Introduction
Conversion of tropical rainforest into cash crops is escalating due to the pressure to increase food and biofuel production (Laurance et al. 2014). Indonesia has experienced particularly high rates of deforestation and forest degradation within the last decades (Hansenet al. 2013; Margono et al. 2014). Conversion of rainforest into rubber and oil palm plantations, which are among the major cash-crops in Indonesia (Juniyanti et al. 2021), is not only accompanied by vast deforestation and above- and belowground carbon losses, but also by substantial inputs of fertilizers, herbicides and pesticides (Guillaume et al. 2018; Darras et al. 2019). Together, this has resulted in decreases in floral and faunal abundance and diversity in multiple taxonomic groups (Fitzherbert et al.2008; Gibson et al. 2011; Chaudhary et al. 2016). However, functional consequences of biodiversity losses remain little understood. Biodiversity across trophic levels is of pivotal importance for maintaining ecosystem multifunctionality (Hunt & Wall 2002; Bardgett & van der Putten 2014; Tilman et al. 2014). Decreases in abundance and diversity of invertebrates may critically impact energy fluxes in food webs and ultimately lead to losses of ecosystem functions (Cardinale et al. 2006; Barnes et al. 2014, 2016; Cloughet al. 2016; Drescher et al. 2016). However, until today we do not know how energy fluxes in tropical canopies, which are one of the most biodiverse habitats on Earth, are affected by land-use change (Erwin 1983; Ellwood & Foster 2004; Nakamura et al. 2017).
Studies on canopy arthropod groups such as ants, beetles, spiders and parasitoid wasps suggest substantial decline in species diversity, density and biomass if rainforest is converted into rubber and oil palm plantations (Nazarreta et al. 2020; Azhar et al. 2022; Kasmiatun et al. 2022; Ramos et al. 2022). Evaluating associated changes in energy fluxes may help to understand functional consequences of this decline (Barnes et al. 2014; Zhou et al. 2022). Ecosystem functioning crucially depends on fluxes of energy and matter, which are closely linked to food web dynamics via trophic interactions (Reiss et al. 2009). Manning et al. (2018) argue that ecosystem functioning should ideally be defined via process rates involving fluxes of energy and matter between trophic levels and the environment. This calls for a holistic food web approach, focussing not only on individual trophic groups, but on energy fluxes through multiple trophic levels and entire communities (Soliveres et al.2016; Barnes et al. 2018; Wan et al. 2022). Energy fluxes through food web nodes, characterizing the energy consumption of different trophic groups, can be used to infer ecosystem functions such as predation, herbivory and detritivory, that are closely related to ecosystem services, such as pest control, decomposition and carbon sequestration (Barnes et al. 2018; Gauzens et al. 2019).
Application of the energy flux approach to invertebrate communities in tropical canopies is challenging, because the trophic structure of arthropod communities in this habitat is poorly explored (Nakamuraet al. 2017). Given the central role of food web topology for the correct estimation of energy fluxes (Jochum et al. 2021), it is crucial to obtain precise information on trophic positions and trophic interactions of multiple trophic groups in the food web (O’Neill 1969; Seibold et al. 2018). Thus, in-situ assessment of trophic positions of taxa and their potential shifts among tropical land-use systems is needed, but rarely done (Zhou et al. 2022). To the best of our knowledge, the trophic structure of arthropod communities in tropical canopies and land-use associated changes in energy fluxes therein have not been investigated in a comprehensive way until today.
Here, using canopy fogging, we collected arthropods in rainforest, and in three major transformation systems in Southeast Asia, i.e. jungle rubber (rubber agroforest system), rubber and oil palm plantations, replicated in two landscapes in Jambi Province, Sumatra. Sampling was conducted once in the dry and once in the rainy season. For all major arthropod taxa (12 orders and 5 families, representing consistent trophic groups) on each site, we recorded abundance and biomass, and used stable isotope ratios (δ15N and δ13C values) to estimate site-specific trophic positions from primary consumers to top predators and to assess the use of basal resources such as plants and epiphytic microorganisms. Stable isotope analysis allowed us to reconstruct realistic trophic positions of major taxa for calculating energy fluxes and to assess ecological functions such as herbivory, algae-microbivory and predation. We compared fluxes and functions among rainforest transformation systems and hypothesized that (1) due to decline in resource diversity, and density and biomass of canopy arthropods under anthropogenic disturbance, land use simplifies the trophic structure and decreases trophic redundancy; and (2) land use results in a strong reduction of the total energy flux in canopy food webs and leads to changes in proportions of key arthropod-driven trophic functions, such as herbivory, predation and microbivory.