Global land-use and land-cover data for ecologists: Historical, current, and future scenarios

Land-use land-cover (LULC) data are important predictors of species occurrence and biodiversity threat. Although there are LULC datasets available for ecologists under current conditions, there is a lack of such data under historical and future climatic conditions. This hinders, for example, projecting niche and distribution models under global change scenarios at different times. The Land Use Harmonization Project (LUH2) is a global terrestrial dataset at 0.25o spatial resolution that provides LULC data from 850 to 2300 for 12 LULC state classes. The dataset, however, is compressed in a file format (NetCDF) that is incompatible with most ecological analysis and intractable for most ecologists. Here we selected and transformed the LUH2 data in order to make it more useful for ecological studies. We provide LULC for every year from 850 to 2100, with data from 2015 on provided under two Shared Socioeconomic Pathways (SSP2 and SSP5). We provide two types of file for each year: separate files with continuous values for each of the 12 LULC state classes, and a single categorical file with all state classes combined. To create the categorical layer, we assigned the state with the highest value in a given pixel among the 12 continuous data. The final dataset provides LULC data for 1251 years that will be of interest for macroecology, ecological niche modeling, global change analysis, and other applications in ecology and conservation. We also provide a description of LUH2 prediction of future LULC change through time.

Visualizing species richness and site similarity from presence-absence matrices

Species richness and similarity of biotas among distinct sites are important quantities in biogeography. Indices derived from presence-absence matrices are used to represent these quantities in so-called diversity-range plots. The most commonly used diversity-range plot, however, has multiple special cases and its interpretation is cumbersome. Here we present an equivalent formulation that is geometrically simpler and has no special cases. In addition, we introduce a method to identify the statistical significance of the dispersion field, an index that represents how similar species composition is in a cell with respect to the whole area. The new diversity-range plot is a promising tool to explore biodiversity and endemism in a region as the values shown in this plot and whether they are statistically significant or not can also be represented in geography.

Predicting multi-species bark beetle (Coleoptera: Curculionidae: Scolytinae) occurrence in Alaska: First use of open access big data mining and open source GIS to provide robust inference and a role model for progress in forest conservation

Native bark beetles (Coleoptera: Curculionidae: Scolytinae) are a multi-species complex that rank among the key disturbances of coniferous forests of western North America. Many landscape-level variables are known to influence beetle outbreaks, such as suitable climatic conditions, spatial arrangement of incipient populations, topography, abundance of mature host trees, and disturbance history that include former outbreaks and fire. We assembled the first open access data, which can be used in open source GIS platforms, for understanding the ecology of the bark beetle organism in Alaska. We used boosted classification and regression tree as a machine learning data mining algorithm to model-predict the relationship between 14 environmental variables, as model predictors, and 838 occurrence records of 68 bark beetle species compared to pseudo-absence locations across the state of Alaska. The model predictors include topography- and climate-related predictors as well as feature proximities and anthropogenic factors. We were able to model, predict, and map the multi-species bark beetle occurrences across the state of Alaska on a 1-km spatial resolution in addition to providing a good quality environmental dataset freely accessible for the public. About 16% of the mixed forest and 59% of evergreen forest are expected to be occupied by the bark beetles based on current climatic conditions and biophysical attributes of the landscape. The open access dataset that we prepared, and the machine learning modeling approach that we used, can provide a foundation for future research not only on scolytines but for other multi-species questions of concern, such as forest defoliators, and small and big game wildlife species worldwide.

The Abundant Niche-centroid Hypothesis: Key Points About Unfilled Niches and the Potential Use of Supraspecfic Modeling Units

Correlative estimates of fundamental niches are gaining momentum as an alternative to predict species’ abundances, particularly via the abundant niche-centroid hypothesis (an expected inverse relationship between species’ abundance variation across its range and the distance to the geometric centroid of its multidimensional ecological niche). The main goal of this review is to recapitulate what has been done, where we are now, and where should we move towards in regards to this hypothesis. Despite evidence in support of the abundance-distance to niche centroid relationship, its usefulness has been highly debated, although with little consideration of the underlying theory regarding the circumstances that might break down the relationship. We address some key points about the conditions needed to test the hypothesis in correlative studies, specifically in relation to nichecharacterization and configurations of the Biotic-Abiotic-Mobility (BAM) framework to illustrate the problem of unfilled niches. Using a created supraspecific modeling unit, we show that species for which only a portion of their fundamental niche is represented in their area of historical accessibility (M)—i.e., when the environmental equilibrium condition is violated—it is impossible to characterize their true niche centroid. Therefore, we strongly recommend to analyze this assumption prior toassess the abundant niche-centroid hypothesis. Finally, we discuss the potential of using modeling units above the species level for cases in which environmental conditions associated with species’ occurrences may not be sufficient to fully characterize their fundamental niches.

Weighing the Evidence for the Abundant-Center Hypothesis

The abundant-center hypothesis posits that species density should be highest in the center of the geographic range or climatic niche of a species, based on the idea that the center of either will be the area with the highest demographic performance (e.g., greater fecundity, survival, or carrying capacity). While intuitive, current support for the hypothesis is quite mixed. Here, we discuss the current state of the abundant-center hypothesis, highlighting the relatively low level of support for the relationship. We then discuss the potential reasons for this lack of empirical support, emphasizing the inherent ecological complexity which may prevent the observation of the abundant-center in natural systems. This includes the role of non-equilibrial population dynamics, species interactions, landscape structure, and dispersal processes, as well as variable data quality and inconsistent methodology. The incorporation of this complexity into studies of the distribution of species densities in geographic or niche space may underlie the limited empirical support for the abundant-center hypothesis. We end by discussing potentially fruitful research avenues. Most notably, we highlight the need for theoretical development and controlled experimental testing of the abundant-center hypothesis.

Presence-only and Presence-absence Data for Comparing Species Distribution Modeling Methods

Species distribution models (SDMs) are widely used to predict and study distributions of species. Many different modeling methods and associated algorithms are used and continue to emerge. It is important to understand how different approaches perform, particularly when applied to species occurrence records that were not gathered in struc­tured surveys (e.g. opportunistic records). This need motivated a large-scale, collaborative effort, published in 2006, that aimed to create objective comparisons of algorithm performance. As a benchmark, and to facilitate future comparisons of approaches, here we publish that dataset: point location records for 226 anonymized species from six regions of the world, with accompanying predictor variables in raster (grid) and point formats. A particularly interesting characteristic of this dataset is that independent presence-absence survey data are available for evaluation alongside the presence-only species occurrence data intended for modeling. The dataset is available on Open Science Framework and as an R package and can be used as a benchmark for modeling approaches and for testing new ways to evaluate the accuracy of SDMs.

General Theory and Good Practices in Ecological Niche Modeling: A Basic Guide

Ecological niche modeling (ENM) and species distribution modeling (SDM) are sets of tools that allow the estimation of distributional areas on the basis of establishing relationships among known occurrences and environmental variables. These tools have a wide range of applications, particularly in biogeography, macroecology, and conservation biology, granting prediction of species potential distributional patterns in the present and dynamics of these areas in different periods or scenarios. Due to their relevance and practical applications, the usage of these methodologies has significantly increased throughout the years. Here, we provide a manual with the basic routines used in this field and a practical example of its implementation to promote good practices and guidance for new users.

Some thoughts about the challenge of inferring ecological interactions from spatial data.

Dr. Luis Escobar asked me to provide a joint review of the submissions by Stephens et al. (2019, this issue) and Peterson et al. (2019, this issue). I pulled thoughts together, but by the time I sent them along, he had received other reviews and made an editorial decision. He felt my perspective might nevertheless warrant publishing as a commentary alongside these two pieces. My review was of the original submissions, which are now appearing with minor, mainly cosmetic changes. I have only lightly edited the text of my review, and added a few additional thoughts and pertinent references. Neither group of authors has seen my commentary, and so I am responsible for any omissions or lapses in interpretation. The protocol developed by Stephens seems to me a potentially valuable exploratory tool in describing patterns of co-occurrence, but I note several potential problems in identifying interactions usingsolely this protocol. I also gently disagree with Peterson et al., who state flatly that co-occurrence data can shed no light at all on interspecific interactions. I suggest there are a number of counter-examples to this claim in the literature. I argue that spatiotemporal data, when available, iprovide a much more powerful tool for discerning interactions, than do staticspatial data. Finally, I use a simple thought experiment to point out that biotic drivers could be playing a key causal role in limitnig distributions, even in equisitlvely accurate SDMs that use only abiotic (scenopoetic) data as input data.

Can Ecological Interactions be Inferred from Spatial Data?

The characterisation and quantication of ecological interactions, and the construction of species distributions and their associated ecological niches, is of fundamental theoretical and practical importance. In this paper we give an overview of a Bayesian inference framework, developed over the last 10 years, which, using spatial data, offers a general formalism within which ecological interactions may be characterised and quantied. Interactions are identied through deviations of the spatial distribution of co-occurrences of spatial variables relative to a benchmark for the non-interacting system, and based on a statistical ensemble of spatial cells. The formalism allows for the integration of both biotic and abiotic factors of arbitrary resolution. We concentrate on the conceptual and mathematical underpinnings of the formalism, showing how, using the Naive Bayes approximation, it can be used to not only compare and contrast the relative contribution from each variable, but also to construct species distributions and niches based on arbitrary variable type. We show how the formalism can be used to quantify confounding and therefore help disentangle the complex causal chains that are present in ecosystems. We also show species distributions and their associated niches can be used to infer standard "micro" ecological interactions, such as predation and parasitism. We present several representative use cases that validate our framework, both in terms of being consistent with present knowledge of a set of known interactions, as well as making and validating predictions about new, previously unknown interactions in the case of zoonoses.

Co-occurrence Networks do not Support Identification of Biotic Interactions

We assess a body of work that has attempted to use co-occurrence networks to infer the existence and type of biotic interactions between species. Although we see considerable interest in the approach as an exploratory tool for understanding patterns of co-occurrence of species, we note and describe numerous problems in the step of inferring biotic interactions from the co-occurrence patterns. These problems are both theoretical and empirical in nature, and limit confidence in inferences about interactions rather severely. We examine a series of examples that demonstrates striking discords between interactions inferred from co-occurrence patterns and previous experimental results and known life-history details.