Differential contribution of frugivorous birds to dispersal patterns of the endangered Chinese yew (Taxus chinensis)

A leading-edge scenario in the phylogeography and evolutionary history of East Asian insular Taxus in Taiwan and the Philippines

The cool temperate origin of gymnosperm Taxus species in East Asia is specifically diverse and widespread. Certain lineages have managed to extend their distribution further south to subtropical and tropical islands such as Taiwan and the Philippines. To address questions including whether these insular lineages, recently identified as T. phytonii, have become genetically distinct from each other and from their continental relatives, and when and how they colonized their residing islands, we sampled over 11 populations, covering 179 Taxus individuals from Taiwan and the Philippines. Using four cpDNA and one nuclear marker, we showed in population genetic and genealogical analyses that the two insular lineages were genetically distinct from each other and also from other continental Taxus and that they represented each other's closest relative. Estimated with the coalescent-based multi-type tree (MTT) analyses, we inferred an origin of Taiwanese T. phytonii more ancient than 2.49 Mya and that of Philippine T. phytonii more ancient than 1.08 Mya. In addition, the divergence demographic history revealed by both MTT and isolation with migration (IM) analyses indicated the presence of recent post-split migrations from a continental taxon, T. mairei, to Taiwanese T. phytonii, as well as from Taiwanese T. phytonii to Philippine T. phytonii. Overall, this study suggests Taiwan as a stepping stone through which the temperate-origin yew trees can extend their distributions to tropical regions such as the Philippines.

Importance of species traits on individual-based seed dispersal networks and dispersal distance for endangered trees in a fragmented forest

Although mutualistic network analyses have sparked a renewed interest in the patterns and drivers of network structures within communities, few studies have explored structural patterns within populations. In an endangered tree species population, plant individuals share their bird seed dispersers; however, the factors affecting individual interaction patterns are poorly understood. In this study, four individual-based networks were built for the endangered Chinese yew, Taxus chinensis, in a fragmented forest based on bird foraging type (swallowing and pecking networks) and habitat type (networks in a bamboo patch and an evergreen broad-leaved forest patch). Species-level network metrics (species degree and specialization, d') were used to evaluate the effects of species traits (bird and plant traits) on species-level networks and dispersal distance for T. chinensis. It was revealed that the interaction networks between T. chinensis individuals and their bird partners were influenced by foraging type and the habitat of plant distribution. Compared to the other two networks, bird swallowing and bird-fruit networks in the evergreen broad-leaved patch habitat had higher nestedness and connectance but lower modules and specialization. Bird (body weight and wing and bill lengths) and plant traits (height, crop size, and cover) significantly affected species-level network metrics such as degree and specialization. Furthermore, seed dispersal distance was influenced by species traits and the species-level metrics of fruit-bird interaction networks. These results provide new insights into individual-based seed dispersal mutualistic networks of endangered plant species under habitat fragmentation. Moreover, these findings have relevant implications for conserving and managing individual endangered trees in increasingly disturbed ecosystems.

Response of distribution patterns of two closely related species in Taxus genus to climate change since last inter-glacial

Climate change affects the species spatio-temporal distribution deeply. However, how climate affects the spatio-temporal distribution pattern of related species on the large scale remains largely unclear. Here, we selected two closely related species in Taxus genus Taxus chinensis and Taxus mairei to explore their distribution pattern. Four environmental variables were employed to simulate the distribution patterns using the optimized Maxent model. The results showed that the highly suitable area of T. chinensis and T. mairei in current period was 1.616 × 10⁵ km² and 3.093 × 10⁵ km², respectively. The distribution area of T. chinensis was smaller than that of T. mairei in different periods. Comparison of different periods shown that the distribution area of the two species was almost in stasis from LIG to the future periods. Temperature and precipitation were the main climate factors that determined the potential distribution of the two species. The centroids of T. chinensis and T. mairei were in Sichuan and Hunan provinces in current period, respectively. In the future, the centroid migration direction of the two species would shift towards northeast. Our results revealed that the average elevation distribution of T. chinensis was higher than that of T. mairei. This study sheds new insights into the habitat preference and limiting environment factors of the two related species and provides a valuable reference for the conservation of these two threatened species.

Random placement models explain species richness and dissimilarity of frog assemblages within Atlantic Forest fragments

Understanding the effects of random versus niche-based processes on biodiversity patterns is a central theme in ecology, and an important tool for predicting effects of habitat loss and fragmentation on biodiversity. We investigated the predictive power of random processes to explain species richness and species dissimilarity of amphibian assemblages in a fragmented tropical landscape of the Atlantic Forest of South America. We analyzed a large database of amphibian abundance and occupancy, sampled in 21 forest fragments ranging in size from 1.9 to 619 ha. We compared observed species richness and species dissimilarity with the outcomes of two null (random placement) models: 1- the traditional Coleman's area-based model and 2- an abundance-based model (based on the number of individuals observed in each fragment). We applied these models for all species combined, and separately for forest-dependent and habitat-generalist species. The abundance-based model fitted the observed species richness data better than the area-based model for all species, forest-dependent species, and generalist species. The area-based and the abundance-based models were also able to significantly explain species dissimilarity for all species and for generalists, but not for forest dependent species. The traditional area-based model assigned too many individuals to large fragments, thus failing to accurately explain species richness within patches across the landscape. Although niche-based processes may be important to structuring the regional pool of species in fragmented landscapes, our results suggest that part of the variation in species richness and species dissimilarity can be successfully explained by random placement models, especially for generalist species. Evaluating which factors cause variation in the number of individuals among patches should be a focus in future studies aiming to understand biodiversity patterns in fragmented landscapes.

Importance of bird traits for seed dispersal patterns of co-fruiting trees in a patchy forest

Habitat fragmentation is globally one of the most important drivers of change in biodiversity. Seed dispersal by birds is crucial for tree regeneration in remnant patchy forests, yet how bird traits affect seed dispersal pattern is still poorly understood. We studied the extent to which bird traits affect seed-removal networks and whether these traits affect seed deposition and seedling recruitment for 3 co-fruiting tree species (Taxus chinensis, Cinnamomum bodinieri and Machilus thunbergii) in a patchy forest. A total of 17, 18 and 10 bird species were recorded foraging for seeds of T. chinensis, M. thunbergii and C. bodinieri, respectively. Frequency of bird visitation increased with tail length, wing length and body length. Furthermore, bird body length, bill length, body weight and wing length were important in the strength of the seed removal network. During foraging, 6 bird species exhibited different patterns of microhabitat utilization and their perching frequency increased with bird weight and tarsus length. As a consequence, frequency of habitat use, bird length and tarsus length were important in determining the number of deposited seeds. For seedling recruitment, seedling number increased with bird tarsus length and weight, but decreased with wing length. Overall, our results showed that various bird traits not only affected seed removal, but also influenced the subsequent processes of seed deposition and seedling distribution in a patchy forest. These results thus highlight the importance of large-bodied birds for plant recruitment and point out the need to prioritize the protection and conservation of these birds in remnant patchy forests.

Impact of multiple bird partners on the seed dispersal effectiveness of China's relic trees

Frugivorous birds generally exhibit an unequal contribution to dispersal effectiveness of plant species as a function of their habitat adaptation and body size. In our study, we compared the effectiveness of multiple bird species that contribute to the dispersal of the endangered relic Chinese yew, Taxus chinensis. Seven bird species dispersed T. chinensis seeds, with Picus canus, Turdus hortulorum, and Urocissa erythrorhyncha being the main dispersers. The quantity part of dispersal effectiveness was strongly influenced by two inherent characteristics of disperser species: body size and habitat adaptation. However, the quality part of dispersal effectiveness was only influenced by disperser type. For instance, small generalist birds and large specialist birds removed more seeds than other type dispersers. Moreover, small birds and specialist birds contributed slightly more to the dispersal quality of T. chinensis than large birds and generalist birds respectively; however, these differences were not significant. Our results suggest that dispersal effectiveness is affected by variety in the body size and habitat adaptation of different dispersers. Therefore, such variation should be incorporated into spatial and temporal management actions of relic plant species in patchy, human-disturbed habitats.