Modeling species co-occurrence by multivariate logistic regression generates new hypotheses on fungal interactions

Combining observational and experimental data to estimate environmental and species drivers of fungal metacommunity dynamics

Understanding the distribution and dynamics of species is central to ecology and important for managing biodiversity. The distributions of species in metacommunities are determined by many factors, including environmental conditions and interactions between species. Yet, it is difficult to quantify the effect of species interactions on metacommunity dynamics from observational data. We present an approach to estimate the importance of species interactions that combines data from two observational presence-absence inventories (providing colonization-extinction data) with data from species interaction experiments (providing informative prior distributions in the Bayesian framework). We further illustrate the approach on wood-decay fungi that interact within a downed log through competition for resources and space, and facilitate the succession of other species by decomposing the wood. Specifically, we estimated the relative importance of species interactions by examining how the presence of a species influenced the colonization and extinction probability of other species. Temporal data on fruit body occurrence of 12 species inventoried twice were jointly analyzed with experimental data from two laboratory experiments that aimed to estimate competitive interactions. Both environmental variables and species interactions affected colonization and extinction dynamics. Late-successional fungi had more colonization interactions with predecessor species than early-successional species. We identified several species interactions, and the presence of certain species changed the probability that later-successional species colonized by -81% to 512%. The presence of certain species increased the probability that other species went extinct from a log by 14%-61%. Including the informative priors from experimental data added two colonization interactions and one extinction interaction for which the observational field data was inconclusive. However, most species had no detectable interactions, either because they did not interact or because of low species occupancy, meaning data limitation. We show how temporal presence-absence data can be combined with experimental data to identify which species influence the colonization-extinction dynamics of others. Accounting for species interactions in metacommunity models, in addition to environmental drivers, is important because interactions can have cascading effects on other species.

Idiosyncratic responses to biotic and environmental filters in wood-inhabiting fungal communities

Quantification of different processes affecting the assembly of ecological communities remains challenging, especially in species-rich communities. While the role of environmental filtering has generally been well established, fewer studies have experimentally shown how other ecological assembly processes, such as biotic filtering, structure species-rich communities. Here, we studied the relative roles of biotic and environmental filtering in the colonization of wood-inhabiting fungi, a species-rich, highly interactive, and environment-sensitive group of species. We conducted a field experiment where we simulated colonization with inoculations of nine fungal species in habitat patches (i.e., logs) with varying biotic and abiotic conditions. We characterized the local resident communities before the inoculations and the colonization success of the inoculated species after one and two years using DNA metabarcoding. We asked what determined the colonization success of the inoculated species by comparing the predictive performance of alternative models. These models included either only abiotic environmental predictors (i.e., physical log properties) or additionally different aspects of the resident fungal communities (i.e., resident fungal species richness, community composition, and DNA amount) as biotic predictors. While all nine species successfully colonized the logs, the rate of success and the factors explaining their colonization success varied among species. The colonization success of four of the inoculated species was explained mostly by the abiotic environmental variables, while the colonization success of three species was additionally explained by the resident communities. The influential biotic predictors varied from the presence of individual species to the collective presence of multiple species. Finally, for two of the inoculated species, all the models showed poor predictive performance. Our results indicate how environmental and biotic filtering may jointly structure species-rich communities. Overall, the results show that species vary idiosyncratically in their response to biotic and environmental factors, highlighting the need to consider the complexity of species-level responses when predicting community-level changes.

Estimating genus-specific effects of non-native honey bees and urbanization on wild bee communities: A case study in Maryland, United States

Non-native species have the potential to detrimentally affect native species through resource competition, disease transmission, and other forms of antagonism. The western honey bee (Apis mellifera) is one such species that has been widely introduced beyond its native range for hundreds of years. There are strong concerns in the United States, and other countries, about the strain that high-density, managed honey bee populations could pose to already imperiled wild bee communities. While there is some experimental evidence of honey bees competing with wild bees for resources, few studies have connected landscape-scale honey bee apiary density with down-stream consequences for wild bee communities. Here, using a dataset from Maryland, US and joint species distribution models, we provide the largest scale, most phylogenetically resolved assessment of non-native honey bee density effects on wild bee abundance to date. As beekeeping in Maryland primarily consists of urban beekeeping, we also assessed the relative impact of developed land on wild bee communities. Six of the 33 wild bee genera we assessed showed a high probability (> 90 %) of a negative association with apiary density and/or developed land. These bees were primarily late-season, specialist genera (several long-horned genera represented) or small, ground nesting, season-long foragers (including several sweat bee genera). Conversely, developed land was associated with an increase in relative abundance for some genera including invasive Anthidium and other urban garden-associated genera. We discuss several avenues to ameliorate potentially detrimental effects of beekeeping and urbanization on the most imperiled wild bee groups. We additionally offer methodological insights based on sampling efficiency of different methods (hand netting, pan trapping, vane trapping), highlighting large variation in effect sizes across genera. The magnitude of sampling effect was very high, relative to the observed ecological effects, demonstrating the importance of integrated sampling, particularly for multi-species or community level assessments.

Predicting responses to climate change using a joint species, spatially dependent physiologically guided abundance model

Predicting the effects of warming temperatures on the abundance and distribution of organisms under future climate scenarios often requires extrapolating species-environment correlations to climatic conditions not currently experienced by a species, which can result in unrealistic predictions. For poikilotherms, incorporating species' thermal physiology to inform extrapolations under novel thermal conditions can result in more realistic predictions. Furthermore, models that incorporate species and spatial dependencies may improve predictions by capturing correlations present in ecological data that are not accounted for by predictor variables. Here, we present a joint species, spatially dependent physiologically guided abundance (jsPGA) model for predicting multispecies responses to climate warming. The jsPGA model uses a basis function approach to capture both species and spatial dependencies. We apply the jsPGA model to predict the response of eight fish species to projected climate warming in thousands of lakes in Minnesota, USA. By the end of the century, the cold-adapted species was predicted to have high probabilities of extirpation across its current range-with 10% of lakes currently inhabited by this species having an extirpation probability >0.90. The remaining species had varying levels of predicted changes in abundance, reflecting differences in their thermal physiology. Though the model did not identify many strong species dependencies, the variation in estimated spatial dependence across species suggested that accounting for both dependencies was important for predicting the abundance of these fishes. The jsPGA model provides a new tool for predicting changes in the abundance, distribution, and extirpation probability of poikilotherms under novel thermal conditions.

Implications of the scale of detection for inferring co-occurrence patterns from paired camera traps and acoustic recorders

Multifunctional landscapes that support economic activities and conservation of biological diversity (e.g., cattle ranches with native forest) are becoming increasingly important because small remnants of native forest may comprise the only habitat left for some wildlife species. Understanding the co-occurrence between wildlife and disturbance factors, such as poaching activity and domesticated ungulates, is key to successful management of multifunctional landscapes. Tools to measure co-occurrence between wildlife and disturbance factors include camera traps and autonomous acoustic recording units. We paired 52 camera-trap stations with acoustic recorders to investigate the association between 2 measures of disturbance (poaching and cattle) and wild ungulates present in multifunctional landscapes of the Colombian Orinoquía. We used joint species distribution models to investigate species-habitat associations and species-disturbance correlations. One model was fitted using camera-trap data to detect wild ungulates and disturbance factors, and a second model was fitted after replacing camera-trap detections of disturbance factors with their corresponding acoustic detections. The direction, significance, and precision of the effect of covariates depended on the sampling method used for disturbance factors. Acoustic monitoring typically resulted in more precise estimates of the effects of covariates and of species-disturbance correlations. Association patterns between wildlife and disturbance factors were found only when disturbance was detected by acoustic recorders. Camera traps allowed us to detect nonvocalizing species, whereas audio recording devices increased detection of disturbance factors leading to more precise estimates of co-occurrence patterns. The collared peccary (Pecari tajacu), lowland tapir (Tapirus terrestris), and white-tailed deer (Odocoileus virginianus) co-occurred with disturbance factors and are conservation priorities due to the greater risk of poaching or disease transmission from cattle.

Incorporating mesopelagic fish into the evaluation of conservation areas for marine living resources under climate change scenarios

Mesopelagic fish (meso-fish) are central species within the Southern Ocean (SO). However, their ecosystem role and adaptive capacity to climate change are rarely integrated into protected areas assessments. This is a pity given their importance as crucial prey and predators in food webs, coupled with the impacts of climate change. Here, we estimate the habitat distribution of nine meso-fish using an ensemble model approach (MAXENT, random forest, and boosted regression tree). Four climate model simulations were used to project their distribution under two representative concentration pathways (RCP4.5 and RCP8.5) for short-term (2006-2055) and long-term (2050-2099) periods. In addition, we assess the ecological representativeness of protected areas under climate change scenarios using meso-fish as indicator species. Our models show that all species shift poleward in the future. Lanternfishes (family Myctophidae) are predicted to migrate poleward more than other families (Paralepididae, Nototheniidae, Bathylagidae, and Gonostomatidae). In comparison, lanternfishes were projected to increase habitat area in the eastern SO but lose area in the western SO; the opposite was projected for species in other families. Important areas (IAs) of meso-fish are mainly distributed near the Antarctic Peninsula and East Antarctica. Negotiated protected area cover 23% of IAs at present and 38% of IAs in the future (RCP8.5, long-term future). Many IAs of meso-fish still need to be included in protected areas, such as the Prydz Bay and the seas around the Antarctic Peninsula. Our results provide a framework for evaluating protected areas incorporating climate change adaptation strategies for protected areas management.

Supplementary Information

The online version contains supplementary material available at 10.1007/s42995-023-00188-9.

Structural and functional microbial diversity in deadwood respond to decomposition dynamics

We investigated the changes in microbial community diversities and functions in natural downed wood at different decay stages in a natural oak forest in the Italian Alps, through metagenomics analysis and in vitro analysis. Alfa diversity of bacterial communities was affected by the decay stage and log characteristics, while beta diversity was mainly driven by log diameter. Fungal and archaeal beta diversities were affected by the size of the sampled wood (log diameter), although, fungi were prominently driven by wood decay stage. The analysis of genes targeting cell wall degradation revealed higher abundances of cellulose and pectin-degrading enzymes in bacteria, while in fungi the enzymes targeting cellulose and hemicellulose were more abundant. The decay class affected the abundance of single enzymes, revealing a shift in complex hydrocarbons degradation pathways along the decay process. Moreover, we found that the genes related to Coenzyme M biosynthesis to be the most abundant especially at early stages of wood decomposition while the overall methanogenesis did not seem to be influenced by the decay stage. Intra- and inter-kingdom interactions between bacteria and fungi revealed complex pattern of community structure in response to decay stage possibly reflecting both direct and indirect interactions.

Joint species distribution models with imperfect detection for high-dimensional spatial data

Determining the spatial distributions of species and communities is a key task in ecology and conservation efforts. Joint species distribution models are a fundamental tool in community ecology that use multi-species detection-nondetection data to estimate species distributions and biodiversity metrics. The analysis of such data is complicated by residual correlations between species, imperfect detection, and spatial autocorrelation. While many methods exist to accommodate each of these complexities, there are few examples in the literature that address and explore all three complexities simultaneously. Here we developed a spatial factor multi-species occupancy model to explicitly account for species correlations, imperfect detection, and spatial autocorrelation. The proposed model uses a spatial factor dimension reduction approach and Nearest Neighbor Gaussian Processes to ensure computational efficiency for data sets with both a large number of species (e.g., >100) and spatial locations (e.g., 100,000). We compared the proposed model performance to five alternative models, each addressing a subset of the three complexities. We implemented the proposed and alternative models in the spOccupancy software, designed to facilitate application via an accessible, well documented, and open-source R package. Using simulations, we found that ignoring the three complexities when present leads to inferior model predictive performance, and the impacts of failing to account for one or more complexities will depend on the objectives of a given study. Using a case study on 98 bird species across the continental US, the spatial factor multi-species occupancy model had the highest predictive performance among the alternative models. Our proposed framework, together with its implementation in spOccupancy, serves as a user-friendly tool to understand spatial variation in species distributions and biodiversity while addressing common complexities in multi-species detection-nondetection data.

Field experimental evidence of sandy beach community changes in response to artificial light at night (ALAN)

Artificial light at night (ALAN) is a pervasive but still under-recognized driver of global change. In coastal settings, a large majority of the studies assessing ALAN impacts has focused on individual species, even though it is unclear whether results gathered from single species can be used to predict community-wide responses. Similarly, these studies often treat species as single life-stage entities, ignoring the variation associated with distinct life stages. This study addresses both limitations by focusing on the effects of ALAN on a sandy beach community consisting of species with distinct early- and late-life stages. Our hypothesis was that ALAN alters community structure and these changes are mediated by individual species and also by their ontogenetic stages. A field experiment was conducted in a sandy beach of north-central Chile using an artificial LED system. Samples were collected at different night hours (8-levels in total) across the intertidal (9-levels) over several days in November and January (austral spring and summer seasons). The abundance of adults of all species was significantly lower in ALAN treatments. Early stages of isopods showed the same pattern, but the opposite was observed for the early stages of the other two species. Clear differences were detected in the zonation of these species during natural darkness versus those exposed to ALAN, with some adult-juvenile differences in this response. These results support our hypothesis and document a series of changes affecting differentially both early and late life stages of these species, and ultimately, the structure of the entire community. Although the effects described correspond to short-term responses, more persistent effects are likely to occur if ALAN sources become established as permanent features in sandy beaches. The worldwide growth of ALAN suggests that the scope of its effect will continue to grow and represents a concern for sandy beach systems.

Trends in bird abundance differ among protected forests but not bird guilds

Improved monitoring and associated inferential tools to efficiently identify declining bird populations, particularly of rare or sparsely distributed species, is key to informed conservation and management across large spatiotemporal regions. We assess abundance trends for 106 bird species in a network of eight forested national parks located within the northeast United States from 2006 to 2019 using a novel hierarchical model. We develop a multispecies, multiregion, removal-sampling model that shares information across species and parks to enable inference on rare species and sparsely sampled parks and to evaluate the effects of local forest structure. Trends in bird abundance over time varied widely across parks, but species showed similar trends within parks. Three parks (Acadia National Park and Marsh-Billings-Rockefeller and Morristown National Historical Parks [NHP]) decreased in bird abundance across all species, while three parks (Saratoga NHP and Roosevelt-Vanderbilt and Weir-Farm National Historic Sites) increased in abundance. Bird abundance peaked at medium levels of basal area and high levels of percent forest and forest regeneration, with percent forest having the largest effect. Variation in these effects across parks could be a result of differences in forest structural stage and diversity. By sharing information across both communities and parks, our novel hierarchical model enables uncertainty-quantified estimates of abundance across multiple geographical (i.e., network, park) and taxonomic (i.e., community, guild, species) levels over a large spatiotemporal region. We found large variation in abundance trends across parks but not across bird guilds, suggesting that local forest condition might have a broad and consistent effect on the entire bird community within a given park. Research should target the three parks with overall decreasing trends in bird abundance to further identify what specific factors are driving observed declines across the bird community. Understanding how bird communities respond to local forest structure and other stressors (e.g., pest outbreaks, climate change) is crucial for informed and lasting management.

Fungal sporocarps house diverse and host-specific communities of fungicolous fungi

Sporocarps (fruit bodies) are the sexual reproductive stage in the life cycle of many fungi. They are highly nutritious and consequently vulnerable to grazing by birds and small mammals, and invertebrates, and can be infected by microbial and fungal parasites and pathogens. The complexity of communities thriving inside sporocarps is largely unknown. In this study, we revealed the diversity, taxonomic composition and host preference of fungicolous fungi (i.e., fungi that feed on other fungi) in sporocarps. We carried out DNA metabarcoding of the ITS2 region from 176 sporocarps of 11 wood-decay fungal host species, all collected within a forest in northeast Finland. We assessed the influence of sporocarp traits, such as lifespan, morphology and size, on the fungicolous fungal community. The level of colonisation by fungicolous fungi, measured as the proportion of non-host ITS2 reads, varied between 2.8-39.8% across the 11 host species and was largely dominated by Ascomycota. Host species was the major determinant of the community composition and diversity of fungicolous fungi, suggesting that host adaptation is important for many fungicolous fungi. Furthermore, the alpha diversity was consistently higher in short-lived and resupinate sporocarps compared to long-lived and pileate ones, perhaps due to a more hostile environment for fungal growth in the latter too. The fungicolous fungi represented numerous lineages in the fungal tree of life, among which a significant portion was poorly represented with reference sequences in databases.

Mycobiomes of Young Beech Trees Are Distinguished by Organ Rather Than by Habitat, and Community Analyses Suggest Competitive Interactions Among Twig Fungi

Beech trees (Fagus sylvatica) are prominent keystone species of great economic and environmental value for central Europe, hosting a diverse mycobiome. The composition of endophyte communities may depend on tree health, plant organ or tissue, and growth habitat. To evaluate mycobiome communalities at local scales, buds, and twigs were sampled from two young healthy mountain beech stands in Bavaria, Germany, four kilometers apart. With Illumina high-throughput sequencing, we found 113 fungal taxa from 0.7 million high-quality reads that mainly consisted of Ascomycota (52%) and Basidiomycota (26%) taxa. Significant correlations between richness and diversity indices were observed (p < 0.05), and mycobiomes did not differ between habitats in the current study. Species richness and diversity were higher in twigs compared to spring buds, and the assemblages in twigs shared most similarities. Interaction network analyses revealed that twig-bound fungi shared similar numbers of (interaction) links with others, dominated by negative co-occurrences, suggesting that competitive exclusion may be the predominant ecological interaction in the highly connected twig mycobiome. Combining community and network analyses strengthened the evidence that plant organs may filter endophytic communities directly through colonization access and indirectly by facilitating competitive interactions between the fungi.

Enhanced Lignocellulolytic Enzyme Activities on Hardwood and Softwood during Interspecific Interactions of White- and Brown-Rot Fungi

Wood decomposition is a sophisticated process where various biocatalysts act simultaneously and synergistically on biopolymers to efficiently break down plant cell walls. In nature, this process depends on the activities of the wood-inhabiting fungal communities that co-exist and interact during wood decay. Wood-decaying fungal species have traditionally been classified as white-rot and brown-rot fungi, which differ in their decay mechanism and enzyme repertoire. To mimic the species interaction during wood decomposition, we have cultivated the white-rot fungus, Bjerkandera adusta, and two brown-rot fungi, Gloeophyllum sepiarium and Antrodia sinuosa, in single and co-cultivations on softwood and hardwood. We compared their extracellular hydrolytic carbohydrate-active and oxidative lignin-degrading enzyme activities and production profiles. The interaction of white-rot and brown-rot species showed enhanced (hemi)cellulase activities on birch and spruce-supplemented cultivations. Based on the enzyme activity profiles, the combination of B. adusta and G. sepiarium facilitated birch wood degradation, whereas B. adusta and A. sinuosa is a promising combination for efficient degradation of spruce wood, showing synergy in β-glucosidase (BGL) and α-galactosidase (AGL) activity. Synergistic BGL and AGL activity was also detected on birch during the interaction of brown-rot species. Our findings indicate that fungal interaction on different woody substrates have an impact on both simultaneous and sequential biocatalytic activities.

The Poisson-Lognormal Model as a Versatile Framework for the Joint Analysis of Species Abundances

Joint Species Distribution Models (JSDM) provide a general multivariate framework to study the joint abundances of all species from a community. JSDM account for both structuring factors (environmental characteristics or gradients, such as habitat type or nutrient availability) and potential interactions between the species (competition, mutualism, parasitism, etc.), which is instrumental in disentangling meaningful ecological interactions from mere statistical associations. Modeling the dependency between the species is challenging because of the count-valued nature of abundance data and most JSDM rely on Gaussian latent layer to encode the dependencies between species in a covariance matrix. The multivariate Poisson-lognormal (PLN) model is one such model, which can be viewed as a multivariate mixed Poisson regression model. Inferring such models raises both statistical and computational issues, many of which were solved in recent contributions using variational techniques and convex optimization tools. The PLN model turns out to be a versatile framework, within which a variety of analyses can be performed, including multivariate sample comparison, clustering of sites or samples, dimension reduction (ordination) for visualization purposes, or inferring interaction networks. This paper presents the general PLN framework and illustrates its use on a series a typical experimental datasets. All the models and methods are implemented in the R package PLNmodels, available from cran.r-project.org.

Polypore fungi as a flagship group to indicate changes in biodiversity - a test case from Estonia

Polyporous fungi, a morphologically delineated group of Agaricomycetes (Basidiomycota), are considered well studied in Europe and used as model group in ecological studies and for conservation. Such broad interest, including widespread sampling and DNA based taxonomic revisions, is rapidly transforming our basic understanding of polypore diversity and natural history. We integrated over 40,000 historical and modern records of polypores in Estonia (hemiboreal Europe), revealing 227 species, and including Polyporus submelanopus and P. ulleungus as novelties for Europe. Taxonomic and conservation problems were distinguished for 13 unresolved subgroups. The estimated species pool exceeds 260 species in Estonia, including at least 20 likely undescribed species (here documented as distinct DNA lineages related to accepted species in, e.g., Ceriporia, Coltricia, Physisporinus, Sidera and Sistotrema). Four broad ecological patterns are described: (1) polypore assemblage organization in natural forests follows major soil and tree-composition gradients; (2) landscape-scale polypore diversity homogenizes due to draining of peatland forests and reduction of nemoral broad-leaved trees (wooded meadows and parks buffer the latter); (3) species having parasitic or brown-rot life-strategies are more substrate-specific; and (4) assemblage differences among woody substrates reveal habitat management priorities. Our update reveals extensive overlap of polypore biota throughout North Europe. We estimate that in Estonia, the biota experienced ca. 3-5% species turnover during the twentieth century, but exotic species remain rare and have not attained key functions in natural ecosystems. We encourage new regional syntheses on long studied fungal groups to obtain landscape-scale understanding of species pools, and for elaborating fungal indicators for biodiversity assessments.

Fungal Communities Are Important Determinants of Bacterial Community Composition in Deadwood

Understanding the interactive dynamics between fungal and bacterial communities is important to gain predictive knowledge on ecosystem functioning. However, little is known about the mechanisms behind fungal-bacterial associations and the directionality of species interactions.

Fungal-bacterial interactions play a key role in the functioning of many ecosystems. Thus, understanding their interactive dynamics is of central importance for gaining predictive knowledge on ecosystem functioning. However, it is challenging to disentangle the mechanisms behind species associations from observed co-occurrence patterns, and little is known about the directionality of such interactions. Here, we applied joint species distribution modeling to high-throughput sequencing data on co-occurring fungal and bacterial communities in deadwood to ask whether fungal and bacterial co-occurrences result from shared habitat use (i.e., deadwood’s properties) or whether there are fungal-bacterial interactive associations after habitat characteristics are taken into account. Moreover, we tested the hypothesis that the interactions are mainly modulated through fungal communities influencing bacterial communities. For that, we quantified how much the predictive power of the joint species distribution models for bacterial and fungal community improved when accounting for the other community. Our results show that fungi and bacteria form tight association networks (i.e., some species pairs co-occur more frequently and other species pairs co-occur less frequently than expected by chance) in deadwood that include common (or opposite) responses to the environment as well as (potentially) biotic interactions. Additionally, we show that information about the fungal occurrences and abundances increased the power to predict the bacterial abundances substantially, whereas information about the bacterial occurrences and abundances increased the power to predict the fungal abundances much less. Our results suggest that fungal communities may mainly affect bacteria in deadwood.

IMPORTANCE Understanding the interactive dynamics between fungal and bacterial communities is important to gain predictive knowledge on ecosystem functioning. However, little is known about the mechanisms behind fungal-bacterial associations and the directionality of species interactions. Applying joint species distribution modeling to high-throughput sequencing data on co-occurring fungal-bacterial communities in deadwood, we found evidence that nonrandom fungal-bacterial associations derive from shared habitat use as well as (potentially) biotic interactions. Importantly, the combination of cross-validations and conditional cross-validations helped us to answer the question about the directionality of the biotic interactions, providing evidence that suggests that fungal communities may mainly affect bacteria in deadwood. Our modeling approach may help gain insight into the directionality of interactions between different components of the microbiome in other environments.

Species co-occurrence networks of ground beetles in managed grasslands

Grassland biodiversity, including traditional rural biotopes maintained by traditional agricultural practices, has become threatened worldwide. Road verges have been suggested to be complementary or compensatory habitats for species inhabiting grasslands. Species co-occurrence patterns linked with species traits can be used to separate between the different mechanisms (stochasticity, environmental filtering, biotic interactions) behind community structure. Here, we study species co-occurrence networks and underlying mechanisms of ground beetle species (Carabidae) in three different managed grassland types (meadows, pastures, road verges, n = 12 in each type) in Central Finland. We aimed to find out whether road verges can be considered as compensatory to traditional rural biotopes (meadows and pastures). We found that stochasticity explained over 90% of the pairwise co-occurrences, and the non-random co-occurrences were best explained by environmental filtering, regardless of the grassland type. However, the identities and traits of the species showing non-random co-occurrences differed among the habitat types. Thus, environmental factors behind environmental filtering differ among the habitat types and are related to the site-specific characteristics and variation therein. This poses challenges to habitat management since the species’ response to management action may depend on the site-specific characteristics. Although road verges are not fully compensatory to meadows and pastures, the high similarity of species richness and the high level of shared species suggest that for carabids road verges may be corridors connecting the sparse network of the remaining traditional rural biotopes.

Protecting Biodiversity (in All Its Complexity): New Models and Methods

We are facing a biodiversity crisis at the same time as we are acquiring an unprecedented view of the world's biodiversity. Vast new datasets (e.g., species distributions, traits, phylogenies, and interaction networks) hold knowledge to better comprehend the depths of biodiversity change, reliably anticipate these changes, and inform conservation actions. To harness this information for conservation, we need to integrate the largely independent fields of biodiversity modeling and conservation. We highlight new developments in each respective field, early examples of how they are being brought together, and ideas for a future synthesis such that conservation decisions can be made with fuller awareness of the biodiversity at stake.

The emergent interactions that govern biodiversity change

Observational studies have not yet shown that environmental variables can explain pervasive nonlinear patterns of species abundance, because those patterns could result from (indirect) interactions with other species (e.g., competition), and models only estimate direct responses. The experiments that could extract these indirect effects at regional to continental scales are not feasible. Here, a biophysical approach quantifies environment- species interactions (ESI) that govern community change from field data. Just as species interactions depend on population abundances, so too do the effects of environment, as when drought is amplified by competition. By embedding dynamic ESI within framework that admits data gathered on different scales, we quantify responses that are induced indirectly through other species, including probabilistic uncertainty in parameters, model specification, and data. Simulation demonstrates that ESI are needed for accurate interpretation. Analysis demonstrates how nonlinear responses arise even when their direct responses to environment are linear. Applications to experimental lakes and the Breeding Bird Survey (BBS) yield contrasting estimates of ESI. In closed lakes, interactions involving phytoplankton and their zooplankton grazers play a large role. By contrast, ESI are weak in BBS, as expected where year-to-year movement degrades the link between local population growth and species interactions. In both cases, nonlinear responses to environmental gradients are induced by interactions between species. Stability analysis indicates stability in the closed-system lakes and instability in BBS. The probabilistic framework has direct application to conservation planning that must weigh risk assessments for entire habitats and communities against competing interests.

Predictor species: Improving assessments of rare species occurrence by modeling environmental co-responses

Designing an effective conservation strategy requires understanding where rare species are located. Because rare species can be difficult to find, ecologists often identify other species called conservation surrogates that can help inform the distribution of rare species. Species distribution models typically rely on environmental data when predicting the occurrence of species, neglecting the effect of species' co-occurrences and biotic interactions. Here, we present a new approach that uses Bayesian networks to improve predictions by modeling environmental co-responses among species. For species from a European peat bog community, our approach consistently performs better than single-species models and better than conventional multi-species approaches that include the presence of nontarget species as additional independent variables in regression models. Our approach performs particularly well with rare species and when calibration data are limited. Furthermore, we identify a group of "predictor species" that are relatively common, insensitive to the presence of other species, and can be used to improve occurrence predictions of rare species. Predictor species are distinct from other categories of conservation surrogates such as umbrella or indicator species, which motivates focused data collection of predictor species to enhance conservation practices.

Sensitivity of comorbidity network analysis

OBJECTIVES

Comorbidity network analysis (CNA) is a graph-theoretic approach to systems medicine based on associations revealed from disease co-occurrence data. Researchers have used CNA to explore epidemiological patterns, differentiate populations, characterize disorders, and more; but these techniques have not been comprehensively evaluated. Our objectives were to assess the stability of common CNA techniques.

MATERIALS AND METHODS

We obtained seven co-occurrence data sets, most from previous CNAs, coded using several ontologies. We constructed comorbidity networks under various modeling procedures and calculated summary statistics and centrality rankings. We used regression, ordination, and rank correlation to assess these properties' sensitivity to the source of data and construction parameters.

RESULTS

Most summary statistics were robust to variation in link determination but somewhere sensitive to the association measure. Some more effectively than others discriminated among networks constructed from different data sets. Centrality rankings, especially among hubs, were somewhat sensitive to link determination and highly sensitive to ontology. As multivariate models incorporated additional effects, comorbid associations among low-prevalence disorders weakened while those between high-prevalence disorders shifted negative.

DISCUSSION

Pairwise CNA techniques are generally robust, but some analyses are highly sensitive to certain parameters. Multivariate approaches expose additional conceptual and technical limitations to the usual pairwise approach.

CONCLUSION

We conclude with a set of recommendations we believe will help CNA researchers improve the robustness of results and the potential of follow-up research.

Parasite associations predict infection risk: incorporating co-infections in predictive models for neglected tropical diseases

BACKGROUND

Schistosomiasis and infection by soil-transmitted helminths are some of the world's most prevalent neglected tropical diseases. Infection by more than one parasite (co-infection) is common and can contribute to clinical morbidity in children. Geostatistical analyses of parasite infection data are key for developing mass drug administration strategies, yet most methods ignore co-infections when estimating risk. Infection status for multiple parasites can act as a useful proxy for data-poor individual-level or environmental risk factors while avoiding regression dilution bias. Conditional random fields (CRF) is a multivariate graphical network method that opens new doors in parasite risk mapping by (i) predicting co-infections with high accuracy; (ii) isolating associations among parasites; and (iii) quantifying how these associations change across landscapes.

METHODS

We built a spatial CRF to estimate infection risks for Ascaris lumbricoides, Trichuris trichiura, hookworms (Ancylostoma duodenale and Necator americanus) and Schistosoma mansoni using data from a national survey of Rwandan schoolchildren. We used an ensemble learning approach to generate spatial predictions by simulating from the CRF's posterior distribution with a multivariate boosted regression tree that captured non-linear relationships between predictors and covariance in infection risks. This CRF ensemble was compared against single parasite gradient boosted machines to assess each model's performance and prediction uncertainty.

RESULTS

Parasite co-infections were common, with 19.57% of children infected with at least two parasites. The CRF ensemble achieved higher predictive power than single-parasite models by improving estimates of co-infection prevalence at the individual level and classifying schools into World Health Organization treatment categories with greater accuracy. The CRF uncovered important environmental and demographic predictors of parasite infection probabilities. Yet even after capturing demographic and environmental risk factors, the presences or absences of other parasites were strong predictors of individual-level infection risk. Spatial predictions delineated high-risk regions in need of anthelminthic treatment interventions, including areas with higher than expected co-infection prevalence.

CONCLUSIONS

Monitoring studies routinely screen for multiple parasites, yet statistical models generally ignore this multivariate data when assessing risk factors and designing treatment guidelines. Multivariate approaches can be instrumental in the global effort to reduce and eventually eliminate neglected helminth infections in developing countries.

Enhanced inference of ecological networks by parameterizing ensembles of population dynamics models constrained with prior knowledge

Background

Accurate network models of species interaction could be used to predict population dynamics and be applied to manage real world ecosystems. Most relevant models are nonlinear, however, and data available from real world ecosystems are too noisy and sparsely sampled for common inference approaches. Here we improved the inference of generalized Lotka–Volterra (gLV) ecological networks by using a new optimization algorithm to constrain parameter signs with prior knowledge and a perturbation-based ensemble method.

Results

We applied the new inference to long-term species abundance data from the freshwater fish community in the Illinois River, United States. We constructed an ensemble of 668 gLV models that explained 79% of the data on average. The models indicated (at a 70% level of confidence) a strong positive interaction from emerald shiner (Notropis atherinoides) to channel catfish (Ictalurus punctatus), which we could validate using data from a nearby observation site, and predicted that the relative abundances of most fish species will continue to fluctuate temporally and concordantly in the near future. The network shows that the invasive silver carp (Hypophthalmichthys molitrix) has much stronger impacts on native predators than on prey, supporting the notion that the invader perturbs the native food chain by replacing the diets of predators.

Conclusions

Ensemble approaches constrained by prior knowledge can improve inference and produce networks from noisy and sparsely sampled time series data to fill knowledge gaps on real world ecosystems. Such network models could aid efforts to conserve ecosystems such as the Illinois River, which is threatened by the invasion of the silver carp.

Improving distribution models of riparian vegetation with mobile laser scanning and hydraulic modelling

This study aimed at illustrating how direct measurements, mobile laser scanning and hydraulic modelling can be combined to quantify environmental drivers, improve vegetation models and increase our understanding of vegetation patterns in a sub-arctic river valley. Our results indicate that the resultant vegetation models successfully predict riparian vegetation patterns (Rho = 0.8 for total species richness, AUC = 0.97 for distribution) and highlight differences between eight functional species groups (Rho 0.46-0.84; AUC 0.79-0.93; functional group-specific effects). In our study setting, replacing the laser scanning-based and hydraulic modelling-based variables with a proxy variable elevation did not significantly weaken the models. However, using directly measured and modelled variables allows relating species patterns to e.g. stream power or the length of the flood-free period. Substituting these biologically relevant variables with proxies mask important processes and may reduce the transferability of the results into other sites. At the local scale, the amount of litter is a highly important driver of total species richness, distribution and abundance patterns (relative influences 49, 72 and 83%, respectively) and across all functional groups (13-57%; excluding lichen species richness) in the sub-arctic river valley. Moreover, soil organic matter and soil water content shape vegetation patterns (on average 16 and 7%, respectively). Fluvial disturbance is a key limiting factor only for lichen, bryophyte and dwarf shrub species in this environment (on average 37, 6 and 10%, respectively). Fluvial disturbance intensity is the most important component of disturbance for most functional groups while the length of the disturbance-free period is more relevant for lichens. We conclude that striving for as accurate quantifications of environmental drivers as possible may reveal important processes and functional group differences and help anticipate future changes in vegetation. Mobile laser scanning, high-resolution digital elevation models and hydraulic modelling offer useful methodology for improving correlative vegetation models.

Untangling the dynamics of persistence and colonization in microbial communities

A central goal of community ecology is to infer biotic interactions from observed distributions of co-occurring species. Evidence for biotic interactions, however, can be obscured by shared environmental requirements, posing a challenge for statistical inference. Here, we introduce a dynamic statistical model, based on probit regression, that quantifies the effects of spatial and temporal covariance in longitudinal co-occurrence data. We separate the fixed pairwise effects of species occurrences on persistence and colonization rates, a potential signal of direct interactions, from latent pairwise correlations in occurrence, a potential signal of shared environmental responses. We first validate our modeling framework with several simulation studies. Then, we apply the approach to a pressing epidemiological question by examining how human papillomavirus (HPV) types coexist. Our results suggest that while HPV types respond similarly to common host traits, direct interactions are sparse and weak, so that HPV type diversity depends largely on shared environmental drivers. Our modeling approach is widely applicable to microbial communities and provides valuable insights that should lead to more directed hypothesis testing and mechanistic modeling.

How sample heterogeneity can obscure the signal of microbial interactions

Microbial community data are commonly subjected to computational tools such as correlation networks, null models, and dynamic models, with the goal of identifying the ecological processes structuring microbial communities. A major assumption of these methods is that the signs and magnitudes of species interactions and vital rates can be reliably parsed from observational data on species' (relative) abundances. However, we contend that this assumption is violated when sample units contain any underlying spatial structure. Here, we show how three phenomena-Simpson's paradox, context-dependence, and nonlinear averaging-can lead to erroneous conclusions about population parameters and species interactions when samples contain heterogeneous mixtures of populations or communities. At the root of this issue is the fundamental mismatch between the spatial scales of species interactions (micrometers) and those of typical microbial community samples (millimeters to centimetres). These issues can be overcome by measuring and accounting for spatial heterogeneity at very small scales, which will lead to more reliable inference of the ecological mechanisms structuring natural microbial communities.

Using co-occurrence network topology in assessing ecological stress in benthic macroinvertebrate communities

Ecological monitoring of streams has often focused on assessing the biotic integrity of individual benthic macroinvertebrate (BMI) communities through local measures of diversity, such as taxonomic or functional richness. However, as individual BMI communities are frequently linked by a variety of ecological processes at a regional scale, there is a need to assess biotic integrity of groups of communities at the scale of watersheds. Using 4,619 sampled communities of streambed BMIs, we investigate this question using co-occurrence networks generated from groups of communities selected within California watersheds under different levels of stress due to upstream land use. Building on a number of arguments in theoretical ecology and network theory, we propose a framework for the assessment of the biotic integrity of watershed-scale groupings of BMI communities using measures of their co-occurrence network topology. We found significant correlations between stress, as described by a mean measure of upstream land use within a watershed, and topological measures of co-occurrence networks such as network size (r = -.81, p < 10-4), connectance (r = .31, p < 10-4), mean co-occurrence strength (r = .25, p < 10-4), degree heterogeneity (r = -.10, p < 10-4), and modularity (r = .11, p < 10-4). Using these five topological measures, we constructed a linear model of biotic integrity, here a composite of taxonomic and functional diversity known as the California Stream Condition Index, of groups of BMI communities within a watershed. This model can account for 66% of among-watershed variation in the mean biotic integrity of communities. These observations imply a role for co-occurrence networks in assessing the current status of biotic integrity for BMI communities, as well as their potential use in assessing other ecological communities.

Site-Specific Microbial Decomposer Communities Do Not Imply Faster Decomposition: Results from a Litter Transplantation Experiment

Microbes drive leaf litter decomposition, and their communities are adapted to the local vegetation providing that litter. However, whether these local microbial communities confer a significant home-field advantage in litter decomposition remains unclear, with contrasting results being published. Here, we focus on a litter transplantation experiment from oak forests (home site) to two away sites without oak in South Tyrol (Italy). We aimed to produce an in-depth analysis of the fungal and bacterial decomposer communities using Illumina sequencing and qPCR, to understand whether local adaptation occurs and whether this was associated with litter mass loss dynamics. Temporal shifts in the decomposer community occurred, reflecting changes in litter chemistry over time. Fungal community composition was site dependent, while bacterial composition did not differ across sites. Total litter mass loss and rates of litter decomposition did not change across sites. Litter quality influenced the microbial community through the availability of different carbon sources. Additively, our results do not support the hypothesis that locally adapted microbial decomposers lead to a greater or faster mass loss. It is likely that high functional redundancy within decomposer communities regulated the decomposition, and thus greater future research attention should be given to trophic guilds rather than taxonomic composition.

Good neighbors aplenty: fungal endophytes rarely exhibit competitive exclusion patterns across a span of woody habitats

Environmental forces and biotic interactions, both positive and negative, structure ecological communities, but their relative roles remain obscure despite strong theory. For instance, ecologically similar species, based on the principle of limiting similarity, are expected to be most competitive and show negative interactions. Specious communities that assemble along broad environmental gradients afford the most power to test theory, but the communities often are difficult to quantify. Microbes, specifically fungal endophytes of wood, are especially suited for testing community assembly theory because they are relatively easy to sample across a comprehensive range of environmental space with clear axes of variation. Moreover, endophytes mediate key forest carbon cycle processes, and although saprophytic fungi from dead wood typically compete, endophytic fungi in living wood may enhance success through cooperative symbioses. To classify interactions within endophyte communities, we analyzed fungal DNA barcode variation across 22 woody plant species growing in woodlands near Richmond, New South Wales, Australia. We estimated the response of endophytes to the measured wood environment (i.e., 11 anatomical and chemical wood traits) and each other using latent-variable models and identified recurrent communities across wood environments using model-based classification. We used this information to evaluate whether (1) co-occurrence patterns are consistent with strong competitive exclusion, and (2) a priori classifications by trophic mode and phylum distinguish taxa that are more likely to have positive vs. negative associations under the principle of limiting similarity. Fungal endophytes were diverse (mean = 140 taxa/sample), with differences in community composition structured by wood traits. Variation in wood water content and carbon concentration were associated with especially large community shifts. Surprisingly, after accounting for wood traits, fungal species were still more than three times more likely to have positive than negative co-occurrence patterns. That is, patterns consistent with strong competitive exclusion were rare, and positive interactions among fungal endophytes were more common than expected. Confirming the frequency of positive vs. negative interactions among fungal taxa requires experimental tests, and our findings establish clear paths for further study. Evidence to date intriguingly suggests that, across a wide range of wood traits, cooperation may outweigh combat for these fungi.

Patterns of richness of freshwater mollusks from Chile: predictions of its distribution based on null models

The freshwater mussels from Chile are characterized by a high percentage of endemism and a fragmented latitudinal diversity, which has been attributed to the features and geomorphological history sculpted by the hydrographic basins. In this work, a set of hypothesis under a macroecological approach is addressed, with the aim to explore environmental, topographic and hydrological factors that define the latitudinal distribution of this mussel group. In order to achieve this goal, Rapoport’s rule, geometrics limits and co-ocurrence were evaluated. In addition, we analyze the source and sink hypotheses through the nested analysis. We observed a noticeable mid-domain effect (MDE), where a major richness than expected was randomly observed between 40 and 41°S. The results revealed that the distribution pattern was not concordant with Rapoport’s rule (r = 0.123; p = 0.128). Regarding to historical dynamic of the distribution, the results show a significant nestedness pattern, suggesting a source-sink dynamic, that is, poorer communities are a subset of richer communities in species. According to the co-occurrence analysis, an aggregate pattern existed, suggesting potential regulatory mechanisms. The specific richness pattern is explained by the variable seasonality of the temperature with a variance percentage explained of 35%. The full model indicated that variables which characterize the heterogeneity of habitat (i.e. range, Shannon), water availability (i.e., precipitation, density of water bodies) and topography (i.e., altitude area available) jointly explain 40% of the variability of the observed richness. This study shows that the geographical distribution of mollusc richness is mainly explained by mainly climatic and topographical environmental components, as well as by the source-sink dynamics.

Inference of Environmental Factor-Microbe and Microbe-Microbe Associations from Metagenomic Data Using a Hierarchical Bayesian Statistical Model

The inference of associations between environmental factors and microbes and among microbes is critical to interpreting metagenomic data, but compositional bias, indirect associations resulting from common factors, and variance within metagenomic sequencing data limit the discovery of associations. To account for these problems, we propose metagenomic Lognormal-Dirichlet-Multinomial (mLDM), a hierarchical Bayesian model with sparsity constraints, to estimate absolute microbial abundance and simultaneously infer both conditionally dependent associations among microbes and direct associations between microbes and environmental factors. We empirically show the effectiveness of the mLDM model using synthetic data, data from the TARA Oceans project, and a colorectal cancer dataset. Finally, we apply mLDM to 16S sequencing data from the western English Channel and report several associations. Our model can be used on both natural environmental and human metagenomic datasets, promoting the understanding of associations in the microbial community.