Mistletoes as drivers of plant community structure and resource heterogeneity in semi-arid savanna ecosystems, Zimbabwe

Abstract

The role of mistletoes in influencing resource heterogeneity has been shown in many environments. Studies have shown that by providing additional litter and weakening their hosts', mistletoes can change plant community structure by providing sub-ordinate subcanopy species a competitive edge thereby increasing species richness and diversity. These studies have mainly focused on differences between mistletoe-infected and uninfected trees, yet it is possible that different mistletoe-infection degrees' influence species composition within and beyond their canopies. Therefore, this thesis investigated the effects of mistletoe-infection degrees on Vachellia karroo (Hayne) Banfi and Glasso on plant community structure in a semi-arid savanna in South-West, Zimbabwe. Firstly, this thesis investigated how high- and low mistletoe-infection degrees on V. karroo trees influence abiotic and biotic conditions within and beyond the canopy patches. Further, it examined whether there were variations in species composition, species and functional diversity, and size measurements of understory plants within high- and low mistletoe-infection canopy patches and intercanopy spaces. Thirdly, it explored whether different mistletoe-infection degrees reduced the reproductive and regeneration capacities of their host trees and the recruitment of host juveniles (seedlings and saplings). Lastly, although studies have investigated how uninfected mature trees influence the spatial distribution of other surrounding woody species of different stage classes, little is known on how mistletoe-infected trees influence the spatial patterns of surrounding woody plants. Therefore, the fourth aim investigated how mistletoe-infected V. karroo trees influence the spatial patterns of their surrounding conspecific and heterospecific woody plants of different stage classes within three 50m × 50m plots with > 30 mistletoe-infected V. karroo trees. The results show that intercanopy spaces had between 18% to 34%, and 18% higher herbaceous biomass and maximum grass height, respectively, compared to canopy patches. Herbaceous biomass and maximum grass height were 8% to 23%, and 13%, respectively, higher in low- compared to high mistletoe-infection microhabitats. Furthermore, high mistletoe-infection canopy patches had between 29% and 49%, and 30% less herbaceous biomass and maximum grass height, respectively compared to low mistletoe-infection intercanopy spaces, which had the highest measurements. Some of these differences are associated with higher grazing/trampling of between 1.24 and 1.39-fold within high mistletoe-infection canopy patches compared to the other three microhabitats. High animal disturbances contributed to the elevated species richness and species and functional diversity (by reducing understory competition) within high mistletoe-infection canopy patches compared to the other microhabitats, as well as being attracted to the shade from the midday sun in all subcanopy patches. Intercanopy spaces were mainly dominated by high grazing value grasses (e.g., Setaria incrassata and Heteropogon contortus) and V. karroo juveniles, whilst canopy patches particularly of high mistletoe-infection had significantly higher grass (of mixed grazing value), forb, and tree diversity. A ‘mistletophily index’, which calculated the affinity of each species to each of the four different microhabitats, was developed. The index showed that 34% of the recorded species had a strong affinity towards canopy patches whilst intercanopy spaces were significantly associated with 9% of the observed species. Low mistletoe-infection canopy patches had higher abundance of decreaser grasses which are associated with low-intermediate disturbance, and intermediate disturbance is often linked to higher species diversity. However, grass, forb, and tree diversity were 17% to 43% higher within high mistletoe-infection canopy patches with higher animal disturbances compared to low mistletoe infected canopy patches. These variations were attributed to higher soil temperature and relative humidity (3% and 0.5%, respectively), measured continuously over 8 months from the wet to the dry season using ibuttons, within high- compared to low mistletoe-infection canopy patches. The raised vii soil temperature and relative humidity is interpreted as a result of higher light incidence and decomposition rates from the relatively higher litter turnover, leading to canopy patches with higher soil nutrients and increased rates of nutrient cycling compared to low mistletoe-infection canopy patches. Therefore, high mistletoe-infection canopy patches had a higher occurrence of species that favour semi-shade, high soil moisture and nutrients (Sporobolus pyramidalis, Asparagus africanus, Ziziphus mucronata and Flueggea virosa) and those that are prevalent on disturbed sites (Cynodon dactylon, Setaria verticillata, Bidens pilosa, Sida alba, and Lantana camara). As a result, 15% of the recorded species showed a strong positive affinity to high mistletoe infection canopy patches, whilst only 10% of the species had a high affinity towards low mistletoe-infection canopy patches. Indeed, an increase in mistletoe infection and canopy presence resulted in different species assemblages, higher species diversity (25% to 45%), functional richness (27% to 42%), functional evenness (7% to 30%), functional dispersion (24% to 58%) and RaoQ (26% to 59%) compared to the other microhabitats. Mistletoe-infection intensity significantly reduced the reproductive and regeneration capacities of V. karroo trees. As mistletoe-infection increased there tended to be a general decline in flower buds, flowers, pods, seeds/pod and most importantly seed and germinable seed production/tree, possibly due to changes in the canopy physio-morphological attributes. Flower buds, flowers, and pods were between 40% and 68% higher in low- compared to high mistletoe-infection canopies. Pods were 1.16- fold longer, whilst the number of seeds/pod was 1.15-fold higher in low- than high mistletoe-infection trees. Consequently, seed and germinable seed production/tree were significantly lower in high- (2273 ± 820; 599 ± 216, respectively) than low mistletoe-infection trees (7088 ± 905; 2947 ± 376). The percentage of intact seeds was higher in low- (70%) compared to high mistletoe-infection trees (58%). High mistletoe-infected seeds had higher bruchid beetle seed predation (21%) and aborted seeds (21%) compared to low mistletoe-infection seeds (11% and 19%, respectively). Seed mass per seed was 20% higher (but not significantly different) in low- (0.037 ± 0.003g) compared to high mistletoe infected trees (0.030 ± 0.003g). The overall percentage germination was 42% and 26% for high- and low mistletoe-infected seeds. However, high mistletoe-infection seeds (12%) had 4-fold higher initial germination rates (pre-scarification) than low infection seeds (3%), whilst after scarification low mistletoe-infection seeds (34%) had 2.83-fold higher germination rates than high mistletoe-infection seeds (12%). This was attributed to variations in seed size and seed coat characteristics. The overall germination rates were very similar at 36.3 ± 2.54 and 36.1 ± 2.57 days for high and low infection seeds, respectively. The soil seed bank was 3.24-fold higher under low- (46.7 ± 10.7 seeds/m2 ) than high mistletoe-infection trees (14.4 ± 5.8 seed/m2 ). The number of understory V. karroo juveniles was 1.41 to 3.51-fold higher within high mistletoe infection canopy (n = 172) patches compared to the three microhabitats. There were no significant differences in the juvenile densities across the microhabitats, however, juvenile densities tended to be 2.86-fold higher in high- (1813 ± 528 juveniles/ha) compared to low mistletoe-infection canopy patches (633 ± 218 juveniles/ha) which had the lowest juvenile densities. Seedlings were significantly higher in high mistletoe-infection canopy patches (1275 ± 344 seedlings/ha) followed by adjacent intercanopy spaces (663 ± 121 seedlings/ha), which were much higher than in the low mistletoe infection canopy patches (294 ± 92 seedlings/ha) and adjacent intercanopy spaces (293 ± 160 seedlings/ha). Although sapling density was higher in low- (706 ± 343 saplings/ha) and high (573 ± 199 saplings/ha) mistletoe-infection intercanopy spaces compared to low- (340 ± 142 saplings/ha) and high mistletoe-infection canopy patches (538 ± 201 saplings/ha), there was no significant difference across microhabitats. This indicates that over time V. karroo juveniles are actually persisting better in the intercanopy spaces than under the tree canopies. Therefore, intercanopy spaces were safer for viii saplings, compared to the adjacent canopy patches that were apparently safer sites for seedlings, at least prior to the cold dry winter period. Moreover, high mistletoe-infection canopy patches were safer for V. karroo juveniles, compared with low mistletoe-infection canopy patches which were safer sites for soil-stored seeds. Lastly, mistletoe infected V. karroo trees had varying relationships with heterospecifics and conspecifics of different stage classes. Mistletoe-infected trees exhibited patterns consistent with a random pattern; this was attributed to bird disperser choices and the already existing random patterns of mature trees. However, due to facilitation, most of the woody conspecific and heterospecifics of seedlings, saplings, and shrubs were clustered around mistletoe-infected trees. Nonetheless, one plot showed significant repulsion of conspecific seedlings and saplings which could be due to both intra and inter-specific competition. The bivariate relationship between all the mature trees and conspecific mature trees showed a random distribution consistent with the existence of both competition and facilitation. Competition can be attributed to trees with low mistletoe-infection competing with uninfected trees thus inclining the pattern towards regular distribution. In contrast, facilitation could be signifying the weakening of hosts due to high mistletoe-infection intensities coupled with a proliferation in soil nutrients and moisture. This increases the competitiveness of sub-ordinate trees, consequently leaning the pattern towards aggregation. Overall, this part of the study shows that mistletoe infection can increase species and functional diversity and alter the spatial patterns of their nearest neighbours of different growth forms. They can augment plant heterogeneity and functional richness which can positively influence the persistence of semi-arid savanna ecosystems which are nearly always resource-limited.