Electronic Theses and Dissertations (PhDs)

Permanent URI for this collection

https://hdl.handle.net/10539/38000

Browse

Browsing Electronic Theses and Dissertations (PhDs) by Author "Balkwil, Kevin"

Now showing 1 - 1 of 1
  • Thumbnail Image
    Item
    Understanding disturbance, vegetation density, seed banks and pollination for the conservation of Protea curvata
    (University of the Witwatersrand, Johannesburg, 2023) Mabuza, Precious Gugulethu Babalwa; Balkwil, Kevin
    Background Aims: Protea curvata (Proteaceae) is a threatened species endemic to Mpumalanga, South Africa. Previous records of the species showed discrepancies in location data and information on population demographics was sparse. At the time of the study (2018), the last IUCN assessment of P. curvata was 20 years ago and one subpopulation of the species was reported to have a low number of mature individuals. In 2017, a census of another subpopulation indicated that the mature individuals were the most abundant cohort and there was no recruitment of P. curvata seedlings. Thus, there was uncertainty regarding how population size, population demographics and threats faced by the species may have changed over the last 20 years. The study aimed to update the IUCN status of P. curvata and to assess factors relating to P. curvata recruitment, namely the breeding biology of the species and the woody species composition on P. curvata sites. Methods: A census was conducted on P. curvata sites. Census data and site composition was compared between sites with contrasting management. Soil samples were collected from six sites. Pollination modes in P. curvata were investigated by manually pollinating inflorescences that were covered at bud phase to exclude animal pollinators. Five treatments were tested: autogamous selfing, tree geitonogamous selfing, inflorescences geitonogamous selfing, natural outcrossing and pollen supplemented outcrossing. Seeds from each treatment were weighed and tested for viability using TTZ staining. Animal pollinators were observed in the field and through camera traps positioned in front of non-covered inflorescences. Key Results: Five subpopulations were identified. P. curvata recruitment was low in two subpopulations and absent in three subpopulations. When comparing the site with the lowest recruitment (Site A) to the site with the highest recruitment (Site C), woody cover and tree density was higher on the site with low recruitment – indicating bush encroachment. This was attributed primarily to long intervals between fires at Site A. Site A showed a concerning net decline in subpopulation size (3% loss per year). Despite having more P. curvata juveniles, Subpopulation C showed a similar rate of decline (2% loss per year). The species was identified as endangered; with severe hailstorms and delayed, intense fires posing the major threats. Hailstorms compounded fire-induced damage on P. curvata bark. Flowering declined significantly after a hailstorm. (x̄pre-hail = 10.94 inflorescences per tree, x̄post-hail = 4.28 inflorescences per tree, p-value = 0.0031). During pollination experiments, Protea curvata was able to self-pollinate. However, seed mass and viability were significantly higher in outcrossed treatments. Viability in naturally outcrossed treatments (x̄=42.7%) was similar to pollen-supplemented outcrossed treatments (x̄=32.1%), but significantly higher than treatments of autogamous selfing (x̄=21.95%), geitonogamous selfing within inflorescences (x̄=27.1%) and geitonogamous selfing within trees (x̄=26.2%) [C.I= 95%; d.f.= 4, 95; Fcalc= 4.5; p = 0.0021]. Colletid bees and Apid bees contributed to P. curvata pollination via geitonogamy and early removal of self-pollen from stigmas. Birds were the most frequent visitors of P. curvata and were the most effective pollinators due to their prominent role in natural outcrossing. Bats infrequently visited inflorescences but may possibly complement outcrossing since they make contact with stigmas and do not spend their visit time restricted to one inflorescence or tree. P. curvata was non-serotinous and possessed a short-lived, soil seed bank. Across all sites, the average viability of seeds in the soil was well below that of freshly collected seeds from pollination experiments (4–9% for soil seed bank; 22.0–42.6% for fresh seeds). Seed bank size was surprisingly low in subpopulations with high flowering. Conclusions: Reducing seed loss after release from the canopy appeared to be more important than high flowering for maintaining a large seed bank. Therefore mild, frequent fires will be essential for reducing bush encroachment, making way for seed deposition and encouraging seedling survival