Spatially autocorrelated sampling falsely inflates measures of accuracy for presence‐only niche models

Aim Environmental niche models that utilize presence‐only data have been increasingly employed to model species distributions and test ecological and evolutionary predictions. The ideal method for evaluating the accuracy of a niche model is to train a model with one dataset and then test model predictions against an independent dataset. However, a truly independent dataset is often not available, and instead random subsets of the total data are used for ‘training’ and ‘testing’ purposes. The goal of this study was to determine how spatially autocorrelated sampling affects measures of niche model accuracy when using subsets of a larger dataset for accuracy evaluation. Location The distribution of Centaurea maculosa (spotted knapweed; Asteraceae) was modelled in six states in the western United States: California, Oregon, Washington, Idaho, Wyoming and Montana. Methods Two types of niche modelling algorithms – the genetic algorithm for rule‐set prediction (GARP) and maximum entropy modelling (as implemented with Maxent) – were used to model the potential distribution of C. maculosa across the region. The effect of spatially autocorrelated sampling was examined by applying a spatial filter to the presence‐only data (to reduce autocorrelation) and then comparing predictions made using the spatial filter with those using a random subset of the data, equal in sample size to the filtered data. Results The accuracy of predictions from both algorithms was sensitive to the spatial autocorrelation of sampling effort in the occurrence data. Spatial filtering led to lower values of the area under the receiver operating characteristic curve plot but higher similarity statistic (I) values when compared with predictions from models built with random subsets of the total data, meaning that spatial autocorrelation of sampling effort between training and test data led to inflated measures of accuracy. Main conclusions The findings indicate that care should be taken when interpreting the results from presence‐only niche models when training and test data have been randomly partitioned but occurrence data were non‐randomly sampled (in a spatially autocorrelated manner). The higher accuracies obtained without the spatial filter are a result of spatial autocorrelation of sampling effort between training and test data inflating measures of prediction accuracy. If independently surveyed data for testing predictions are unavailable, then it may be necessary to explicitly account for the spatial autocorrelation of sampling effort between randomly partitioned training and test subsets when evaluating niche model predictions.     

Spind: a package for computing spatially corrected accuracy measures

Using an appropriate accuracy measure is essential for assessing prediction accuracy in species distribution modelling. Therefore, model evaluation as an analytical uncertainty is a challenging problem. Although a variety of accuracy measures for the assessment of prediction errors in presence/absence models is available, there is a lack of spatial accuracy measures, i.e. measures that are sensitive to the spatial arrangement of the predictions. We present ‘spind’, a new software package (based on the R software program) that provides spatial performance measures for grid‐based models. These accuracy measures are generalized, spatially corrected versions of the classical ones, thus enabling comparisons between them. Our method for evaluation consists of the following steps: 1) incorporate additional autocorrelation until spatial autocorrelation in predictions and actuals is balanced, 2) cross‐classify predictions and adjusted actuals in a 4 × 4 contingency table, 3) use a refined weighting pattern for errors, and 4) calculate weighted Kappa, sensitivity, specificity and subsequently ROC, AUC, TSS to get spatially corrected indices. To illustrate the impact of our spatial method we present an example of simulated data as well as an example of presence/absence data of the plant species Dianthus carthusianorum across Germany. Our analysis includes a statistic for the comparison of spatial and classical (non‐spatial) indices. We find that our spatial indices tend to result in higher values than classical ones. These differences are statistically significant at medium and high autocorrelation levels. We conclude that these spatial accuracy measures may contribute to evaluate prediction errors in presence/absence models, especially in case of medium or high degree of similarity of adjacent data, i.e. aggregated (clumped) or continuous species distributions.     

Model-based uncertainty in species range prediction

Aim Many attempts to predict the potential range of species rely on environmental niche (or ‘bioclimate envelope’) modelling, yet the effects of using different niche‐based methodologies require further investigation. Here we investigate the impact that the choice of model can have on predictions, identify key reasons why model output may differ and discuss the implications that model uncertainty has for policy‐guiding applications. Location The Western Cape of South Africa. Methods We applied nine of the most widely used modelling techniques to model potential distributions under current and predicted future climate for four species (including two subspecies) of Proteaceae. Each model was built using an identical set of five input variables and distribution data for 3996 sampled sites. We compare model predictions by testing agreement between observed and simulated distributions for the present day (using the area under the receiver operating characteristic curve (AUC) and kappa statistics) and by assessing consistency in predictions of range size changes under future climate (using cluster analysis). Results Our analyses show significant differences between predictions from different models, with predicted changes in range size by 2030 differing in both magnitude and direction (e.g. from 92% loss to 322% gain). We explain differences with reference to two characteristics of the modelling techniques: data input requirements (presence/absence vs. presence‐only approaches) and assumptions made by each algorithm when extrapolating beyond the range of data used to build the model. The effects of these factors should be carefully considered when using this modelling approach to predict species ranges. Main conclusions We highlight an important source of uncertainty in assessments of the impacts of climate change on biodiversity and emphasize that model predictions should be interpreted in policy‐guiding applications along with a full appreciation of uncertainty.     

Scaling down distribution maps from atlas data: a test of different approaches with virtual species

Aim Conservation managers are increasingly looking for modelled projections of species distributions to inform management strategies; however, the coarse resolution of available data usually compromises their helpfulness. The aim of this paper is to delineate and test different approaches for converting coarse‐grain occurrence data into high‐resolution predictions, and to clarify the conceptual circumstances affecting the accuracy of downscaled models. Location We used environmental data from a real landscape, southern Africa, and simulated species distributions within this landscape. Methods We built 10 virtual species at a resolution of 5 arcmin, and for each species we simulated atlas range maps at four decreasing resolutions (15, 30, 60, 120 arcmin). We tested the ability of three downscaling strategies to produce high‐resolution predictions using two modelling techniques: generalized linear models and generalized boosted models. We calibrated reference models with high‐resolution data and we compared the relative reduction of predictive performance in the downscaled models by using a null model approach. We also estimated the applicability of downscaling procedures to different situations by using distribution data for Mediterranean reptiles. Results All reference models achieved high performance measures. For all strategies, we observed a reduction of predictive performance proportional to the degree of downscaling. The differences in evaluation indices between reference models and downscaled projections obtained from atlases at 15 and 30 arcmin were never statistically significant. The accuracy of projections scaled down from 60 arcmin largely depended on the combination of approach and algorithm adopted. Projections scaled down from 120 arcmin gave misleading results in all cases. Main conclusions Moderate levels of downscaling allow for reasonably accurate results, regardless of the technique used. The most general effect of scaling down coarse‐grain data is the reduction of model specificity. The models can successfully delineate a species’ environmental association up until a 12‐fold downscaling, although with an increasing approximation that causes the overestimation of true distributions. We suggest appropriate procedures to mitigate the commission error introduced by downscaling at intermediate levels (approximately 12‐fold). Reductions of grain size > 12‐fold are discouraged.     

Informing management decisions for ecological networks,using dynamic models calibrated to noisy time‐series data

Well‐intentioned environmental management can backfire, causing unforeseen damage. To avoid this, managers and ecologists seek accurate predictions of the ecosystem‐wide impacts of interventions, given small and imprecise datasets, which is an incredibly difficult task. We generated and analysed thousands of ecosystem population time series to investigate whether fitted models can aid decision‐makers to select interventions. Using these time‐series data (sparse and noisy datasets drawn from deterministic Lotka‐Volterra systems with two to nine species, of known network structure), dynamic model forecasts of whether a species’ future population will be positively or negatively affected by rapid eradication of another species were correct > 70% of the time. Although 70% correct classifications is only slightly better than an uninformative prediction (50%), this classification accuracy can be feasibly improved by increasing monitoring accuracy and frequency. Our findings suggest that models may not need to produce well‐constrained predictions before they can inform decisions that improve environmental outcomes.     

Distribution models of estuarine fish species: The effect of sampling bias,species ecology and threshold selection on models' accuracy

Species distribution models (SDMs) relate presence/absence data to environmental variables, allowing to predict species environmental requirements and potential distribution. They have been increasingly used in fields such as ecology, biogeography and evolution, and often support conservation priorities and strategies. Thus, it becomes crucial to understand how trustworthy and reliable their predictions are. Different approaches, such as using ensemble methods (combining forecasts of different single models), or applying the most suitable threshold to transform continuous probability maps into species presences or absences, have been used to reduce model-based uncertainty. Taking into account the influence of biased sampling imprecision in species location, small datasets and species ecological characteristics, may also help to detect and compensate for uncertainty in the model building process. To investigate the effect of applying an ensemble approach, several threshold selection criteria and different datasets representing seasonal and spatial sampling bias, on models' accuracy, SDMs were built for four estuarine fish species with distinct use of the estuarine systems. Overall, predictions obtained with the ensemble approach were more accurate. Variability in accuracy metrics obtained with the nine threshold selection criteria applied was more pronounced for species with low prevalence and when sensitivity was calculated. Higher values of accuracy measures were registered with the threshold that maximizes the sum of sensitivity and specificity, and the threshold where the predicted prevalence equals the observed, whereas the 0.5 cut-off was unreliable, originating the lowest values for these metrics. Accuracy of models created from a spatially biased sampling was overall higher than accuracy of models created with a seasonally biased sampling or with the multi-year database created and this pattern was consistently obtained for marine migrant species, which use estuaries as nursery areas, presenting a seasonally and regular use of these ecosystems. The ecological dependence between these fish species and estuaries may add difficulties in the model building process, and needs to be taken into account, to improve their accuracy. The present study highlights the need for a thorough analysis of the critical underlying issues of the complete model building process to predict the distribution of estuarine fish species, due to the particular and dynamic nature of these ecosystems.     

Assessing the impacts of imperfect detection on estimates of diversity and community structure through multispecies occupancy modeling

Detecting all species in a given survey is challenging, regardless of sampling effort. This issue, more commonly known as imperfect detection, can have negative impacts on data quality and interpretation, most notably leading to false absences for rare or difficult‐to‐detect species. It is important that this issue be addressed, as estimates of species richness are critical to many areas of ecological research and management. In this study, we set out to determine the impacts of imperfect detection, and decisions about thresholds for inclusion in occupancy, on estimates of species richness and community structure. We collected data from a stream fish assemblage in Algonquin Provincial Park to be used as a representation of ecological communities. We then used multispecies occupancy modeling to estimate species‐specific occurrence probabilities while accounting for imperfect detection, thus creating a more informed dataset. This dataset was then compared to the original to see where differences occurred. In our analyses, we demonstrated that imperfect detection can lead to large changes in estimates of species richness at the site level and summarized differences in the community structure and sampling locations, represented through correspondence analyses.     

Spatial scale,topography and thermoregulatory behaviour interact when modelling species’ thermal niches

The spatial scale at which climate and species’ occupancy data are gathered, and the resolution at which ecological models are run, can strongly influence predictions of species performance and distributions. Running model simulations at coarse rather than fine spatial resolutions, for example, can determine if a model accurately predicts the distribution of a species. The impacts of spatial scale on a model's accuracy are particularly pronounced across mountainous terrain. Understanding how these discrepancies arise requires a modelling approach in which the underlying processes that determine a species’ distribution are explicitly described. Here we use a process‐based model to explore how spatial resolution, topography and behaviour alter predictions of a species thermal niche, which in turn constrains its survival and geographic distribution. The model incorporates biophysical equations to predict the operative temperature (T_e), thermal‐dependent performance and survival of a typical insect, with a complex life‐cycle, in its microclimate. We run this model with geographic data from a mountainous terrain in South Africa using climate data at three spatial resolutions. We also explore how behavioural thermoregulation affects predictions of a species performance and survival by allowing the animal to select the optimum thermal location within each coarse‐grid cell. At the regional level, coarse‐resolution models predicted lower T_e at low elevations and higher T_e at high elevations than models run at fine‐resolutions. These differences were more prominent on steep, north‐facing slopes. The discrepancies in T_e in turn affected estimates of the species thermal niche. The modelling framework revealed how spatial resolution and topography influence predictions of species distribution models, including the potential impacts of climate change. These systematic biases must be accounted for when interpreting the outputs of future modelling studies, particularly when species distributions are predicted to shift from uniform to topographically heterogeneous landscapes.     

Geographic selection bias of occurrence data influences transferability of invasive Hydrilla verticillata distribution models

Due to socioeconomic differences, the accuracy and extent of reporting on the occurrence of native species differs among countries, which can impact the performance of species distribution models. We assessed the importance of geographical biases in occurrence data on model performance using Hydrilla verticillata as a case study. We used Maxent to predict potential North American distribution of the aquatic invasive macrophyte based upon training data from its native range. We produced a model using all available native range occurrence data, then explored the change in model performance produced by omitting subsets of training data based on political boundaries. We also compared those results with models trained on data from which a random sample of occurrence data was omitted from across the native range. Although most models accurately predicted the occurrence of H. verticillata in North America (AUC > 0.7600), data omissions influenced model predictions. Omitting data based on political boundaries resulted in larger shifts in model accuracy than omitting randomly selected occurrence data. For well‐documented species like H. verticillata, missing records from single countries or ecoregions may minimally influence model predictions, but for species with fewer documented occurrences or poorly understood ranges, geographic biases could misguide predictions. Regardless of focal species, we recommend that future species distribution modeling efforts begin with a reflection on potential spatial biases of available occurrence data. Improved biodiversity surveillance and reporting will provide benefit not only in invaded ranges but also within under‐reported and unexplored native ranges.     

Differences in spatial predictions among species distribution modeling methods vary with species traits and environmental predictors

Prediction maps produced by species distribution models (SDMs) influence decision‐making in resource management or designation of land in conservation planning. Many studies have compared the prediction accuracy of different SDM modeling methods, but few have quantified the similarity among prediction maps. There has also been little systematic exploration of how the relative importance of different predictor variables varies among model types and affects map similarity. Our objective was to expand the evaluation of SDM performance for 45 plant species in southern California to better understand how map predictions vary among model types, and to explain what factors may affect spatial correspondence, including the selection and relative importance of different environmental variables. Four types of models were tested. Correlation among maps was highest between generalized linear models (GLMs) and generalized additive models (GAMs) and lowest between classification trees and GAMs or GLMs. Correlation between Random Forests (RFs) and GAMs was the same as between RFs and classification trees. Spatial correspondence among maps was influenced the most by model prediction accuracy (AUC) and species prevalence; map correspondence was highest when accuracy was high and prevalence was intermediate (average prevalence for all species was 0.124). Species functional type and the selection of climate variables also influenced map correspondence. For most (but not all) species, climate variables were more important than terrain or soil in predicting their distributions. Environmental variable selection varied according to modeling method, but the largest differences were between RFs and GLMs or GAMs. Although prediction accuracy was equal for GLMs, GAMs, and RFs, the differences in spatial predictions suggest that it may be important to evaluate the results of more than one model to estimate the range of spatial uncertainty before making planning decisions based on map outputs. This may be particularly important if models have low accuracy or if species prevalence is not intermediate.     

How to assess the prediction accuracy of species presence–absence models without absence data?

It is very common that only presence data are available in ecological niche modeling. However, most existing methods for evaluating the accuracy of presence–absence (binary) predictions of species require presence–absence data. The aim of this study is to present a new method for accuracy assessment that does not rely on absence data. Two new statistics F_pb and F_cpb were derived based on presence–background data. With generated six virtual species, we used DOMAIN, generalized linear modeling (GLM), and maximum entropy (MAXENT) to produce different species presence–absence predictions. To investigate the effectiveness of the new statistics in accuracy assessment, we used F_pb, F_cpb, the traditional F‐measure (F), kappa coefficient, true skill statistic (TSS), area under the receiver operating characteristic curve (AUC), and the contrast validation index (CVI) to evaluate the accuracy of predictions, and the behaviors of these accuracy measures were compared. The effectiveness of F_pb for threshold selection and estimation of species prevalence was also investigated. Experimental results show that F_cpb is an estimate of F. The Pearson's correlation coefficient (COR) between F_cpb and F is 0.9882, with a root‐mean‐square error (RMSE) of 0.0171. In general, F_pb, F_cpb, F, kappa coefficient, TSS, and CVI can sort models by the accuracy of binary prediction, but AUC is not appropriate to evaluate the accuracy of binary prediction. For DOMAIN, GLM, and MAXENT, finding the threshold by maximizing F_pb and by maximizing F result in similar accuracies. In addition, the estimation of species prevalence based on binary output with maximizing F_pb as the thresholding method is significantly more accurate than simply averaging the original continuous output. The best estimate of prevalence is provided by the binary output of MAXENT, with an RMSE of 0.0116. Finally, we conclude that the new method is promising in accuracy assessment, threshold selection, and estimation of species prevalence, all of which are important but challenging problems with presence‐only data. Because it does not require absence data, the new method will have important applications in ecological niche modeling.     

Inferring macro‐ecological patterns from local presence/absence data

Biodiversity provides support for life, vital provisions, regulating services and has positive cultural impacts. It is therefore important to have accurate methods to measure biodiversity, in order to safeguard it when we discover it to be threatened. For practical reasons, biodiversity is usually measured at fine scales whereas diversity issues (e.g. conservation) interest regional or global scales. Moreover, biodiversity may change across spatial scales. It is therefore a key challenge to be able to translate local information on biodiversity into global patterns. Many databases give no information about the abundances of a species within an area, but only its occurrence in each of the surveyed plots. In this paper, we introduce an analytical framework (implemented in a ready‐to‐use R code) to infer species richness and abundances at large spatial scales in biodiversity‐rich ecosystems when species presence/absence information is available on various scattered samples (i.e. upscaling). This framework is based on the scale‐invariance property of the negative binomial. Our approach allows to infer and link within a unique framework important and well‐known biodiversity patterns of ecological theory, such as the species accumulation curve (SAC) and the relative species abundance (RSA) as well as a new emergent pattern, which is the relative species occupancy (RSO). Our estimates are robust and accurate, as confirmed by tests performed on both in silico‐generated and real forests. We demonstrate the accuracy of our predictions using data from two well‐studied forest stands. Moreover, we compared our results with other popular methods proposed in the literature to infer species richness from presence to absence data and we showed that our framework gives better estimates. It has thus important applications to biodiversity research and conservation practice.     

Performance tradeoffs in target‐group bias correction for species distribution models

Species distribution models (SDMs) are often calibrated using presence‐only datasets plagued with environmental sampling bias, which leads to a decrease of model accuracy. In order to compensate for this bias, it has been suggested that background data (or pseudoabsences) should represent the area that has been sampled. However, spatially‐explicit knowledge of sampling effort is rarely available. In multi‐species studies, sampling effort has been inferred following the target‐group (TG) approach, where aggregated occurrence of TG species informs the selection of background data. However, little is known about the species‐ specific response to this type of bias correction. The present study aims at evaluating the impacts of sampling bias and bias correction on SDM performance. To this end, we designed a realistic system of sampling bias and virtual species based on 92 terrestrial mammal species occurring in the Mediterranean basin. We manipulated presence and background data selection to calibrate four SDM types. Unbiased (unbiased presence data) and biased (biased presence data) SDMs were calibrated using randomly distributed background data. We used real and TG‐estimated sampling efforts in background selection to correct for sampling bias in presence data. Overall, environmental sampling bias had a deleterious effect on SDM performance. In addition, bias correction improved model accuracy, and especially when based on spatially‐explicit knowledge of sampling effort. However, our results highlight important species‐specific variations in susceptibility to sampling bias, which were largely explained by range size: widely‐distributed species were most vulnerable to sampling bias and bias correction was even detrimental for narrow‐ranging species. Furthermore, spatial discrepancies in SDM predictions suggest that bias correction effectively replaces an underestimation bias with an overestimation bias, particularly in areas of low sampling intensity. Thus, our results call for a better estimation of sampling effort in multispecies system, and cautions the uninformed and automatic application of TG bias correction.     

Keep collecting: accurate species distribution modelling requires more collections than previously thought

Aim Species distribution models (SDMs) use the locations of collection records to map the distributions of species, making them a powerful tool in conservation biology, ecology and biogeography. However, the accuracy of range predictions may be reduced by temporally autocorrelated biases in the data. We assess the accuracy of SDMs in predicting the ranges of tropical plant species on the basis of different sample sizes while incorporating real‐world collection patterns and biases. Location Tropical South American moist forests. Methods We use dated herbarium records to model the distributions of 65 Amazonian and Andean plant species. For each species, we use the first 25, 50, 100, 125 and 150 records collected and available for each species to analyse changes in spatial aggregation and climatic representativeness through time. We compare the accuracy of SDM range estimates produced using the time‐ordered data subsets to the accuracy of range estimates generated using the same number of collections but randomly subsampled from all available records. Results We find that collections become increasingly aggregated through time but that additional collecting sites are added resulting in progressively better representations of the species’ full climatic niches. The range predictions produced using time‐ordered data subsets are less accurate than predictions from random subsets of equal sample sizes. Range predictions produced using time‐ordered data subsets consistently underestimate the extent of ranges while no such tendency exists for range predictions produced using random data subsets. Main conclusions These results suggest that larger sample sizes are required to accurately map species ranges. Additional attention should be given to increasing the number of records available per species through continued collecting, better distributed collecting, and/or increasing access to existing collections. The fact that SDMs generally under‐predict the extent of species ranges means that extinction risks of species because of future habitat loss may be lower than previously estimated.     

Minimum required number of specimen records to develop accurate species distribution models

Species distribution models (SDMs) are widely used to predict the occurrence of species. Because SDMs generally use presence‐only data, validation of the predicted distribution and assessing model accuracy is challenging. Model performance depends on both sample size and species’ prevalence, being the fraction of the study area occupied by the species. Here, we present a novel method using simulated species to identify the minimum number of records required to generate accurate SDMs for taxa of different pre‐defined prevalence classes. We quantified model performance as a function of sample size and prevalence and found model performance to increase with increasing sample size under constant prevalence, and to decrease with increasing prevalence under constant sample size. The area under the curve (AUC) is commonly used as a measure of model performance. However, when applied to presence‐only data it is prevalence‐dependent and hence not an accurate performance index. Testing the AUC of an SDM for significant deviation from random performance provides a good alternative. We assessed the minimum number of records required to obtain good model performance for species of different prevalence classes in a virtual study area and in a real African study area. The lower limit depends on the species’ prevalence with absolute minimum sample sizes as low as 3 for narrow‐ranged and 13 for widespread species for our virtual study area which represents an ideal, balanced, orthogonal world. The lower limit of 3, however, is flawed by statistical artefacts related to modelling species with a prevalence below 0.1. In our African study area lower limits are higher, ranging from 14 for narrow‐ranged to 25 for widespread species. We advocate identifying the minimum sample size for any species distribution modelling by applying the novel method presented here, which is applicable to any taxonomic clade or group, study area or climate scenario.     

Prediction of plant species distributions across six millennia

The usefulness of species distribution models (SDMs) in predicting impacts of climate change on biodiversity is difficult to assess because changes in species ranges may take decades or centuries to occur. One alternative way to evaluate the predictive ability of SDMs across time is to compare their predictions with data on past species distributions. We use data on plant distributions, fossil pollen and current and mid-Holocene climate to test the ability of SDMs to predict past climate-change impacts. We find that species showing little change in the estimated position of their realized niche, with resulting good model performance, tend to be dominant competitors for light. Different mechanisms appear to be responsible for among-species differences in model performance. Confidence in predictions of the impacts of climate change could be improved by selecting species with characteristics that suggest little change is expected in the relationships between species occurrence and climate patterns.     

Assessing the effect of prevalence on the predictive performance of species distribution models using simulated data

Aim The proportion of sampled sites where a species is present is known as prevalence. Empirical studies have shown that prevalence can affect the predictive performance of species distribution models. This paper uses simulated species data to examine how prevalence and the form of species environmental dependence affect the assessment of the predictive performance of models. Methods Simulated species data were based on various functions of simulated environmental data with differing degrees of spatial correlation. Seven model performance measures – sensitivity, specificity, class‐average (CA), overall prediction success, kappa (κ), normalized mutual information (NMI) and area under the receiver operating characteristic curve (AUC) – were applied to species models fitted by three regression methods. The response of the performance measures to prevalence was then assessed. Three probability threshold selection methods used to convert fitted logistic model values to presence or absence were also assessed. Results The study shows that the extent to which prevalence affects model performance depends on the modelling technique and its degree of success in capturing dominant environmental determinants. It also depends on the statistic used to measure model performance and the probability threshold method. The response based on κ generally preferred models with medium prevalence. All performance measures were least affected by prevalence when the probability threshold was chosen to maximize predictive performance or was based directly on prevalence. In these cases, the responses based on AUC, CA and NMI generally preferred models with small or large prevalence. Main conclusions The effect of prevalence on the predictive performance of species distribution models has a methodological basis. Relevant factors include the success of the fitted distribution model in capturing the dominant environmental determinant, the model performance measure and the probability threshold selection method. The fixed probability threshold method yields a marked response of model performance to prevalence and is therefore not recommended. The study explains previous empirical results obtained with real data.     

Effects of the number of presences on reliability and stability of MARS species distribution models: the importance of regional niche variation and ecological heterogeneity

Question: What are the effects of the number of presences on models generated with multivariate adaptive regression splines (MARS)? Do these effects vary with data quality and quantity and species ecology? Location: Spain and Ecuador. Methods: We used two data sets: (1) two trees from Spain, representing high‐occurrence number data sets with real absences and unbalanced prevalence; (2) two herbs from Ecuador, representing low‐occurrence number data sets without real absences and balanced prevalence. For model quality, we used two different measures: reliability and stability. For each sample size, different replicates were generated at random and then used to generate a consensus model. Results: Model reliability and stability decrease with sample size. Optimal minimum sample size varies depending on many factors, many of which are unknown. Regional niche variation and ecological heterogeneity are critical. Conclusions: (1) Model predictive power improves greatly with more than 18‐20 presences. (2) Model reliability depends on data quantity and quality as well as species ecological characteristics. (3) Depending on the number of presences in the data set, investigators must carefully distinguish between models that should be treated with skepticism and those whose predictions can be applied with reasonable confidence. (4) For species combining few initial presences and wide environmental range variation, it is advisable to generate several replicate models that partition the initial data and generate a consensus model. (5) Models of species with a narrow environmental range variation can be highly stable and reliable, even when generated with few presences.     

Model‐based integration of observed and expert‐based information for assessing the geographic and environmental distribution of freshwater species

Freshwater ecosystems harbor specialized and vulnerable biodiversity, and the prediction of potential impacts of freshwater biodiversity to environmental change requires knowledge of the geographic and environmental distribution of taxa. To date, however, such quantitative information about freshwater species distributions remains limited. Major impediments include heterogeneity in available species occurrence data, varying detectability of species in their aquatic environment, scarcity of contiguous freshwater‐specific predictors, and methods that support addressing these issues in a single framework. Here we demonstrate the use of a hierarchical Bayesian modeling (HBM) framework that combines disparate species occurrence information with newly‐developed 1 km freshwater‐specific predictors, to account for imperfect species detection and make fine‐grain (1 km) estimates of distributions in freshwater organisms. The approach integrates a Bernoulli suitability and a Binomial observability process into a hierarchical zero‐inflated Binomial model. The suitability process includes point presence observations, records of site visits, 1 km environmental predictors and expert‐derived species range maps integrated with a distance‐decay function along the within‐stream distance as covariates. The observability process uses repeated observations to estimate a probability of observation given that the species was present. The HBM accounts for the spatial autocorrelation in species habitat suitability projections using an intrinsic Gaussian conditional autoregressive model. We used this framework for three fish species native to different regions and habitats in North America. Model comparison shows that HBMs significantly outperformed non‐spatial GLMs in terms of AUC and TSS scores, and that expert information when appropriately included in the model can provide an important refinement. Such ancillary species information and an integrative, hierarchical Bayesian modeling framework can therefore be used to advance fine‐grain habitat suitability predictions and range size estimates in the freshwater realm. Our approach is extendable in terms of data availability and generality and can be used on other freshwater organisms and regions.     

A review of the occurrence of inter‐colony segregation of seabird foraging areas and the implications for marine environmental impact assessment

Understanding the determinants of species’ distributions is a fundamental aim in ecology and a prerequisite for conservation but is particularly challenging in the marine environment. Advances in bio‐logging technology have resulted in a rapid increase in studies of seabird movement and distribution in recent years. Multi‐colony studies examining the effects of intra‐ and inter‐colony competition on distribution have found that several species exhibit inter‐colony segregation of foraging areas, rather than overlapping distributions. These findings are timely given the increasing rate of human exploitation of marine resources and the need to make robust assessments of likely impacts of proposed marine developments on biodiversity. Here we review the occurrence of foraging area segregation reported by published tracking studies in relation to the density‐dependent hinterland (DDH) model, which predicts that segregation occurs in response to inter‐colony competition, itself a function of colony size, distance from the colony and prey distribution. We found that inter‐colony foraging area segregation occurred in 79% of 39 studies. The frequency of occurrence was similar across the four seabird orders for which data were available, and included species with both smaller (10–100 km) and larger (100–1000 km) foraging ranges. Many predictions of the DDH model were confirmed, with examples of segregation in response to high levels of inter‐colony competition related to colony size and proximity, and enclosed landform restricting the extent of available habitat. Moreover, as predicted by the DDH model, inter‐colony overlap tended to occur where birds aggregated in highly productive areas, often remote from all colonies. The apparent prevalence of inter‐colony foraging segregation has important implications for assessment of impacts of marine development on protected seabird colonies. If a development area is accessible from multiple colonies, it may impact those colonies much more asymmetrically than previously supposed. Current impact assessment approaches that do not consider spatial inter‐colony segregation will therefore be subject to error. We recommend the collection of tracking data from multiple colonies and modelling of inter‐colony interactions to predict colony‐specific distributions.