Land-use changes impact root–fungal network connectivity in a global biodiversity hotspot

Abstract

1 INTRODUCTION

Tropical rainforests are highly diverse ecosystems that contain a wide array of micro-habitats and organisms (Eiserhardt et al., 2017; Wardle et al., 2004). Yet, rainforests face major threats due to their rapid replacement with cash crop plantations such as oil palm and rubber (Rembold et al., 2017; Zemp et al., 2023). For example, the ongoing agricultural expansion has resulted in ~21 million hectares of oil palm plantations globally (Descals et al., 2021; Tedersoo et al., 2014). Overall, the taxonomic diversity of native species has decreased due to land-use changes (Ballauff et al., 2021; Barnes et al., 2017; Brinkmann et al., 2019; Felipe-Lucia et al., 2020; Newbold et al., 2016). However, to better understand the multiple impacts of agricultural expansion on tropical ecosystems, we urgently need to strengthen our knowledge of the impacts of land-use conversion on species interactions (Brinkmann et al., 2019; Felipe-Lucia et al., 2020; Newbold et al., 2016; Romdhane et al., 2022).

Plant–fungal associations play a fundamental role in shaping the structure and functioning of tropical ecosystems (Põlme et al., 2018; Tedersoo et al., 2014, 2022), providing insights into the resilience, adaptive capacity and health of forest ecosystems (Trivedi et al., 2020). Root–microbial interactions involving different kingdoms and functional guilds (Brunel et al., 2020; Ferlian et al., 2018; McLaren & Callahan, 2020; Trivedi et al., 2020; Vorburger & Perlman, 2018) have been linked to mutualistic preferences, evolutionary and trait differences among hosts, and phylogenetic relatedness and competitive exclusion between fungi (Alzarhani et al., 2019; Francioli et al., 2021). For example, the abundance and diversity of arbuscular mycorrhizal fungi (AMF) have been found to respond to changes in plant diversity and composition (Deyn et al., 2011; Gui et al., 2017), with host–AMF preferences likely modulated by plant functional groups, instead of reflecting individual species interactions (Edy et al., 2022; Francioli et al., 2021; Zanne et al., 2020). Moreover, an increase in the abundance of AMF in the rhizosphere is expected to promote a decrease in pathogen abundance (Francioli et al., 2021; Sweeney et al., 2021; Zanne et al., 2020) and to be positively correlated with the abundance of saprotrophic fungi (Francioli et al., 2021; van der Heijden & Hartmann, 2016). Although in tropical lowland forests, most tree species are associated with AMF, plant species in the Dipterocarpaceae and Fagaceae families are known to harbour ectomycorrhizal fungi (EMF; Bahram et al., 2014; Matchado et al., 2021). Nevertheless, our comprehension of root-fungal associations in lowland tropical rainforests stems predominantly from studies concentrating on root-AMF (Gui et al., 2017; Koorem et al., 2017), with EMF studies mainly from temperate ecosystems (Encinas-Viso et al., 2016; Hogan et al., 2023; Kuang et al., 2021; Van Geel et al., 2018).

Large-scale conversion of tropical rainforests to monocultures reduces microhabitat diversity, leading to species turnover (Ballauff et al., 2021; Barnes et al., 2017; Brinkmann et al., 2019) and loss of plant and fungal diversity (Barnes et al., 2017). For example, when analysing species-specific root–AMF associations in individual root samples, there was a significant decrease in AMF richness in roots of oil palm and rubber compared with roots from plants sampled in rainforests (Edy et al., 2022). Environmental filters associated with forest conversion may select tolerant and opportunistic species (plants and fungal) that can occupy large niche ranges and out-compete native species in converted landscapes (Gioria et al., 2023; Hulme, 2017; McLaren & Callahan, 2020; Nayanakantha, 2007; Vilà & Weiner, 2004; Walther et al., 2009), resulting in changes in community composition. Therefore, insights toward comprehending community-level interactions in natural and human-modified ecosystems are highly needed (Chen et al., 2021; Delavaux et al., 2017; Sepp et al., 2019), yet studies in tropical diversity hotspots on below-ground associations are rare.

By combining amplicon sequencing on DNA from identical samples for plant and fungal communities' assessment with network analysis, we can provide valuable insights into the mechanisms driving the coexistence of species from various kingdoms within biological communities, revealing trophic and non-trophic interactions under specific environmental conditions (Grass et al., 2020; Montoya-Sánchez et al., 2023; Tiede et al., 2016; Zemp et al., 2023). Network analysis plays a crucial role in identifying highly connected nodes, facilitating the identification of key species and interactions essential for shaping community structure (Guimerà & Amaral, 2005; Põlme et al., 2018; van der Heijden & Hartmann, 2016). An effective approach for identifying land-use-specific species (i.e. indicator species) is the use of the indicator value index (IndVal), which evaluates the distribution patterns that best match the taxa in question (De Cáceres et al., 2010), determining whether a species is specific to a particular land use or associated with multiple land-use types.

Here, by using an unprecedented dataset for the tropics containing roots and fungal communities from four land-use types in Sumatra, Indonesia, we aim to understand the impact of rainforest transformation on the composition and structure of root–fungal networks in a global tropical biodiversity hotspot. To do that, we first identified patterns in root–fungal communities and estimated indicator species associated with each land-use type. Second, we estimated root-fungal associations through co-occurrence network analysis. Third, we examined highly connected taxa that shape the composition and structure of ecological networks. We hypothesized that forest transformation is a main driver of root–fungal associations via changes in root diversity, composition and traits (Ballauff et al., 2021). Plantations, for instance, have 59% fewer plant species than rainforests, whereas the latter consist predominantly of native species, including forest specialists, versus higher non-native species richness and abundance in other land-use types (Rembold et al., 2017). Thereby, we expected rainforest plots to be associated with roots from rare and endemic species. In contrast, roots from non-native and generalist species, along with fungal groups adapted to withstand or thrive in highly disturbed niches, may serve as indicators of land-use intensity due to their resilience to adverse conditions. Finally, across land-use types, we hypothesized that taxa associated with rainforests have greater ecological connectivity than those in plantations, thereby contributing to a more structured ecosystem.