Large-scale deforestation in the tropics, triggered by logging and subsequent agricultural monoculture has a significant adverse impact on biodiversity due to habitat degradation. Here, we measured the diversity of butterfly species in three agricultural landscapes, agroforestry orchards, oil palm, and rubber tree plantations. Butterfly species were counted at 127 sampling points over the course of a year using the point count method. We found that agroforestry orchards supported a greater number of butterfly species (74 species) compared to rubber tree (61 species) and oil palm plantations (54 species) which were dominated by generalist (73%) followed by forest specialists (27%). We found no significant difference of butterfly species composition between agroforestry orchards and rubber tree plantation, with both habitats associated with more butterfly species compared to oil palm plantations. This indicates butterflies were able to persist better in certain agricultural landscapes. GLMMs suggested that tree height, undergrowth coverage and height, and elevation determined butterfly diversity. Butterfly species richness was also influenced by season and landscape-level variables such as proximity to forest, mean NDVI, and habitat. Understanding the factors that contributed to butterfly species richness in an agroecosystem, stakeholders should consider management practices to improve biodiversity conservation such as ground vegetation management and retaining adjacent forest areas to enhance butterfly species richness. Furthermore, our findings suggest that agroforestry system should be considered to enhance biodiversity in agricultural landscapes.
Mutualistic and antagonistic plant-animal interactions differentially contribute to the maintenance of species diversity in ecological communities. Although both seed dispersal and predation by fruit-eating animals are recognized as important drivers of plant population dynamics, the mechanisms underlying how seed dispersers and predators jointly affect plant diversity remain largely unexplored. Based on mediating roles of seed size and species abundance, we investigated the effects of seed dispersal and predation by two sympatric primates (Nomascus concolor and Trachypithecus crepusculus) on local plant recruitment in a subtropical forest of China. Over a 26 month period, we confirmed that these primates were functionally distinct: gibbons were legitimate seed dispersers who dispersed seeds of 44 plant species, while langurs were primarily seed predators who destroyed seeds of 48 plant species. Gibbons dispersed medium-seeded species more effectively than small- and large-seeded species, and dispersed more seeds of rare species than common and dominant species. Langurs showed a similar predation rate across different sizes of seeds, but destroyed a large number of seeds from common species. Due to gut passage effects, gibbons significantly shortened the duration of seed germination for 58% of the dispersed species; however, for 54% of species, seed germination rates were reduced significantly. Our study underlined the contrasting contributions of two primate species to local plant recruitment processes. By dispersing rare species and destroying the seeds of common species, both primates might jointly maintain plant species diversity. To maintain healthy ecosystems, the conservation of mammals that play critical functional roles needs to receive further attention.
Tropical rainforest disturbance and conversion are critical drivers of biodiversity loss. A key knowledge gap is understanding the impacts of habitat modification on mechanisms of community assembly, which are predicted to respond differently between taxa and across spatial scales. We use a null model approach to detect trait assembly of species at local- and landscape-scales, and then subdivide communities with different habitat associations and foraging guilds to investigate whether the detection of assembly mechanisms varies between groups. We focus on two indicator taxa, dung beetles and birds, across a disturbance gradient of primary rainforest, selectively logged rainforest, and oil palm plantations in Borneo, Southeast Asia. Random community assembly was predominant for dung beetles across habitats, whereas trait convergence, indicative of environmental filtering, occurred across the disturbance gradient for birds. Assembly patterns at the two spatial scales were similar. Subdividing for habitat association and foraging guild revealed patterns hidden when focusing on the overall community. Dung beetle forest specialists and habitat generalists showed opposing assembly mechanisms in primary forest, community assembly of habitat generalists for both taxa differed with disturbance intensity, and insectivorous birds strongly influenced overall community assembly relative to other guilds. Our study reveals the sensitivity of community assembly mechanisms to anthropogenic disturbance via a shift in the relative contribution of stochastic and deterministic processes. This highlights the need for greater understanding of how habitat modification alters species interactions and the importance of incorporating species' traits within assessments.
Most canopy species in lowland tropical rain forests in Southeast Asia, represented by Dipterocarpaceae, undergo mast reproduction synchronously at community level during a general flowering event. Such events occur at irregular intervals of 2-10 years. Some species do not necessarily participate in every synchronous mast reproduction, however. This may be due to a lack of carbohydrate resources in the trees for masting. We tested the hypothesis that interspecific differences in the time required to store assimilates in trees for seed production are due to the frequency of masting and/or seed size in each species. We examined the relationship between reproductive frequency and the carbon accumulation period necessary for seed production, and between the seed size and the period, using radiocarbon analysis in 18 dipterocarp canopy species. The mean carbon accumulation period was 0.84 years before seed maturation in all species studied. The carbon accumulation period did not have any significant correlation with reproductive frequency or seed size, both of which varied widely across the species studied. Our results show that for seed production, dipterocarp masting species do not use carbon assimilates stored for a period between the masting years, but instead use recent photosynthates produced primarily in a masting year, regardless of the masting interval or seed size of each species. These findings suggest that storage of carbohydrate resources is not a limiting factor in the masting of dipterocarps, and that accumulation and allocation of other resources is important as a precondition for participation in general flowering.