Effects of Local- and Landscape-level Drivers in Influencing Bird Diversity and Persistence in Cashew Plantations of the Northern Western Ghats

Abstract

Agricultural expansion is a significant driver of deforestation and biodiversity loss, particularly in the tropics. The Western Ghats-Sri Lanka biodiversity hotspot is no exception. Most forest patches in the northern Western Ghats are being cleared for cashew cultivation. However, here, cashew plantations host varying proportions of cashew and native trees and have varying extents of ambient forest cover. Investigating the effects of such local and landscape-level drivers across land use types on taxonomic, functional, and phylogenetic diversities, as well as species- and trait-specific responses of birds, as implemented in this study, can give comprehensive insights on the impacts of forest conversion. I conducted this study in the low-elevation forests and cashew plantation landscape of Tillari, Maharashtra. I compared tree species diversity and vegetation structural attributes across land use types using Hill-Shannon diversity and linear models (LMs). To assess how land use type affects bird species composition and taxonomic diversity, I used non-metric multidimensional sampling (NMDS) and Hill-Shannon diversity. To compare functional and phylogenetic impacts across land use types, I used SES (Standardised Effect Size) values of fMPD (functional Mean Pairwise Distance) and pMPD (phylogenetic Mean Pairwise Distance) for LMs and Tukey’s post-hoc analysis. Using Hierarchical Modelling of Species Communities (HMSC), I looked for 1) bird species responses, 2) trait responses, and 3) a phylogenetic signal in the residuals. I collected bird and vegetation data from 100 points distributed across forest, mixed cashew plantations, and pure cashew plantations. My field collaborator (Vishal Sadekar) and I sampled vegetation within a 10 m radius of circular plots around all points. We sampled birds, implementing 10-minute point counts at a point, replicated 5 times across 4 months between January to May 2024. Vegetation differed significantly across all three land use types in terms of tree species diversity and tree height, where the forest had the highest estimates followed by mixed cashew. NMDS showed that the community in mixed cashew plantations overlapped with that of pure cashew plantations. There was no significant effect of land use type on functional diversity. But taxonomic and phylogenetic diversities were lowest in pure cashew. The HMSC model revealed a low explained variation (Tjur R2 = 0.1372), but I found land use type, a site-level variable, to exhibit the highest explained variation of 58% compared to the landscape-level variables. Forest species like Malabar Trogon (Harpactes fasciatus) and Black-naped Monarch (Hypothymis azurea) responded negatively to mixed and pure cashew plantations. Species like Golden-fronted Leafbird (Chloropsis auriforns) and Tickell’s Blue Flycatcher (Cyornis tickelliae) benefitted from mixed cashew and open country species like Indian Peafowl (Pavo cristatus) and Plain Prinia (Prinia inornata) responded positively to pure cashew. Trait response analysis revealed that evergreen species responded positively to the forest and negatively to both cashew habitats. Sedentary species responded positively to forest cover in 800 m. No evidence of a phylogenetic signal was found in the residuals (⍴ = 0.23, 95% CI: 0 - 0.68). This study highlights the value of mixed cashew plantations which are more habitable than pure cashew plantations due to native tree prevalence. However, this study also emphasises that even mixed cashew plantations cannot replace forests as habitats for evergreen forest species. This message is now more crucial than ever in the light of ongoing forest clearance in the northern Western Ghats.