An understanding of how biodiversity is distributed across space is key to much of ecology and conservation. Many predictive modelling approaches have been developed to estimate the distribution of biodiversity over various spatial scales. Community modelling techniques may offer many benefits over single-species modelling. However, techniques capable of estimating precise species makeups of communities are highly data intensive and thus often limited in their applicability. Here we present an R package, spectre, which can predict regional community composition at a fine spatial resolution using only sparsely sampled biological data. The package can predict the presence and absence of all species in an area, both known and unknown, at the sample site scale. Underlying the spectre package is a min-conflicts optimisation algorithm that predicts species’ presences and absences throughout an area using estimates of α-, β-, and γ-diversity. We demonstrate the utility of the spectre package using a spatially-explicit simulated ecosystem to assess the accuracy of the package’s results. spectre offers a simple-to-use tool with which to accurately predict community compositions across varying scales, facilitating further research and knowledge acquisition into this fundamental aspect of ecology.

The simulated community datasets were built using the virtualspecies V1.5.1 R package (Leroy et al., 2016), which generates spatially-explicit presence/absence matrices from habitat suitability maps. We simulated these suitability maps using Gaussian fields neutral landscapes produced using the NLMR V1.0 R package (Sciaini et al., 2018). To allow for some level of overlap between species suitability maps, we divided the γ-diversity (i.e., the total number of simulated species) by an adjustable correlation value to create several species groups that share suitability maps. Using a full factorial design, we developed 81 presence/absence maps varying across four axes (see Supplemental Table 1 and Supplemental Figure 1): 1) landscape size, representing the number of sites in the simulated landscape; 2) γ-diversity; 3) the level of correlation among species suitability maps, with greater correlations resulting in fewer shared species groups among suitability maps; and 4) the habitat suitability threshold of the virtual species distribution function. The latter corresponds to the level to which a species is a generalist or a specialist represented by the degree a species distribution can be outside its preferred habitat type from a suitability map. Every variable set in the factorial design was replicated three times. Species richness, pairwise dissimilarity and γ-diversity measures (used as the inputs for the spectre algorithm) were taken directly from the simulated community composition maps, thus avoiding any errors produced in the process of estimating these values. 

The data is stored here in .RDS format and as such can be directly loaded into any R environment using the `readRDS()` function. Additionally, the data can be recreated and fully re-analysed by accessing it via the spectre usecase GitHub repo https://github.com/r-spatialecology/spectre_usecase.