Abstract
AIM: Localised variation in soil properties can play an important role in shaping vegetation structure and plant community composition. However, in degraded ecosystems, these vegetation patterns may not be apparent due to the homogenization of local plant communities. Thus, defining restoration targets may prove challenging. By comparing the root development of Portulacaria afra (L) Jacq. cuttings grown in three different soils collected from degraded subtropical succulent thicket habitats, we aim to test whether soil chemical properties act as an ecological filter limiting root growth, which may in turn influence community composition and restoration success. This study focuses on root biomass differences as a proxy for potential vegetation structure rather than directly assessing plant community composition. LOCATION: Eastern Cape, South Africa. TAXON: Portulacaria afra (L) Jacq. METHODS: Soil was collected from the top 20 cm at three sites within a degraded succulent thicket landscape (two with historically closed canopies, and one with a historically open woody canopy). A total of 196 P. afra cuttings were grown in each soil (588 cuttings in total) across two growing conditions (glasshouse and growth chamber). Root development was evaluated by destructive harvesting of 14 cuttings per treatment (soil type and growth condition) per week, and root dry weight was compared across treatments for each harvest date. The soil properties from each site were analysed to identify possible drivers for any differences in root development and visualised via a principal components analysis. RESULTS: Significant differences in root dry weight were detected (all tests: F(5,74) = 4.11-11.45, p < 0.01). Root biomass was significantly lower in cuttings grown in soil from Site C (calcareous; historically open canopy) compared to Sites A and B (slightly saline and non-saline, respectively; historically closed-canopy thicket), suggesting that edaphic factors may have historically influenced vegetation structure. The soil from Site C showed notable differences from the soils at Sites A and B, with a higher pH (7.9 vs 6.5, 6.8, respectively), increased Ca(2+) concentration (25.4 vs 8.8, 6.4 cmol(+)/kg), Ca saturation % (83 vs 62, 53), and a lower P concentration (<2.2 vs 116, 43 mg/kg). These factors, particularly high pH and low P availability, likely suppressed root initiation and development, which may limit P. afra establishment in restoration efforts on calcareous soils. CONCLUSION: Local variation in soil properties plays an important role in the regeneration dynamics and restoration of succulent thicket vegetation. Calcareous soils likely supported an open canopy vegetation with relatively low P. afra cover. This possible vegetation structure should be accounted for when setting restoration targets and measuring restoration success.