Abstract:

Aims As an anomalous feature of the species-area relationship (SAR), the small-island effect (SIE) is an important pattern in the fields of island biogeography and biodiversity science. However, the existing work has rarely explored the SIE in fragmented mountaintops. Moreover, the underlying factors determining the SIE remain largely untested. In this paper, we aimed to evaluate the prevalence and underlying factors determining the occurrence of SIEs in mosses on mountaintops of the Middle Taihang Mountains.
Methods We investigated and compiled data on the species richness of all mosses and six most common moss families on 19 mountaintops of the Middle Taihang Mountains. For each of the seven taxonomic groups, we applied six SAR models, including four piecewise regressions with two segments, a linear regression, and a zero-slope regression as a null model, to detect the SIE and then used the corrected Akaike’s information criterion (AICc) as a criterion to select the best model. We obtained three environmental variables, including island height, annual temperature range, and net primary productivity per unit area that are linked to the habitat diversity hypothesis, the extinction hypothesis, and the subsidized island biogeography hypothesis, respectively, in order to explore the underlying drivers of the SIE. We applied iterative multiple linear regression and variation partitioning to determine the effects of island area and each of environmental variables on species richness for islands of varying sizes. Finally, we analyzed the trends of environmental variables’ contributions to species richness variation over the iterative process by using generalized linear regressions with polynomials of area up to the second degree.
Important findings A total of 131 moss species, belonging to 23 genera under 68 families, were identified. SIEs were detected for all mosses and six most common moss families, with the area thresholds varing between 0.36 km² and 106.91 km². The driving factors of the SIE differentiate among groups. Among them, Grimmiaceae did not support the habitat diversity hypothesis; Pottiaceae, Bryaceae, Grimmiaceae, and Hypnaceae did not support the extinction hypothesis; whereas, the subsidized island biogeography hypothesis was generally supported by each group to varying degrees. On the whole, mountaintops with an area of ​​over 10 km² maintain a large diversity of moss species and thus be identified as site conservation targets. For mosses with relatively unique habitat requirement, such as Grimmiaceae, protecting the stony habitat is the key to maintain its species diversity; as for other mosses, conservation of habitat diversity is the key to maintain their species diversity. The general applicability of the subsidized island biogeography hypothesis suggests that the resource replenishment of the forest ecosystem below the mountaintops counteracts the removal of mosses’ species richness due to area reduction. Therefore, conservation of species diversity and community stability in the forest below the mountaintops is essential for sustaining moss species diversity in fragmented mountaintops.

Key words: species-area relationship, small-island effect, area threshold, piecewise regression, iteration, variation partitioning