Termite mound vegetation patterns are largely driven by stochastic rather than deterministic processes at regional scale in Africa: A meta‐analysis. (15th May 2020)

• Record Type: Journal Article
• Authors: Muvengwi, Justice; Witkowski, Edward T. F.
• Editors: Ward, David
• Is Part Of: Journal of vegetation science. Volume 31:Number 4(2020:Jul.)
• Page Start: 540
• Page End: 550
• Publication Date: 2020-05-15
• Subjects: altitudinal gradient -- beta diversity -- ecosystem heterogeneity -- latitudinal gradient -- Macrotermes -- nestedness -- species richness -- species turnover -- termite mounds; Plant ecology -- Periodicals; Plant communities -- Periodicals; Plant populations -- Periodicals; 581.7
• DOI: 10.1111/jvs.12884
• Languages: English
• ISSNs: 1100-9233

Abstract

Abstract: Questions: Although the biogeography of plant communities has been well studied, mechanisms controlling plant community diversity are still poorly understood. Stochastic and deterministic processes are hotly debated as being key in structuring communities across taxa. Here we ask, to what extent are plant communities on termite mounds shaped by either stochastic or deterministic processes within and between regions on the African continent? Methods: We performed a meta‐analysis on 23 studies that satisfied our selection criteria out of 369 that we initially identified. All these studies sampled woody plants that occurred on termite mounds and the surrounding matrix. In order to disentangle the relative importance of deterministic and stochastic processes in generating beta diversity between mound–mound and matrix–matrix plots, within and between regions, we applied null models. We also assessed the relationship between species richness and environmental variables for each site. Results: Of the environmental variables assessed at each site, only altitude had a significant (negative) relationship with species richness. Within each region comparison of mound–mound, matrix–matrix and matrix–mound species composition did not significantly deviate from the null models of random expectation. Between regions, the observed pattern was significantly higher than null expectation for both mound–mound and matrix–matrix plots between East and West Africa. Surprisingly, the comparison between southern Africa and both East and West Africa did not significantly deviate from the null expectation, regardless of increased spatial separation. We attributed the within‐region pattern to homogenizing dispersal and reliance on the same species pool (γ) for matrix and mound plots. The between‐regions pattern was influenced by both deterministic and stochastic processes. Conclusion: Our results suggest that the processes responsible for vegetation composition between mound–mound, matrix–matrix and mound–matrix plots are stochastic in nature due to lack of difference from the null expectation. However, regional patterns are governed by both deterministic and stochastic processes. Abstract : Our results suggest that the processes responsible for vegetation composition between mound–mound, matrix–matrix and mound–matrix plots are stochastic in nature due to lack of difference from the null model expectation. However, regional patterns are governed by both deterministic and stochastic processes. The change in composition of woody plant communities between mound–mound, matrix–matrix and matrix–mound plots was dominated by turnover within and between regions.