7 2
 OS11 93 60 28 2.7 3.4 3.1 3.3 2.9 3.7 1.7 2.1 1.6 2.1
 OS12 90 160 120 3.1 3.5 3.6 3.5 3.1 3.6 3.4 3.8 3.3 3.4
 OS13 58 68 32 2 2.9 2.1 2.8 2.4 3.2 2.6 3.1 2.3 2.8
 OS14 135 110 45 2.6 2.8 2.3 2.6 2.3 2.8 2.2 2.6 1.8 2
 OS15 122 110 103 2.3 2.4 2.7 2.8 2.3 2.4 1.8 1.8 2.9 2.6
 OS16 50 56 48 2.9 3.2 2.8 3.1 3 3.1 3.3 3.8 3.6 3.8
 OS17 85 65 63 2.2 2.3 2.2 2.3 2.2 2.4 3.1 3.4 3.3 3.6
 OS18 91 65 38 2.8 3.2 2.5 2 3 3.1 2.9 3.2 2.7 2.9
 OS19 117 64 43 2.8 2.3 3.3 2.3 3.8 3.6 2.6 2.6 3.9 3.2
 OS20 85 71 55 2.1 2 1.9 1.8 2 1.8 2.5 2.2 1.8 1.8
 OS21 83 62 68 2 1.8 1.9 2.3 2.4 2.7 2.3 2.3 1.8 2.2
 OS22 85 54 52 3.5 2.4 1.8 1.8 2 1.6 1.8 1.6 2 1.8
 OS23 84 78 56 1.5 1.6 2.8 3.2 2.6 2.7 2.5 2.3 3 2.9
 OS24 92 73 60 2.1 1.9 2.6 2.1 2.2 1.8 2.4 1.7 2 2.1
 OS25 82 38 33 3.7 2.5 3.1 2.5 4 2.9 1.6 1.9 1.8 2
 OS26 92 30 46 2.9 2.5 3.3 2.6 3.2 3.4 3.4 2.8 3.1 3.5
 OS27 114 113 107 1.6 1.8 1.3 1.9 1.5 1.4 1.5 2 2.6 2.5
 OS28 90 28 26 1.6 1.4 1.4 1.4 2.6 2.1 1.5 1.6 2.4 2.3
 OS29 127 84 110 2.1 2.7 1.3 1.5 2.9 2.8 2 2.6 2.4 2.3
 OS30 122 93 105 4.1 3.2 4.1 2.3 2.7 2.3 3.1 2.3 2.8 2.3
 OS31 123 77 36 2.6 2.5 2.7 2.8 2.3 2.5 2.7 2.4 2.6 2.6
 OS32 91 57 60 2.6 2.2 3.2 2.7 3.1 2.6 2.1 2.2 3 2.2
 OS33 123 76 45 1.6 1.5 1.3 1.2 2.1 1.8 2.6 2.6 2.3 2
 OS34 111 52 49 3.2 2.4 2.7 2.5 3.4 2.6 3 2.4 2.8 2.4
 OS35 118 67 42 2.9 2.8 3.7 3.2 3.6 3 2.5 2.6 2.3 2
  
 154 
APPENDIX B - Raw genetic data
 1.  Indigofera adenoides
 RW1 0 0 1 0 0 1 1 1 0 1 1 1 1 1 1 1 0 1 1 1 0 0 1 0 0 1 1 1 1 1 0 1 1 0 1 0 1 1 0 1 1 0 1 1
 RW2 0 0 1 0 0 0 1 1 0 0 0 1 1 1 1 1 0 1 1 0 1 0 0 0 1 0 0 1 1 1 0 1 1 0 1 0 1 1 0 1 1 0 1 1
 RW3 1 0 1 0 0 0 1 1 0 1 1 1 1 1 1 1 0 1 1 0 0 0 0 0 1 1 0 1 1 1 1 1 1 0 1 0 1 1 0 1 1 0 1 1
 RW4 0 0 1 0 0 0 1 1 0 1 1 1 1 1 1 1 0 1 1 1 1 1 0 1 0 1 1 1 1 0 1 1 1 0 1 0 1 1 0 1 1 0 1 1
 RW5 0 0 1 0 0 1 1 1 0 1 1 1 1 1 1 1 1 1 1 0 1 0 1 0 0 1 0 1 1 0 0 1 1 0 1 0 1 1 0 1 1 0 1 1
 RW6 0 0 1 1 0 1 1 1 0 0 0 1 1 1 1 1 1 1 1 0 1 0 0 1 0 0 0 1 1 0 1 1 1 0 1 0 1 1 0 1 1 0 1 1
 RW7 0 0 1 0 0 0 1 1 0 1 1 1 1 1 1 1 1 1 1 0 1 0 1 1 0 0 0 1 1 0 1 1 1 0 1 0 1 1 0 1 1 0 1 1
 RW8 1 0 1 0 0 0 1 1 0 0 0 1 1 1 1 0 1 1 1 1 1 0 1 0 0 1 1 1 1 1 1 1 1 0 0 0 1 1 0 1 1 0 1 1
 RW9 1 0 1 1 0 0 1 0 0 0 1 1 1 1 1 1 1 1 1 1 1 0 1 1 1 1 0 1 1 0 1 0 1 0 1 0 1 1 1 1 1 0 1 1
 TPW1 1 0 1 0 1 0 1 1 0 1 1 1 1 1 1 1 1 1 1 0 1 0 1 0 0 0 0 1 1 1 1 1 1 0 1 0 1 1 1 1 1 0 1 1
 TPW2 1 0 1 1 1 0 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 0 0 0 0 1 1 1 1 1 0 1 1 0 1 0 1 1 0 1 1 0 1 1
 TPW3 1 0 1 1 1 0 1 1 0 0 1 1 1 0 1 1 1 1 1 0 1 0 0 1 0 1 1 1 1 1 1 1 1 0 1 0 1 1 0 1 1 0 1 1
 TPW4 1 0 1 0 1 0 1 0 1 1 1 1 1 1 1 1 1 1 1 1 1 0 0 1 1 1 1 1 1 1 1 1 1 1 1 0 1 1 0 1 1 0 1 1
 TPW5 0 0 1 1 1 0 1 1 0 0 1 1 1 1 1 1 1 1 1 1 1 0 0 1 1 1 1 1 1 1 1 1 0 1 1 0 1 1 0 1 1 0 1 1
 TPW6 0 0 1 0 0 0 1 0 1 1 0 1 1 1 1 0 1 1 1 0 1 0 0 0 1 1 0 1 1 1 1 1 1 0 1 0 1 0 0 1 1 0 1 1
 TPW7 0 0 1 0 1 0 1 1 0 1 1 1 1 1 1 0 1 1 1 0 0 1 0 0 1 0 0 1 1 1 1 1 1 0 1 0 1 1 0 1 1 0 1 1
 TPW8 1 0 1 1 1 0 1 0 0 0 1 1 1 1 1 0 1 1 1 0 0 0 1 1 0 1 1 1 1 0 1 1 1 0 1 0 1 1 0 1 1 0 1 1
 TPW9 1 0 1 1 1 0 1 0 0 0 1 1 1 1 1 0 1 1 1 0 1 0 0 1 1 1 1 1 1 1 0 1 1 0 1 0 1 1 1 1 1 0 1 1
 TPW101 0 1 1 0 0 1 0 1 0 1 1 1 1 1 0 1 1 1 1 1 0 0 1 1 0 1 1 1 1 1 1 1 1 1 0 1 1 1 1 1 0 1 1
 TPW111 0 1 0 1 0 1 0 0 1 1 1 1 1 1 1 0 1 1 1 1 0 0 1 1 1 0 1 1 0 0 1 1 0 1 0 1 1 1 1 1 0 1 1
 TPW120 0 1 0 1 0 1 0 0 1 0 1 1 1 1 1 1 1 1 1 1 0 1 0 1 1 1 1 1 0 0 1 1 0 1 0 1 1 1 1 1 0 1 1
 TPW130 0 1 0 1 0 1 0 0 1 0 1 1 1 1 1 1 1 1 0 1 0 0 0 1 1 1 1 1 1 0 1 1 0 1 0 1 1 0 1 1 0 1 1
 TPW141 0 1 1 1 0 1 1 0 0 1 1 1 1 1 1 1 1 1 1 1 0 0 0 1 1 1 1 1 0 0 1 1 0 1 0 1 1 0 1 1 0 1 1
 TPW150 0 1 0 1 0 1 0 0 1 0 1 1 1 1 0 1 1 1 1 1 0 0 1 1 1 1 1 1 1 1 1 1 1 1 0 1 1 0 1 1 0 1 1
 TP1 0 0 1 0 0 0 1 1 0 0 0 1 1 1 1 1 0 1 1 1 0 0 1 1 0 1 0 1 1 1 0 1 1 0 1 0 1 1 0 1 1 0 1 1
 TP2 0 0 1 0 0 0 1 1 0 1 0 1 1 0 1 1 0 1 1 0 1 0 0 0 0 0 0 1 1 1 1 1 1 0 1 0 1 1 0 1 1 0 1 1
 TP3 0 0 1 1 0 0 1 1 0 1 0 1 1 1 1 1 0 1 1 0 1 1 0 0 1 1 0 1 1 1 0 1 1 0 1 0 1 1 0 1 1 0 1 1
 TP4 0 0 1 0 0 0 1 1 0 0 0 1 1 1 1 1 1 1 1 1 1 0 1 1 0 1 0 1 1 1 0 1 1 0 1 0 1 1 0 1 1 0 1 1
 TP5 1 0 1 1 0 0 1 0 0 0 0 1 1 1 1 1 0 0 1 0 1 0 0 0 1 1 0 1 1 1 1 1 1 0 1 0 1 1 0 1 1 0 1 1
 TP6 0 0 1 1 0 0 1 1 0 0 0 1 1 1 1 1 0 1 1 0 0 0 0 0 0 0 1 1 1 1 0 1 1 0 1 0 1 1 0 1 1 0 1 1
 TP7 0 0 1 0 0 0 1 1 0 0 0 1 1 1 1 1 0 1 1 0 1 0 0 0 0 0 1 1 1 0 0 1 1 0 1 0 1 1 0 1 1 0 1 1
 TP8 1 0 1 1 0 1 1 1 0 0 0 1 1 1 1 1 0 1 1 0 0 0 1 1 0 0 1 1 1 0 1 1 1 0 1 0 1 1 0 1 1 0 1 1
 TP9 0 0 1 1 0 1 1 1 0 1 1 1 0 0 1 1 0 1 1 1 0 0 0 0 0 1 1 1 1 0 1 1 1 0 1 0 1 1 1 1 1 0 1 1
 TP10 1 0 1 1 1 1 1 1 0 1 1 1 1 1 1 1 0 1 1 1 0 1 0 0 0 1 0 1 1 0 1 1 1 0 1 0 1 1 0 1 1 0 1 1
 TP11 0 0 1 1 0 1 1 1 0 1 0 1 1 1 1 1 1 1 1 1 0 0 0 1 0 0 1 1 1 0 1 1 1 0 1 0 1 1 0 1 1 0 1 1
 TP12 1 0 1 1 0 0 1 1 1 0 1 1 1 1 1 1 1 1 1 0 0 0 1 1 0 1 1 1 1 1 1 1 0 1 1 0 1 1 0 1 1 0 1 1
 TP13 1 0 1 1 0 0 1 1 1 0 1 1 1 1 1 1 1 1 1 0 0 0 1 1 0 1 1 1 1 1 1 1 0 1 1 0 1 1 0 1 1 0 1 1
 TP14 0 0 1 1 1 0 1 1 1 0 1 1 1 1 1 1 1 1 1 0 0 0 1 1 0 1 1 1 1 1 1 1 0 1 1 0 1 1 0 1 1 0 1 1
 TP15 1 0 1 1 1 0 1 1 1 0 1 1 1 1 1 1 1 1 1 0 0 0 1 1 0 1 1 1 1 1 1 1 0 1 1 0 1 1 0 1 1 0 1 1
 TP16 1 0 1 1 0 0 1 1 0 1 1 1 1 1 1 1 1 1 1 1 1 0 1 1 1 1 1 1 1 1 1 1 1 0 1 0 1 1 1 1 1 0 1 1
 TP17 1 0 1 1 1 0 1 1 0 0 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 0 1 1 1 1 1 0 1 1
 TP18 1 0 1 1 1 1 1 1 1 1 0 1 1 1 1 1 1 1 1 0 1 0 1 1 0 1 0 1 1 1 0 1 1 1 1 0 1 1 0 1 1 0 1 1
 TP19 0 0 1 0 1 0 1 1 1 0 0 1 1 1 1 1 1 1 1 1 1 0 0 1 1 1 1 1 1 1 0 0 1 0 1 0 1 1 0 1 1 0 1 1
 TP20 1 0 1 1 1 0 1 1 1 1 1 1 0 1 1 1 1 0 1 1 1 0 1 0 1 1 1 1 1 1 1 1 1 0 1 0 1 1 0 1 1 0 1 1
 TP21 0 0 1 1 1 0 1 0 0 1 1 1 0 1 1 0 1 1 1 0 1 0 0 1 0 1 0 1 1 0 1 1 1 0 1 0 1 1 0 1 1 0 1 1
 TP22 1 0 1 1 1 0 1 1 0 1 1 1 1 1 1 1 1 1 1 1 1 0 0 0 1 1 1 1 1 1 0 1 1 0 1 0 1 1 0 1 1 0 1 1
 TP23 1 0 1 0 1 0 1 1 0 1 0 1 1 1 1 0 1 0 1 0 0 0 1 0 0 1 0 1 1 0 1 0 1 0 1 0 1 1 0 1 1 0 1 1
 TP24 0 0 1 0 1 0 1 0 0 0 1 1 1 1 1 1 1 0 1 1 0 0 0 0 1 1 1 1 1 1 1 1 1 0 1 0 1 1 0 1 1 0 1 1
 TP25 0 0 1 1 0 0 1 1 0 1 1 1 1 1 1 1 1 0 1 0 1 0 0 0 1 0 0 1 1 1 0 1 1 0 1 0 1 1 1 1 1 0 1 1
  
 155 
TP26 1 0 1 1 1 0 1 0 1 1 0 1 1 1 1 1 1 1 1 0 1 0 1 1 0 1 0 1 1 1 1 0 1 0 1 0 1 1 0 1 1 0 1 1
 TP27 1 0 1 0 1 0 1 1 0 1 1 1 1 1 1 1 1 0 1 1 1 0 0 1 0 1 1 1 1 1 0 1 0 1 1 0 1 1 0 1 1 0 1 1
 TP28 0 0 1 1 1 0 1 1 0 0 1 1 1 1 1 1 1 1 1 0 1 0 1 1 1 1 1 1 1 1 0 1 1 1 1 0 1 1 0 1 1 0 1 1
 TP29 0 0 1 0 1 0 1 0 0 0 1 1 1 1 1 0 1 1 1 0 0 0 0 0 1 1 1 1 1 1 1 0 1 0 0 0 1 1 0 1 1 0 1 1
 TP30 1 0 1 0 1 0 1 1 0 0 0 1 1 1 1 1 1 1 1 0 1 0 0 0 1 1 0 1 1 1 0 1 1 0 1 0 1 1 0 1 1 0 1 1
 TP31 0 0 1 0 0 0 1 1 0 0 1 1 1 1 1 1 1 0 1 1 0 0 0 0 1 1 1 1 1 1 0 1 1 1 1 0 1 1 1 1 1 0 1 1
 TP32 0 0 1 0 1 0 1 1 0 0 1 1 1 1 1 1 1 1 1 0 1 0 0 1 1 1 0 1 1 1 0 1 1 1 0 0 1 1 0 1 1 0 1 1
 TP33 0 0 1 0 1 0 1 1 0 0 0 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 0 1 1 1 1 0 1 1 0 1 1 0 1 1
 TP34 0 0 1 1 1 0 1 1 0 0 0 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 0 0 1 0 1 0 1 1 0 1 1 0 1 1
 TP35 0 0 1 1 1 0 1 1 0 0 1 1 1 1 1 1 1 1 1 0 1 0 1 1 1 1 1 1 1 1 0 1 1 1 1 0 1 1 1 1 1 0 1 1
 TP36 0 0 1 1 0 0 1 0 0 0 1 1 1 1 1 1 1 1 1 0 1 0 1 1 1 1 1 1 1 1 1 1 1 0 1 0 1 1 0 1 1 0 1 1
 TP37 1 0 1 1 1 0 1 1 1 0 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 0 1 1 0 1 1 0 1 1
 TP38 0 0 1 0 1 0 1 1 1 0 0 1 1 1 1 0 1 0 1 1 1 0 1 1 1 1 1 1 1 1 0 1 1 0 1 0 1 1 1 1 1 0 1 1
 TP39 1 0 1 0 0 1 1 1 0 0 1 1 1 1 1 1 1 1 1 0 1 1 1 1 0 1 1 1 1 1 1 1 1 0 1 0 1 1 0 1 1 0 1 1
 TP40 1 0 1 0 1 0 1 1 1 0 0 1 1 1 1 0 1 1 1 1 1 0 0 1 1 1 0 1 1 1 0 1 1 0 1 0 1 1 0 1 1 0 1 1
 TP41 0 0 1 0 1 0 1 0 0 0 1 1 1 1 1 0 1 1 1 0 1 1 1 0 1 1 1 1 1 1 0 1 1 0 1 0 1 1 1 1 1 0 1 1
 TP42 1 0 1 0 1 0 1 1 0 0 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 0 1 1 0 1 0 1 1 0 1 1 0 1 1
 TP43 0 0 1 0 1 0 1 1 0 0 1 1 1 1 1 1 1 1 1 0 1 0 0 0 0 1 1 1 1 1 0 1 1 0 1 0 1 1 0 1 1 0 1 1
 TP44 1 0 1 1 1 0 1 1 1 1 0 1 1 1 1 1 1 1 1 1 1 1 0 1 1 0 1 1 1 0 1 1 1 0 1 0 1 1 0 1 1 0 1 1
 TP45 0 0 1 1 0 0 1 1 0 0 0 1 1 1 1 1 1 1 1 1 1 0 0 0 1 1 0 1 1 1 1 1 1 0 1 0 1 0 1 1 1 0 1 1
 TP46 1 0 1 1 1 0 1 1 1 1 1 1 1 1 1 0 1 1 1 1 1 1 0 0 1 1 1 1 1 1 0 1 1 0 0 0 1 1 1 1 1 0 1 1
 TP47 1 0 1 0 0 0 1 1 1 1 1 1 1 1 1 0 0 1 1 1 1 1 0 0 1 1 1 1 1 1 0 1 1 0 1 0 1 1 1 1 1 0 1 1
 TP48 1 0 1 1 1 0 1 1 1 1 0 1 1 1 1 1 1 1 1 1 1 0 0 1 1 1 1 1 1 1 0 1 1 1 1 0 1 1 1 1 1 0 1 1
 TP49 1 0 1 0 1 0 1 1 1 1 0 1 1 1 1 0 1 1 1 1 1 1 0 1 1 1 1 1 1 1 0 1 1 1 1 0 1 1 0 1 1 0 1 1
 TP50 1 0 1 1 1 0 1 1 1 1 0 1 1 1 1 1 1 0 1 1 1 1 0 0 1 1 0 1 1 1 1 1 1 1 1 0 1 1 1 1 1 0 1 1
 TP51 1 0 1 0 1 0 1 1 1 0 1 1 1 1 1 0 1 1 1 1 1 0 1 1 1 1 0 1 1 1 1 1 1 0 1 0 1 1 0 1 1 0 1 1
 TP52 0 0 1 0 1 0 1 1 0 1 1 1 1 1 1 1 1 1 1 1 1 0 0 1 1 1 0 1 1 1 1 1 1 1 1 0 1 1 0 1 1 0 1 1
 TP53 1 0 1 1 1 0 1 1 0 0 1 1 1 1 1 1 1 1 1 1 1 0 0 1 1 1 1 1 1 1 1 1 1 0 1 0 1 1 0 1 1 0 1 1
 TP54 0 0 1 0 1 0 1 0 0 0 0 1 1 1 1 0 0 1 1 0 0 0 1 1 1 0 0 1 1 1 1 1 1 1 1 0 1 1 1 1 1 0 1 1
 OS1 0 0 1 1 1 1 1 0 1 0 0 1 1 1 1 1 1 1 1 1 1 0 0 1 1 1 1 1 1 0 1 0 1 0 0 0 1 1 0 1 1 0 1 1
 OS2 0 0 1 1 0 0 1 1 1 1 0 1 1 1 1 0 1 1 1 0 1 0 0 1 1 1 0 1 1 0 0 1 1 0 0 0 1 1 0 1 1 0 1 1
 OS3 0 0 1 0 1 1 1 0 0 1 1 1 1 1 1 0 1 0 1 0 1 0 0 0 0 1 1 1 1 1 0 1 1 0 1 0 1 1 1 1 1 0 1 1
 OS4 1 0 1 0 0 1 1 1 0 1 0 1 1 1 1 1 1 0 1 1 1 0 0 1 0 1 1 1 1 1 1 1 1 0 0 0 1 0 1 1 1 0 1 1
 OS5 0 0 1 1 1 0 1 1 0 0 1 1 1 1 1 0 0 0 0 1 1 0 1 1 1 1 1 1 1 1 1 1 1 0 0 0 1 1 1 1 1 0 1 1
 OS6 0 0 1 0 0 1 1 0 0 0 0 1 1 1 1 0 0 0 1 1 1 0 0 0 0 1 1 1 0 0 0 1 1 0 1 0 1 1 0 1 1 0 1 1
 OS7 0 0 1 0 0 1 1 1 0 1 1 1 1 1 1 1 0 1 1 1 0 0 1 1 0 1 0 1 1 0 1 1 1 0 1 0 1 1 0 1 1 0 1 1
 OS8 0 0 1 0 0 1 1 1 0 1 1 1 1 0 1 1 1 1 1 1 0 0 0 0 0 1 1 1 1 0 1 1 1 0 1 0 1 1 0 1 1 0 1 1
 OS9 0 0 1 0 0 0 1 1 0 0 0 1 1 1 0 1 0 1 0 1 0 0 0 1 0 0 0 1 1 0 1 1 1 0 1 0 1 1 0 1 1 0 1 1
 OS10 1 0 1 0 1 0 1 0 0 0 1 1 1 1 1 1 1 1 1 0 0 0 0 0 0 1 1 0 1 1 1 1 1 0 1 0 0 1 0 1 1 0 1 1
 OS11 0 0 1 1 0 0 1 1 1 0 1 1 1 1 1 0 1 1 1 1 1 0 1 1 0 1 0 1 1 1 1 0 1 0 1 0 1 1 1 1 1 0 1 1
 OS12 0 0 1 0 1 0 1 1 1 1 1 1 1 1 1 1 1 1 1 0 1 1 0 0 0 1 1 1 1 1 1 1 1 0 1 0 1 1 0 1 1 0 1 1
 OS13 0 0 1 1 1 0 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 0 1 0 1 1 1 1 1 1 0 1 0 1 0 1 1 0 1 1 0 1 1
 OS14 1 0 1 1 1 0 1 1 0 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 0 1 1 1 1 1 1 1 1 0 1 0 0 1 0 1 1 0 1 1
 OS15 1 0 1 1 0 0 1 1 1 0 1 1 1 1 1 1 1 1 1 1 1 0 0 0 1 1 0 1 1 1 0 1 1 0 0 0 1 0 1 1 1 0 1 1
 OS16 1 0 1 1 0 0 1 1 1 1 0 1 1 1 1 1 1 0 1 1 1 0 1 1 0 1 0 1 1 1 1 1 1 0 1 0 1 1 1 1 1 0 1 1
 OS17 1 0 1 1 1 0 1 1 1 0 0 1 1 1 1 0 1 0 1 0 1 0 0 0 1 1 1 1 1 1 1 1 1 0 1 0 1 1 0 1 1 0 1 1
 OS18 1 0 1 1 1 1 1 1 1 0 1 1 1 1 1 1 1 1 1 1 1 0 1 1 1 1 1 1 1 1 1 1 1 0 1 0 1 1 0 1 1 0 1 1
 OS19 1 0 1 0 1 0 1 0 1 0 0 1 1 1 1 0 1 0 1 0 1 1 1 1 0 0 0 1 1 0 1 1 1 0 1 0 1 0 1 1 1 0 1 1
 OS20 0 0 1 1 1 0 1 0 1 0 1 1 1 1 1 1 1 1 1 1 1 0 1 0 1 1 1 1 1 1 1 1 1 0 1 0 1 1 0 1 1 0 1 1
 OS21 0 0 1 1 0 0 1 1 0 1 1 1 1 1 1 1 1 1 1 0 1 0 0 1 1 1 1 1 1 1 0 1 0 1 1 0 1 1 0 1 1 0 1 1
 OS22 0 0 1 0 1 0 1 1 0 1 1 1 1 1 1 1 1 0 1 1 1 0 0 1 0 1 1 1 1 1 0 1 0 1 1 0 1 1 0 1 1 0 1 1
 OS23 1 0 1 0 1 0 1 1 0 0 1 1 1 1 1 1 1 1 1 1 1 0 1 1 0 0 0 1 1 1 1 1 1 0 1 0 1 1 0 1 1 0 1 1
 OS24 1 0 1 1 1 0 1 1 1 0 1 1 1 1 1 1 1 1 1 1 1 1 0 0 1 1 1 1 1 1 1 1 1 1 1 0 1 0 0 1 1 0 1 1
  
 
 156 
FOSA1 1 0 1 0 1 0 1 0 0 0 0 1 1 1 1 0 1 1 1 0 1 0 0 1 1 1 0 0 1 1 1 1 0 1 1 0 1 1 0 1 1 0 1 1
 FOSA2 0 0 1 0 1 1 1 0 1 1 0 1 1 0 1 1 1 1 1 0 1 0 0 0 1 1 0 1 1 1 1 1 1 1 1 0 1 1 0 1 1 0 1 1
 FOSA3 0 0 0 1 1 0 1 0 1 0 0 1 1 1 1 0 1 1 1 0 1 0 0 1 1 1 1 1 1 1 0 1 0 1 1 0 1 1 0 1 1 0 1 1
 FOSA4 0 0 1 0 1 0 1 0 1 0 1 1 1 0 1 1 1 1 1 0 1 0 0 0 1 1 1 1 1 1 0 1 1 1 1 0 1 1 1 1 1 1 1 1
 FOSA5 1 0 1 0 1 0 1 0 1 0 1 1 1 1 1 1 1 0 1 0 1 0 0 1 1 1 1 1 1 1 1 1 0 1 0 0 1 1 0 1 1 0 1 1
 FOSA6 1 1 1 0 0 0 1 1 0 0 1 1 1 1 1 1 1 1 1 0 1 0 0 1 1 1 1 1 1 1 0 1 0 1 1 0 1 1 1 1 1 0 1 1
 FOSA7 0 0 1 0 1 0 1 1 0 1 1 1 1 1 1 1 1 1 1 1 1 0 1 1 0 1 1 1 1 1 0 1 0 1 1 0 1 1 1 1 1 1 1 1
 FOSA8 0 0 1 0 1 0 1 1 0 1 1 1 1 1 1 1 1 1 1 0 1 0 0 1 1 1 1 1 1 1 0 0 0 1 1 0 1 1 0 1 1 0 1 1
 FOSA9 1 0 1 1 1 0 1 0 0 0 1 1 1 1 1 1 1 1 1 0 0 0 1 1 0 1 1 1 1 1 0 1 1 1 1 0 1 1 1 1 1 0 1 1
 FOSA10 0 0 1 0 1 0 1 1 0 1 0 1 1 1 1 1 0 1 1 0 0 0 1 1 1 1 1 1 1 1 0 1 0 1 1 0 1 1 0 1 1 0 1 1
 FOSA11 1 0 1 0 1 0 1 0 0 1 0 1 1 1 1 1 1 1 1 1 1 0 0 0 1 0 1 1 1 1 1 1 1 1 1 0 1 1 0 1 1 1 1 1
 FOSA12 0 1 1 1 0 0 1 1 0 1 1 1 1 1 1 1 0 1 1 0 0 0 0 1 1 1 1 1 1 0 1 1 0 1 1 1 1 1 0 1 1 0 1 1
 FOSA13 0 0 1 1 0 0 1 0 0 0 1 1 1 1 1 0 1 1 0 0 1 0 0 0 1 1 1 1 1 1 0 0 1 0 1 0 1 1 1 1 1 1 1 1
 FOSA14 0 0 1 1 1 0 1 0 0 0 1 1 1 1 1 0 1 1 1 0 1 0 0 0 1 1 0 1 1 1 1 1 1 1 1 0 1 1 1 1 1 0 1 1
 FOSA15 0 0 1 0 0 0 1 0 1 0 1 1 1 1 1 0 0 1 1 0 1 0 0 1 1 1 1 1 1 1 1 1 0 1 1 0 1 1 0 1 1 0 1 1
 FOSA16 0 1 1 0 1 0 1 1 0 0 0 1 1 1 1 0 1 1 1 1 1 0 0 1 1 1 0 1 1 1 0 1 0 1 1 0 1 1 0 1 1 0 1 1
 FOSA17 0 0 1 1 0 0 0 0 0 0 1 1 1 1 1 0 0 1 1 1 1 0 1 1 1 1 1 1 1 1 1 1 1 0 1 0 1 1 0 1 1 0 1 1
 FOSA18 0 0 1 0 1 0 1 0 0 0 0 1 1 1 1 0 0 1 1 0 1 0 1 1 1 1 1 1 1 1 0 1 1 0 1 0 1 1 0 1 1 0 1 1
 FOSA19 0 1 1 0 1 0 1 0 0 1 0 1 1 1 1 0 0 1 1 0 1 0 0 1 1 1 1 1 1 1 0 1 1 0 1 0 1 1 0 1 1 0 1 1
 FOSA20 1 0 1 0 1 0 1 0 0 1 0 1 1 1 1 0 0 1 1 0 1 0 0 0 1 1 1 1 1 1 0 1 1 1 1 0 1 1 0 1 1 0 1 1
 FOSA21 0 1 1 0 1 0 1 0 0 0 0 1 1 1 1 0 0 1 1 0 1 0 0 1 1 1 1 1 1 1 0 1 1 0 1 0 1 1 0 1 1 0 1 1
 FOSA22 1 0 1 1 1 0 1 0 0 0 0 1 1 1 1 0 0 1 1 1 1 0 1 1 1 1 0 1 1 1 0 1 1 0 1 0 1 1 0 1 1 0 1 1
 FOSA23 1 0 1 1 1 0 1 1 0 0 0 1 1 1 1 0 0 1 1 1 1 0 0 1 1 1 1 1 1 1 0 1 1 0 1 0 1 1 0 1 1 0 1 1
 FOSA24 0 1 1 1 0 0 1 1 0 1 0 1 1 1 1 0 1 1 1 1 1 0 1 1 1 1 1 1 1 1 1 1 1 1 1 0 1 1 0 1 1 0 1 1
 FOSA25 1 1 1 0 0 0 1 0 1 0 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 0 1 1 0 1 0 1 1 0 1 1 0 1 1
 FOSB1 1 0 1 0 0 0 1 1 0 1 1 1 1 1 1 1 1 1 1 1 0 0 1 0 0 1 1 1 1 1 0 1 1 0 0 0 1 1 1 1 1 0 1 1
 FOSB2 1 0 1 1 0 0 1 1 0 1 1 1 1 1 1 1 1 1 1 1 0 0 1 1 0 0 1 1 1 1 0 1 1 0 1 0 1 1 1 1 1 0 1 1
 FOSB3 1 0 1 1 0 1 1 1 0 1 1 1 1 1 1 1 0 1 1 1 0 0 1 1 0 1 0 1 1 1 1 1 1 0 1 0 1 1 1 1 1 0 1 1
 FOSB4 0 0 1 0 0 1 1 1 1 0 0 1 1 1 1 1 0 1 1 1 0 0 0 0 0 1 1 1 1 1 1 1 1 0 1 0 1 1 0 1 1 0 1 1
 FOSB5 0 0 1 0 0 1 1 1 0 1 1 0 1 1 1 1 1 1 1 1 1 0 1 1 0 1 1 1 1 1 1 1 1 0 1 0 1 1 0 1 1 1 1 1
 FOSB6 1 0 1 0 0 1 1 0 0 0 0 0 0 1 1 0 1 1 1 0 1 0 1 1 1 1 1 1 1 1 1 1 1 0 1 0 1 1 0 1 1 0 1 1
 FOSB7 0 0 1 0 1 1 1 0 0 1 1 0 1 1 1 1 1 1 1 0 1 0 1 1 0 0 1 1 1 1 1 1 1 0 1 0 1 1 0 1 1 0 1 1
 FOSB8 0 0 1 1 1 0 1 0 0 1 1 1 1 0 1 1 1 1 1 1 1 0 0 1 0 1 1 1 1 1 1 1 1 0 1 0 1 1 1 1 1 1 1 1
 FOSB9 1 0 1 1 1 0 0 1 0 1 1 1 1 1 1 1 1 1 1 1 1 0 0 0 1 1 1 1 1 1 0 1 1 0 1 0 1 1 0 1 1 0 1 1
 FOSB10 1 0 1 0 1 0 1 0 1 0 0 1 1 1 1 0 1 1 1 1 1 0 1 1 0 0 1 1 1 1 1 0 1 0 1 0 1 1 1 1 1 0 1 1
 FOSB11 1 0 1 1 1 0 1 1 1 1 1 0 0 1 1 0 1 1 1 1 1 0 1 0 0 0 1 0 1 1 1 0 1 0 1 0 1 1 0 1 1 0 1 1
 FOSB12 1 0 1 0 0 0 1 0 0 0 0 1 0 1 1 1 1 1 1 1 1 0 0 0 0 1 1 1 1 1 1 1 1 0 1 0 1 1 1 1 1 1 1 1
 FOSB13 0 0 1 1 1 0 1 1 0 0 1 1 1 1 0 1 1 1 1 1 1 0 1 1 0 1 1 1 1 1 1 1 1 0 1 0 1 1 1 1 1 1 1 1
 FOSB14 0 0 1 0 1 1 1 0 0 1 0 1 1 1 1 1 1 1 1 1 1 0 0 1 0 1 1 1 1 1 0 1 0 1 0 0 1 1 0 1 1 0 1 1
 FOSB15 1 0 1 1 1 0 1 1 0 1 1 1 1 1 1 1 0 1 1 0 0 0 1 1 0 1 0 1 1 1 1 1 1 1 1 0 1 1 0 1 1 0 1 1
 FOSB16 0 0 1 1 0 0 1 1 1 0 1 1 1 1 1 1 1 0 1 0 1 0 0 1 0 1 1 1 1 1 0 1 0 1 1 0 1 1 0 1 1 0 1 1
 FOSB17 1 0 1 1 0 0 1 1 0 1 0 1 1 1 1 0 1 0 1 0 1 0 0 0 1 1 0 1 1 1 0 1 0 1 1 0 1 1 0 1 1 0 1 1
 FOSB18 0 0 0 1 1 0 1 0 0 1 1 1 1 1 1 1 1 1 1 0 1 0 0 1 1 1 1 1 1 1 1 1 0 1 1 0 1 1 0 1 1 0 1 1
 FOSB19 0 0 0 1 1 0 1 1 1 1 0 1 1 1 1 1 1 1 1 1 1 0 0 1 1 1 1 1 1 1 1 1 1 1 1 0 1 1 1 1 1 0 1 1
  
 
 
 
 
 
 
 
 
 
 157 
2. Indigofera zeyheri 
 
TP1 0 1 0 1 1 0 0 1 0 1 1 1 1 1 1 0 0 1 0 1 1 0 1 1 0 1 1 1 1 0 0 1 1 1
 TP2 0 0 0 0 1 0 0 0 0 0 0 1 1 1 0 1 0 1 0 1 0 0 1 1 0 0 1 1 1 1 0 1 1 1
 TP3 0 0 0 1 0 1 0 1 0 1 0 1 1 1 0 0 0 1 0 0 1 0 1 1 0 1 1 1 1 0 0 1 1 1
 TP4 0 0 0 0 0 0 0 1 0 0 0 1 1 1 0 0 0 1 0 1 0 1 1 1 0 1 1 1 0 1 0 1 1 1
 TP5 0 1 0 1 1 1 0 1 0 1 1 1 1 1 0 0 1 1 0 1 1 0 1 1 0 0 1 1 1 0 0 0 1 1
 TP6 0 1 0 1 0 0 0 0 0 1 1 1 1 1 0 0 0 1 0 0 0 0 1 1 0 0 1 1 0 1 0 1 1 1
 TP7 0 1 0 1 1 1 0 1 0 1 0 1 1 1 1 0 1 1 0 1 1 0 1 1 1 0 1 1 0 1 1 1 0 1
 TP8 0 1 0 1 0 1 0 0 0 1 0 1 1 1 0 0 0 1 0 1 1 0 1 1 0 0 1 1 1 1 1 1 0 1
 TP9 0 1 0 1 1 0 0 0 0 1 0 1 1 1 0 0 0 1 0 1 1 1 1 1 1 1 1 1 1 1 0 0 0 1
 TP10 0 0 0 0 1 0 0 0 0 1 0 1 1 1 0 0 1 1 0 1 1 0 1 1 1 1 1 1 1 1 1 1 0 1
 TP11 0 1 0 1 1 0 0 0 0 1 1 1 1 1 0 0 0 1 0 1 1 0 1 1 0 1 1 1 1 1 0 1 0 1
 TP12 0 1 1 1 1 0 0 1 1 1 0 1 1 1 1 1 0 1 0 1 1 1 1 1 1 1 1 1 1 1 0 1 0 1
 TP13 0 1 0 1 0 0 0 0 0 1 0 1 1 1 0 1 0 1 0 1 1 1 1 1 0 0 1 1 0 1 1 1 0 1
 TP14 0 1 1 0 0 1 0 0 1 1 0 1 1 1 1 0 0 1 0 1 1 0 1 1 1 1 1 1 1 1 1 1 0 1
 TP15 0 1 0 1 1 1 0 0 0 1 1 1 1 1 1 0 0 1 0 1 1 0 1 1 1 1 1 1 1 1 1 1 0 1
 TP16 0 1 0 1 1 0 0 0 0 1 1 1 1 1 0 1 0 1 0 1 1 0 1 1 1 0 1 0 0 0 0 0 0 1
 TP17 0 0 0 1 0 0 0 0 0 1 0 1 1 1 1 0 0 1 0 1 1 0 1 1 1 1 1 0 0 1 0 0 0 1
 TP18 0 0 0 0 0 0 0 0 0 0 1 1 1 1 0 0 0 0 0 1 1 0 1 1 0 1 1 1 0 1 0 1 0 1
 TP19 0 1 0 0 1 1 0 0 0 0 1 1 1 1 0 0 0 1 0 1 1 0 1 1 1 1 1 0 1 0 0 1 0 1
 TP20 0 1 0 1 1 0 0 0 0 1 0 0 1 1 1 0 0 0 0 1 1 0 1 1 0 0 1 1 0 1 0 1 0 1
 TP21 0 1 0 1 1 0 0 0 0 1 0 1 1 1 1 0 0 0 0 1 1 0 1 1 0 0 1 1 0 1 0 1 0 1
 TP22 0 1 0 1 1 0 0 0 0 1 0 0 1 1 0 0 0 1 0 1 1 0 1 1 0 0 1 1 1 1 0 1 0 1
 TP23 0 1 0 1 1 1 0 1 0 1 0 1 1 1 0 0 0 1 0 1 1 0 1 1 0 1 1 1 0 1 0 1 0 1
 TP24 0 1 1 1 1 1 0 0 0 1 1 1 1 1 0 0 0 1 0 1 1 0 1 1 1 0 1 1 1 1 0 1 0 1
 TP25 0 1 0 0 1 1 0 0 0 1 0 1 1 1 0 1 1 1 0 1 1 0 1 1 0 0 1 1 0 1 0 1 0 1
 TP26 0 0 0 1 1 0 0 1 0 1 0 1 1 1 0 0 0 0 0 1 1 0 1 1 0 0 1 1 0 1 0 1 0 1
 OS1 0 1 0 0 1 0 0 0 0 1 1 1 1 1 1 0 0 0 1 1 1 0 1 1 0 1 1 1 0 1 0 1 1 1
 OS2 0 0 0 0 1 1 0 1 0 1 1 1 0 1 1 0 0 1 0 1 1 0 1 1 0 0 1 1 1 0 0 1 1 1
 OS3 0 1 0 0 1 0 0 0 0 1 0 1 0 1 0 0 1 1 0 1 1 0 1 1 0 1 1 1 0 0 0 1 1 1
 OS4 0 1 0 1 1 0 0 1 0 1 0 1 0 1 0 0 0 1 0 1 1 1 1 1 0 1 1 1 0 0 0 0 1 1
 OS5 0 1 0 0 1 0 0 1 0 1 1 1 1 1 0 0 1 1 0 1 1 0 1 1 0 1 1 1 0 1 1 1 1 1
 OS6 0 1 0 0 1 1 1 1 0 1 1 1 1 1 0 0 1 1 0 1 1 0 1 1 1 0 1 1 1 1 1 0 0 1
 OS7 0 1 0 0 0 1 0 0 0 1 0 1 1 1 0 0 1 1 0 1 1 0 1 1 0 0 1 1 0 1 0 1 0 1
 OS8 0 1 0 1 1 0 0 1 0 1 0 1 1 1 0 0 0 1 0 1 1 0 1 1 1 1 1 1 0 1 0 1 0 1
 OS9 0 1 1 1 1 1 0 1 1 1 0 0 1 1 1 1 0 1 0 1 1 0 1 1 1 0 1 1 1 1 0 1 1 1
 OS10 1 1 0 0 1 0 0 0 0 1 0 1 1 1 0 0 0 1 0 1 1 0 1 1 1 0 1 1 1 1 1 1 1 1
 OS11 0 1 0 1 0 0 0 0 1 1 0 1 1 1 0 1 0 1 0 1 1 0 1 1 1 0 1 1 0 1 0 1 0 1
 OS12 0 1 0 1 1 0 0 1 0 1 0 1 1 1 0 1 0 1 0 1 1 0 1 1 0 0 1 1 0 1 0 1 0 1
 OS13 0 1 0 0 1 1 0 0 0 1 1 1 1 1 1 0 0 0 0 1 1 0 1 1 0 1 1 0 0 0 0 1 0 1
 OS14 0 1 0 0 1 1 0 0 0 0 1 1 1 1 0 0 0 0 0 1 1 0 1 1 0 1 1 1 0 1 0 1 0 1
 OS15 0 1 0 0 1 1 0 0 0 0 1 1 1 1 0 0 0 1 0 1 1 0 0 1 1 1 1 0 1 0 0 0 0 1
 OS16 0 1 0 1 1 1 0 0 0 1 0 1 1 1 0 0 0 1 0 1 1 0 1 1 0 1 1 1 0 1 0 1 0 1