Abstract
Questions: Does vegetation on the dried Aral Sea bed resolve into recurrent floristic configurations that are robust to analytical choices, and do these configurations show structural and phylogenetic differentiation relevant to competing models of vegetation development? Location: Desiccated bed of the Aral Sea (Aralkum), Kazakhstan and Uzbekistan, Central Asia. Methods: We analyzed 577 vegetation relevés (10 m 2 ) containing 233 vascular plant taxa. Modified TWINSPAN classifications were compared across eight preprocessing variants and evaluated by repeated subsampling. The retained partition was examined using ordination, multivariate tests, diagnostic species, growth‐form structure, and Sørensen beta‐diversity partitioning. Correspondence with 17 ecological types from the historical Uzgiprozem typology was assessed, and phylogenetic structure was analyzed from a 232‐taxon phylogeny using Faith's PD, SES.MPD, and SES.MNTD. Results: Six recurrent floristic configurations were retained, with unequal reproducibility and separation. All six occurred in both sectors and in at least two survey periods. After accounting for survey period, data source, latitude, and longitude, vegetation group explained an additional 16.6% of Bray–Curtis variation. The configurations differed in local richness and growth‐form structure. Mean Sørensen dissimilarity was 0.905, with 94.3% attributable to turnover. The Uzgiprozem typology showed substantial but incomplete correspondence with the six‐group classification. Phylogenetic diversity and standardized phylogenetic structure also differed among configurations; Faith's PD was strongly related to species richness, whereas most relevés did not depart from null expectations. Conclusions: Vegetation across the Aralkum is organized into six recurrent but unequally separated floristic configurations. Their diagnostic composition, growth‐form structure, high species turnover, historical‐typological correspondence, and phylogenetic differentiation provide an empirical framework for competing models of vegetation development. The six‐group classification can be used to test whether vegetation change is governed primarily by environmental differentiation, a broadly shared successional sequence, or contrasting developmental trajectories.
Funding Information
  • Academy of Sciences Republic of Uzbekistan