The geographical variation of network structure is scale dependent: understanding the biotic specialization of host-parasitoid networks

Latitudinal and anthropogenic effects on the structuring of networks linking blood-feeding flies and their vertebrate hosts

Biting flies (Diptera) transmit pathogens that cause many important diseases in humans as well as domestic and wild animals. The networks of feeding interactions linking these insects to their hosts, and how they vary geographically and in response to human land-use, are currently poorly documented but are relevant to understanding cross-species disease transmission. We compiled a database of biting Diptera-host interactions from the literature to investigate how key interaction network metrics vary latitudinally and with human land-use. Interaction evenness and H2' (a measure of the degree of network specificity) did not vary significantly with latitude. Compared to near-natural habitats, interaction evenness was significantly lower in agricultural habitats, where networks were dominated by relatively few species pairs, but there was no evidence that the presence of humans and their domesticated animals within networks led to systematic shifts in network structure. We discuss the epidemiological relevance of these results and the implications for predicting and mitigating future spill-over events.

Moving towards a comprehensive view of the spatial processes in seed dispersal networks: Embracing metacommunities

The current biodiversity crisis requires efficient approaches to address the ongoing impoverishment of natural communities and the depletion of ecosystem services and functions. In this sense, identifying key species that promote the functioning of ecological processes can be strategic to guide actions aiming at the conservation and restoration of biodiversity. Node-level metrics in interaction networks can be helpful to identify those key species, as they measure the role each species plays in organizing the interactions. Moreover, ecological correlates of species structural roles may vary between local and global networks of interactions, reflecting distinct mechanisms acting at different spatial scales. By studying local seed dispersal networks and one global meta-network combining those local networks, Moulatlet et al. identified the most important traits explaining bird species centrality at varying spatial scales. They found that body mass was the main trait explaining centrality at the local scale, whereas range size was the main predictor of species centrality at the global scale. In this contribution, besides assessing local interaction networks, Moulatlet et al. adopt a biogeographical perspective to seed dispersal systems, extending our knowledge about the possible mechanisms that underlie the organization of interacting assemblages when changing the spatial scale of observation. Future efforts on this field could include an intermediate scale, comprising the level of metacommunities, shedding light on the interplay between local and spatial processes, both embedded in biogeographical realms, when determining the organization of interactions and the ecological correlates of species roles.

Latitudinal gradient in species diversity provides high niche opportunities for a range-expanding phytophagous insect

When species undergo poleward range expansions in response to anthropogenic change, they likely encounter less diverse communities in new locations. If low diversity communities provide weak biotic interactions, such as reduced competition or predation, range-expanding species may experience high niche opportunities. Here, we investigated if oak gall wasp communities follow a latitudinal diversity gradient (LDG) and if lower diversity communities provide weaker interactions at the poles for a range-expanding community member, Neuroterus saltatorius. We performed systematic surveys of gall wasps on a dominant oak, Quercus garryana, throughout most of its range, from northern California to Vancouver Island, British Columbia. On 540 trees at 18 sites, we identified 23 oak gall wasp morphotypes in three guilds (leaf detachable, leaf integral, and stem galls). We performed regressions between oak gall wasp diversity, latitude, and other abiotic (e.g. temperature) and habitat (e.g. oak patch size) factors to reveal if gall wasp communities followed an LDG. To uncover patterns in local interactions, we first performed partial correlations of gall wasp morphotype occurrences on trees within regions). We then performed regressions between abundances of co-occurring gall wasps on trees to reveal if interactions are putatively competitive or antagonistic. Q. garryana-gall wasp communities followed an LDG, with lower diversity at higher latitudes, particularly with a loss of detachable leaf gall morphotypes. Detachable leaf gall wasps, including the range-expanding species, co-occurred most on trees, with weak co-occurrences on trees in the northern expanded region. Abundances of N. saltatorius and detachable and integral leaf galls co-occurring on trees were negatively related, suggesting antagonistic interactions. Overall, we found that LDGs create communities with weaker associations at the poles that might facilitate ecological release in a range-expanding community member. Given the ubiquity of LDGs in nature, poleward range-expanding species are likely moving into low diversity communities. Yet, understanding if latitudinal diversity pattern provides weak biotic interactions for range-expanding species is not well explored. Our large-scale study documenting diversity in a related community of phytophagous insects that co-occur on a host plant reveals that LDGs create high niche opportunities for a range-expanding community member. Biogeographical patterns in diversity and species interactions are likely important mechanisms contributing to altered biotic interactions under range-expansions.

Functional biogeography of parasite traits: hypotheses and evidence

Functional biogeography, or the study of trait-based distributional patterns, not only complements our understanding of spatial patterns in biodiversity, but also sheds light on the underlying processes generating them. In parallel with the well-studied latitudinal diversity gradient, decades-old ecogeographical rules also postulate latitudinal variation in species traits. Notably, species in the tropics are predicted to have smaller body sizes (Bergmann's rule), narrower niches (MacArthur's rule) and smaller geographical ranges (Rapoport's rule) than their counterparts at higher latitudes. Although originally proposed for free-living organisms, these rules have been extended to parasitic organisms as well. In this review, I discuss the mechanistic hypotheses most likely to explain latitudinal gradients in parasite traits, and assess the empirical evidence obtained from comparative studies testing the above three rules as well as latitudinal gradients in other parasite traits. Overall, there is only weak empirical support for latitudinal gradients in any parasite trait, with little consistency among comparative analyses. The most parsimonious explanation for the existence of geographical patterns in parasite traits is that they are primarily host-driven, i.e. ecological traits of parasites track those of their hosts, with a direct influence of bioclimatic factors playing a secondary role. Thus, geographical patterns in parasite traits probably emerge as epiphenomena of parallel patterns in their hosts.

This article is part of the theme issue ‘Infectious disease macroecology: parasite diversity and dynamics across the globe’.

Modularity and specialization in bat-fly interaction networks are remarkably consistent across patches within urbanized landscapes and spatial scales

Patterns of specialization and the structure of interactions between bats and ectoparasitic flies have been studied mostly on non-urban environments and at local scales. Thus, how anthropogenic disturbances influence species interactions and network structure in this system remain poorly understood. Here, we investigated patterns of interaction between Phyllostomidae bats and ectoparasitic Streblidae flies, and variations in network specialization and structure across Cerrado patches within urbanized landscapes in Brazil and between local and regional scales. We found high similarity in the richness and composition of bat and fly species across communities, associated with low turnover of interactions between networks. The high specialization of bat–streblid interactions resulted in little connected and modular networks, with the emergence of modules containing subsets of species that interact exclusively or primarily with each other. Such similarities in species and interaction composition and network structure across communities and scales suggest that bat–fly interactions within Cerrado patches are little affected by the degree of human modification in the surrounding matrix. This remarkable consistency is likely promoted by specific behaviors, the tolerance of Phyllostomidae bats to surrounding urbanized landscapes as well as by the specificity of the streblid–bat interactions shaped over evolutionary time.

Assessing Biotic and Abiotic Interactions of Microorganisms in Amazonia through Co-Occurrence Networks and DNA Metabarcoding

Species may co-occur due to responses to similar environmental conditions, biological associations, or simply because of coincident geographical distributions. Disentangling patterns of co-occurrence and potential biotic and abiotic interactions is crucial to understand ecosystem function. Here, we used DNA metabarcoding data from litter and mineral soils collected from a longitudinal transect in Amazonia to explore patterns of co-occurrence. We compared data from different Amazonian habitat types, each with a characteristic biota and environmental conditions. These included non-flooded rainforests (terra-firme), forests seasonally flooded by fertile white waters (várzeas) or by unfertile black waters (igapós), and open areas associated with white sand soil (campinas). We ran co-occurrence network analyses based on null models and Spearman correlation for all samples and for each habitat separately. We found that one third of all operational taxonomic units (OTUs) were bacteria and two thirds were eukaryotes. The resulting networks were nevertheless mostly composed of bacteria, with fewer fungi, protists, and metazoans. Considering the functional traits of the OTUs, there is a combination of metabolism modes including respiration and fermentation for bacteria, and a high frequency of saprotrophic fungi (those that feed on dead organic matter), indicating a high turnover of organic material. The organic carbon and base saturation indices were important in the co-occurrences in Amazonian networks, whereas several other soil properties were important for the co-exclusion. Different habitats had similar network properties with some variation in terms of modularity, probably associated with flooding pulse. We show that Amazonian microorganism communities form highly interconnected co-occurrence and co-exclusion networks, which highlights the importance of complex biotic and abiotic interactions in explaining the outstanding biodiversity of the region.

Species diversity and food web structure jointly shape natural biological control in agricultural landscapes

Land-use change and agricultural intensification concurrently impact natural enemy (e.g., parasitoid) communities and their associated ecosystem services (ESs), i.e., biological pest control. However, the extent to which (on-farm) parasitoid diversity and food webs mediate landscape-level influences on biological control remains poorly understood. Here, drawing upon a 3-year study of quantitative parasitoid-hyperparasitoid trophic networks from 25 different agro-landscapes, we assess the cascading effects of landscape composition, species diversity and trophic network structure on ecosystem functionality (i.e., parasitism, hyperparasitism). Path analysis further reveals cascaded effects leading to biological control of a resident crop pest, i.e., Aphis gossypii. Functionality is dictated by (hyper)parasitoid diversity, with its effects modulated by food web generality and vulnerability. Non-crop habitat cover directly benefits biological control, whereas secondary crop cover indirectly lowers hyperparasitism. Our work underscores a need to simultaneously account for on-farm biodiversity and trophic interactions when investigating ESs within dynamic agro-landscapes.

Low host specificity and broad geographical ranges in a community of parasitic non-pollinating fig wasps (Sycoryctinae; Chalcidoidea)

Plants, phytophagous insects and their parasitoids form the most diverse assemblages of macroscopic organisms on earth. Enclosed assemblages in particular represent a tractable system for studying community assembly and diversification. Communities associated with widespread plant species are especially suitable as they facilitate a comparative approach. Pantropical fig-wasp communities represent a remarkably well-replicated system, ideal for studying these historical processes. We expect high dispersal ability in non-pollinating fig wasps to result in lower geographical turnover in comparison to pollinating fig wasps. The ability of non-pollinating wasps to utilise a number of hosts (low host specificity) is a key determinant of overall geographical range, with intraspecific competition becoming a constraining factor should diet breadth overlap among species. Finally, we expect conserved community structure throughout the host range. We aim to test these expectations, derived from population genetic and community studies, using the multi-trophic insect community associated with Ficus hirta throughout its 3,500 km range across continental and insular Asia. We collect molecular evidence from one coding mitochondrial gene, one non-coding nuclear gene and multiple microsatellites across 25 geographical sites. Using these data, we establish species boundaries, determine levels of host specificity among non-pollinating fig wasps and quantify geographical variation in community composition. We find low host specificity in two genera of non-pollinating fig wasps. Functional community structure is largely conserved across the range of the host fig, despite limited correspondence between the ranges of non-pollinator and pollinator species. While nine pollinators are associated with Ficus hirta, the two non-pollinator tribes developing in its figs each contained only four species. Contrary to predictions, we find stronger isolation by distance in non-pollinators than pollinators. Long-lived non-pollinators may disperse more gradually and be less reliant on infrequent long-distance dispersal by wind currents. Segregation among non-pollinating species across their range is suggestive of competitive exclusion and we propose that this may be a result of increased levels of local adaptation and moderate, but regular, rates of dispersal. Our findings provide one more example of lack of strict codiversification in the geographical diversification of plant-associated insect communities.

Species-area and network-area relationships in host-helminth interactions

The scaling relationship observed between species richness and the geographical area sampled (i.e. the species-area relationship (SAR)) is a widely recognized macroecological relationship. Recently, this theory has been extended to trophic interactions, suggesting that geographical area may influence the structure of species interaction networks (i.e. network-area relationships (NARs)). Here, we use a global dataset of host-helminth parasite interactions to test existing predictions from macroecological theory. Scaling between single locations to the global host-helminth network by sequentially adding networks together, we find support that geographical area influences species richness and the number of species interactions in host-helminth networks. However, species-area slopes were larger for host species relative to their helminth parasites, counter to theoretical predictions. Lastly, host-helminth network modularity-capturing the tendency of the network to form into separate subcommunities-decreased with increasing area, also counter to theoretical predictions. Reconciling this disconnect between existing theory and observed SAR and NAR will provide insight into the spatial structuring of ecological networks, and help to refine theory to highlight the effects of network type, species distributional overlap, and the specificity of trophic interactions on NARs.

Geographical variation of multiplex ecological networks in marine intertidal communities

Understanding the drivers of geographical variation in species distributions, and the resulting community structure, constitutes one of the grandest challenges in ecology. Geographical patterns of species richness and composition have been relatively well studied. Less is known about how the entire set of trophic and non-trophic ecological interactions, and the complex networks that they create by gluing species together in complex communities, change across geographical extents. Here, we compiled data of species composition and three types of ecological interactions occurring between species in rocky intertidal communities across a large spatial extent (~970 km of shoreline) of central Chile, and analyzed the geographical variability in these multiplex networks (i.e., comprising several interaction types) of ecological interactions. We calculated nine network summary statistics common across interaction types, and additional network attributes specific to each of the different types of interactions. We then investigated potential environmental drivers of this multivariate network organization. These included variation in sea surface temperature and coastal upwelling, the main drivers of productivity in nearshore waters. Our results suggest that structural properties of multiplex ecological networks are affected by local species richness and modulated by factors influencing productivity and environmental predictability. Our results show that non-trophic negative interactions are more sensitive to spatially structured temporal environmental variation than feeding relationships, with non-trophic positive interactions being the least labile to it. We also show that environmental effects are partly mediated through changes in species richness and partly through direct influences on species interactions, probably associated to changes in environmental predictability and to bottom-up nutrient availability. Our findings highlight the need for a comprehensive picture of ecological interactions and their geographical variability if we are to predict potential effects of environmental changes on ecological communities.