General methods for sensitivity analysis of equilibrium dynamics in patch occupancy models

Sensitivity analysis is a useful tool for the study of ecological models that has many potential applications for patch occupancy modeling. Drawing from the rich foundation of existing methods for Markov chain models, I demonstrate new methods for sensitivity analysis of the equilibrium state dynamics of occupancy models. Estimates from three previous studies are used to illustrate the utility of the sensitivity calculations: a joint occupancy model for a prey species, its predators, and habitat used by both; occurrence dynamics from a well-known metapopulation study of three butterfly species; and Golden Eagle occupancy and reproductive dynamics. I show how to deal efficiently with multistate models and how to calculate sensitivities involving derived state variables and lower-level parameters. In addition, I extend methods to incorporate environmental variation by allowing for spatial and temporal variability in transition probabilities. The approach used here is concise and general and can fully account for environmental variability in transition parameters. The methods can be used to improve inferences in occupancy studies by quantifying the effects of underlying parameters, aiding prediction of future system states, and identifying priorities for sampling effort.     

Predictive vegetation modeling for conservation: impact of error propagation from digital elevation data.

The effect of digital elevation model (DEM) error on environmental variables, and subsequently on predictive habitat models, has not been explored. Based on an error analysis of a DEM, multiple error realizations of the DEM were created and used to develop both direct and indirect environmental variables for input to predictive habitat models. The study explores the effects of DEM error and the resultant uncertainty of results on typical steps in the modeling procedure for prediction of vegetation species presence/absence. Results indicate that all of these steps and results, including the statistical significance of environmental variables, shapes of species response curves in generalized additive models (GAMs), stepwise model selection, coefficients and standard errors for generalized linear models (GLMs), prediction accuracy (Cohen's kappa and AUC), and spatial extent of predictions, were greatly affected by this type of error. Error in the DEM can affect the reliability of interpretations of model results and level of accuracy in predictions, as well as the spatial extent of the predictions. We suggest that the sensitivity of DEM-derived environmental variables to error in the DEM should be considered before including them in the modeling processes.     

Predicting abundance of desert riparian birds: validation and calibration of the Effective Area Model.

Reliable prediction of the effects of landscape change on species abundance is critical to land managers who must make frequent, rapid decisions with long-term consequences. However, due to inherent temporal and spatial variability in ecological systems, previous attempts to predict species abundance in novel locations and/or time frames have been largely unsuccessful. The Effective Area Model (EAM) uses change in habitat composition and geometry coupled with response of animals to habitat edges to predict change in species abundance at a landscape scale. Our research goals were to validate EAM abundance predictions in new locations and to develop a calibration framework that enables absolute abundance predictions in novel regions or time frames. For model validation, we compared the EAM to a null model excluding edge effects in terms of accurate prediction of species abundance. The EAM outperformed the null model for 83.3% of species (N=12) for which it was possible to discern a difference when considering 50 validation sites. Likewise, the EAM outperformed the null model when considering subsets of validation sites categorized on the basis of four variables (isolation, presence of water, region, and focal habitat). Additionally, we explored a framework for producing calibrated models to decrease prediction error given inherent temporal and spatial variability in abundance. We calibrated the EAM to new locations using linear regression between observed and predicted abundance with and without additional habitat covariates. We found that model adjustments for unexplained variability in time and space, as well as variability that can be explained by incorporating additional covariates, improved EAM predictions. Calibrated EAM abundance estimates with additional site-level variables explained a significant amount of variability (P < 0.05) in observed abundance for 17 of 20 species, with R2 values >25% for 12 species, >48% for six species, and >60% for four species when considering all predictive models. The calibration framework described in this paper can be used to predict absolute abundance in sites different from those in which data were collected if the target population of sites to which one would like to statistically infer is sampled in a probabilistic way.     

Area and mammalian elevational diversity

Elevational gradients hold enormous potential for understanding general properties of biodiversity. Like latitudinal gradients, the hypotheses for diversity patterns can be grouped into historical explanations, climatic drivers, and spatial hypotheses. The spatial hypotheses include the species-area effect and spatial constraint (mid-domain effect null models). I test these two spatial hypotheses using regional diversity patterns for mammals (non-volant small mammals and bats) along 34 elevational gradients spanning 24.4 degrees S-40.4 degrees N latitude. There was high variability in the fit to the species-area hypothesis and the mid-domain effect. Both hypotheses can be eliminated as primary drivers of elevational diversity. Area and spatial constraint both represent sources of error rather than mechanisms underlying these mammalian diversity patterns. Similar results are expected for other vertebrate taxa, plants, and invertebrates since they show comparable distributions of elevational diversity patterns to mammalian patterns.     

The importance of spatial autocorrelation,extent and resolution in predicting forest bird occurrence

Concerns about declines in forest biodiversity underscore the need for accurate estimates of the distribution and abundance of organisms at large scales and at resolutions that are fine enough to be appropriate for management. This paper addresses three major objectives: (i) to determine whether the resolution of typical air photo-derived forest inventory is sufficient for the accurate prediction of site occupancy by forest birds. We compared prediction success of habitat models using air photo variables to models with variables derived from finer resolution, ground-sampled vegetation plots. (ii) To test whether incorporating spatial autocorrelation into habitat models via autologistic regression increases prediction success. (iii) To determine whether landscape structure is an important factor in predicting bird distribution in forest-dominated landscapes. Models were tested locally (Greater Fundy Ecosystem [GFE]) using cross-validation, and regionally using an independent data set from an area located ca. 250 km to the northwest (Riley Brook [RB]). We found significant positive spatial autocorrelation in the residuals of at least one habitat model for 76% (16/21) of species examined. In these cases, the logistic regression assumption of spatially independent errors was violated. Logistic models that ignored spatial autocorrelation tended to overestimate habitat effects. Though overall prediction success was higher for autologistic models than logistic models in the GFE, the difference was only significantly improved for one species. Further, the inclusion of spatial covariates did little to improve model performance in the geographically discrete study area. For 62% (13/21) of species examined, landscape variables were significant predictors of forest bird occurrence even after statistically controlling for stand-level variability. However, broad spatial extents explained less variation than local factors. In the GFE, 76% (16/21) of air photo and 81% (17/21) of ground plot models were accurate enough to be of practical utility (AUC > 0.7). When applied to RB, both model types performed effectively for 55% (11/20) of the species examined. We did not detect an overall difference in prediction success between air photo and ground plot models in either study area. We conclude that air photo data are as effective as fine resolution vegetation data for predicting site occupancy for the majority of species in this study. These models will be of use to forest managers who are interested in mapping species distributions under various timber harvest scenarios, and to protected areas planners attempting to optimize reserve function.     

Building statistical models to analyze species distributions.

Models of the geographic distributions of species have wide application in ecology. But the nonspatial, single-level, regression models that ecologists have often employed do not deal with problems of irregular sampling intensity or spatial dependence, and do not adequately quantify uncertainty. We show here how to build statistical models that can handle these features of spatial prediction and provide richer, more powerful inference about species niche relations, distributions, and the effects of human disturbance. We begin with a familiar generalized linear model and build in additional features, including spatial random effects and hierarchical levels. Since these models are fully specified statistical models, we show that it is possible to add complexity without sacrificing interpretability. This step-by-step approach, together with attached code that implements a simple, spatially explicit, regression model, is structured to facilitate self-teaching. All models are developed in a Bayesian framework. We assess the performance of the models by using them to predict the distributions of two plant species (Proteaceae) from South Africa's Cape Floristic Region. We demonstrate that making distribution models spatially explicit can be essential for accurately characterizing the environmental response of species, predicting their probability of occurrence, and assessing uncertainty in the model results. Adding hierarchical levels to the models has further advantages in allowing human transformation of the landscape to be taken into account, as well as additional features of the sampling process.     

Relationship among the species richness of different taxa

Spatially explicit forecasting of changes in species richness is key to designing informative scenarios on the development of diversity on our planet. It might be possible to predict changes in the richness of inadequately investigated groups from that of groups for which enough information is available. Here we evaluate the reliability of this approach by reviewing 237 richness correlations extracted from the recent literature. Of the 43 taxa covered, beetles, vascular plants, butterflies, birds, ants, and mammals (in that order) were the most common ones examined. Forests and grasslands strongly dominated the ecosystem types studied. The variance explanation (R2) could be calculated for 152 cases, but only 53 of these were significant. An average correlation effect size of 0.374 (95% CI = +/- 0.0678) indicates positive but weak correlations between taxa within the very heterogeneous data set; None of the examined explanatory variables (spatial scale, taxonomic distance, trophic position, biome) could account for this heterogeneity. However, studies focusing on 10-km2 grid cells had the highest variance explanation. Moreover, within-phylum between-class comparisons had marginally significantly lower correlations than between-phylum comparisons. And finally, the explanatory power of studies conducted in the tropics was significantly higher than that of studies conducted in temperate regions. It is concluded that the potential of a correlative approach to species richness is strongly diminished by the overall low level of variance explanation. So far, no taxon has proved to be a universal or even particularly good predictor for the richness of other taxa. Some suggestions for future research are inclusion of several taxa in models aiming at regional richness predictions, improvement of knowledge on species correlations in human dominated systems, and a better understanding of mechanisms underlying richness correlations.     

Species Richness, Endemism, and the Choice of Areas for Conservation

Although large reserve networks will be integral components in successful biodiversity conservation, implementation of such systems is hindered by the confusion over the relative importance of endemism and species richness. There is evidence (  Prendergast et al. 1993) that regions with high richness for a taxon tend to be different from those with high endemism. I tested this finding using distribution and richness data for 368 species from Mammalia, Lasioglossum, Plusiinae, and Papilionidae. The study area, subdivided into 336 quadrats, was the continuous area of North America north of Mexico. I also tested the hypothesis that the study taxa exhibit similar diversity patterns in North America. I found that endemism and richness patterns within taxa were generally similar. Therefore, the controversy over the relative importance of endemism and species richness may not be necessary if an individual taxon were the target of conservation efforts. I also found, however, that richness and endemism patterns were not generally similar between taxa. Therefore, centering nature reserves around areas that are important for mammal diversity (the umbrella species concept) may lead to large gaps in the overall protection of biodiversity because the diversity and endemism of other taxa tend to be concentrated elsewhere. I investigated this further by selecting four regions in North America that might form the basis of a hypothetical reserve system for Carnivora. I analyzed the distribution of the invertebrate taxa relative to these regions and found that this preliminary carnivore reserve system did not provide significantly different protection for these invertebrates than randomly selected quadrats. I conclude that the use of Carnivora as an umbrella taxon is an unreliable method for invertebrate conservation.     

Detecting Biodiversity Hotspots by Species-Area Relationships: a Case Study of Mediterranean Beetles

Abstract:  Any method of identifying hotspots should take into account the effect of area on species richness. I examined the importance of the species-area relationship in determining tenebrionid (Coleoptera: Tenebrionidae) hotspots on the Aegean Islands (Greece). Thirty-two islands and 170 taxa (species and subspecies) were included in this study. I tested several species-area relationship models with linear and nonlinear regressions, including power, exponential, negative exponential, logistic, Gompertz, Weibull, Lomolino, and He-Legendre functions. Islands with positive residuals were identified as hotspots. I also analyzed the values of the C parameter of the power function and the simple species-area ratios. Species richness was significantly correlated with island area for all models. The power function model was the most convenient one. Most functions, however, identified certain islands as hotspots. The importance of endemics in insular biotas should be evaluated carefully because they are of high conservation concern. The simple use of the species-area relationship can be problematic when areas with no endemics are included. Therefore the importance of endemics should be evaluated according to different methods, such as percentages, to take into account different levels of endemism and different kinds of "endemics" (e.g., endemic to single islands vs. endemic to the archipelago). Because the species-area relationship is a key pattern in ecology, my findings can be applied at broader scales.     

Modeling spatial patterns in fisheries bycatch: improving bycatch maps to aid fisheries management.

Fisheries bycatch, or incidental take, of large vertebrates such as sea turtles, seabirds, and marine mammals, is a pressing conservation and fisheries management issue. Identifying spatial patterns of bycatch is an important element in managing and mitigating bycatch occurrences. Because bycatch of these taxa involves rare events and fishing effort is highly variable in space and time, maps of raw bycatch rates (the ratio of bycatch to fishing effort) can be misleading. Here we show how mapping bycatch can be enhanced through the use of Bayesian hierarchical spatial models. We compare model-based estimates of bycatch rates to raw rates. The model-based estimates were more precise and fit the data well. Using these results, we demonstrate the utility of this approach for providing information to managers on bycatch probabilities and cross-taxa bycatch comparisons. To illustrate this approach, we present an analysis of bycatch data from the U.S. gill net fishery for groundfish in the northwest Atlantic. The goals of this analysis are to produce more reliable estimates of bycatch rates, assess similarity of spatial patterns between taxa, and identify areas of elevated risk of bycatch.     

Hierarchical Bayesian Spatial Models for Multispecies Conservation Planning and Monitoring

Abstract: Biologists who develop and apply habitat models are often familiar with the statistical challenges posed by their data's spatial structure but are unsure of whether the use of complex spatial models will increase the utility of model results in planning. We compared the relative performance of nonspatial and hierarchical Bayesian spatial models for three vertebrate and invertebrate taxa of conservation concern (Church's sideband snails [Monadenia churchi], red tree voles [Arborimus longicaudus], and Pacific fishers [Martes pennanti pacifica]) that provide examples of a range of distributional extents and dispersal abilities. We used presence–absence data derived from regional monitoring programs to develop models with both landscape and site‐level environmental covariates. We used Markov chain Monte Carlo algorithms and a conditional autoregressive or intrinsic conditional autoregressive model framework to fit spatial models. The fit of Bayesian spatial models was between 35 and 55% better than the fit of nonspatial analogue models. Bayesian spatial models outperformed analogous models developed with maximum entropy (Maxent) methods. Although the best spatial and nonspatial models included similar environmental variables, spatial models provided estimates of residual spatial effects that suggested how ecological processes might structure distribution patterns. Spatial models built from presence–absence data improved fit most for localized endemic species with ranges constrained by poorly known biogeographic factors and for widely distributed species suspected to be strongly affected by unmeasured environmental variables or population processes. By treating spatial effects as a variable of interest rather than a nuisance, hierarchical Bayesian spatial models, especially when they are based on a common broad‐scale spatial lattice (here the national Forest Inventory and Analysis grid of 24 km² hexagons), can increase the relevance of habitat models to multispecies conservation planning.     

Breeding systems and seed size in a neotropical flora: testing evolutionary hypotheses

A well-known, but largely untested, prediction in plant reproductive ecology is that dioecious taxa should produce larger, more, higher-quality, or better-defended seeds than cosexual taxa. Using a data set composed of 972 species in 104 families, representing the flora of the Tambopata Wildlife Reserve (Madre de Dios, Peru), we evaluated the first component of this prediction, examining ecological and evolutionary relationships between breeding system and mean seed size with two kinds of tests. First, we conducted cross-species analyses to determine whether species with different breeding systems differed significantly with respect to mean individual seed size. Second, we used a hypothesized phylogeny to identify pairs of the most closely related taxa or clades within the Tambopata community that differed with respect to breeding system. Comparing pair members allowed us to determine whether evolutionary divergence in breeding system (between taxa with unisexual vs. cosexual individuals) was consistently associated with evolutionary change in seed size. In both analyses, we controlled for potentially confounding effects of growth form by examining these relationships within woody and nonwoody taxa. Cross-species analyses revealed that dioecious species produced larger seeds than cosexual species among woody species, shrubs, lianas (each growth form analyzed separately), and all species pooled, but not among trees. Phylogenetically independent contrasts upheld the significant association between breeding system and seed size among woody taxa, lianas, and all taxa pooled, but not among shrubs. We discuss the implications of our findings for evolutionary hypotheses regarding associations between dioecy and seed size.     

Mapping shifts in spatial synchrony in grassland birds to inform conservation planning

Spatial synchrony, defined as the correlated fluctuations in abundance of spatially separated populations, can be caused by regional fluctuations in natural and anthropogenic environmental population drivers. Investigations into the geography of synchrony can provide useful insight to inform conservation planning efforts by revealing regions of common population drivers and metapopulation extinction vulnerability. We examined the geography of spatial synchrony and decadal changes in these patterns for grassland birds in the United States and Canada, which are experiencing widespread and persistent population declines. We used Bayesian hierarchical models and over 50 years of abundance data from the North American Breeding Bird Survey to generate population indices within a 2° latitude by 2° longitude grid. We computed and mapped mean local spatial synchrony for each cell (mean detrended correlation of the index among neighboring cells), along with associated uncertainty, for 19 species in 2, 26-year periods, 1968–1993 and 1994–2019. Grassland birds were predicted to increase in spatial synchrony where agricultural intensification, climate change, or interactions between the 2 increased. We found no evidence of an overall increase in synchrony among grassland bird species. However, based on the geography of these changes, there was considerable spatial heterogeneity within species. Averaging across species, we identified clusters of increasing spatial synchrony in the Prairie Pothole and Shortgrass Prairie regions and a region of decreasing spatial synchrony in the eastern United States. Our approach has the potential to inform continental-scale conservation planning by adding an additional layer of relevant information to species status assessments and spatial prioritization of policy and management actions. Our work adds to a growing literature suggesting that global change may result in shifting patterns of spatial synchrony in population dynamics across taxa with broad implications for biodiversity conservation.     

Interhemispheric comparison of recruitment to intertidal communities: pattern persistence and scales of variation

Recruitment variation can be a major source of fluctuation in populations and communities, making it difficult to generalize results. Determining the scales of variation and whether spatial patterns in the supply of individuals are persistent over time can provide insight into spatial generality and the application of conservation and metacommunity models. We examined these issues using eight-year-long data sets of monthly recruitment of intertidal mussels (Mytilus spp., Perumytilus purpuratus, Semimytilus algosus, Brachidontes granulata) and barnacles (Balanus glandula, Chthamalus dalli, Jehlius cirratus, Notochthamalus scabrosus) at sites spanning > 900 km along the coasts of Oregon-northern California (OR-NCA, 45.47-39.43 degrees N) and central Chile (CC, 29.5-34.65 degrees S). We evaluated four general "null" hypotheses: that despite different phylogenies and great spatial separation of these taxa, their similar life history strategies and environmental settings lead to similar patterns of recruitment (1) between hemispheres, (2) in time, (3) in space, and (4) at larger and smaller spatial scales. Hypothesis 1 was rejected: along the OR-NCA coast, rates of recruitment were between two and three orders of magnitude higher, and patterns of seasonality were generally stronger and more coherent across space and time than along CC. Surprisingly, however, further analysis revealed regularities in both time and space for all species, supporting hypotheses 2 and 3. Temporal decorrelation scales were 1-3 months, and characteristic spatial scales of recruitment were approximately 250 km. Contrary to hypothesis 4, for the ecologically dominant species in both hemispheres, recruitment was remarkably persistent at larger mesoscales (kilometers) but was highly stochastic at smaller microscales (meters). Across species, increased recruitment variation at large scales was positively associated with increased persistence. Our results have several implications. Although the two regions span distinct latitudinal ranges, potential forcing processes behind these patterns include similar large-scale climates and topographically locked hydrographic features, such as upwelling. Further, spatial persistence of the recruitment patterns of most species at the mesoscale supports the view that marine protected areas can be powerful conservation and management tools. Finally, persistent and yet contrasting spatial patterns of recruitment among competing species suggest that recent metacommunity models might provide useful representations of the mechanisms involved in species coexistence.     

Recovery of Faunal Communities During Tropical Forest Regeneration

Abstract:  As mature tropical forests are cleared, secondary forests may play an important role in the conservation of animal species, depending on how fast animal communities recover during forest regeneration. I reviewed published studies on the recovery of animal species richness and composition during tropical forest regeneration. In 38 of the 39 data sets I examined, conversion of forest to agriculture or pasture substantially reduced species richness. Given suitable conditions for forest recovery, the species richness of the animal taxa considered can be predicted to resemble that of mature forests roughly 20–40 years after land abandonment. At least for ants and birds, however, recovery of species composition appears to take substantially longer than recovery of species richness. Because species richness for many taxa appears to recover relatively rapidly in secondary forests, conservation of secondary forests may be an effective investment in future diversity. The slower recovery of species composition indicates, however, that some species will require stands of mature forest to persist.     

Impacts of phylogenetic nomenclature on the efficacy of the U.S. Endangered Species Act

Cataloging biodiversity is critical to conservation efforts because accurate taxonomy is often a precondition for protection under laws designed for species conservation, such as the U.S. Endangered Species Act (ESA). Traditional nomenclatural codes governing the taxonomic process have recently come under scrutiny because taxon names are more closely linked to hierarchical ranks than to the taxa themselves. A new approach to naming biological groups, called phylogenetic nomenclature (PN), explicitly names taxa by defining their names in terms of ancestry and descent. PN has the potential to increase nomenclatural stability and decrease confusion induced by the rank‐based codes. But proponents of PN have struggled with whether species and infraspecific taxa should be governed by the same rules as other taxa or should have special rules. Some proponents advocate the wholesale abandonment of rank labels (including species); this could have consequences for the implementation of taxon‐based conservation legislation. I examined the principles of PN as embodied in the PhyloCode (an alternative to traditional rank‐based nomenclature that names biological groups based on the results of phylogenetic analyses and does not associate taxa with ranks) and assessed how this novel approach to naming taxa might affect the implementation of species‐based legislation by providing a case study of the ESA. The latest version of the PhyloCode relies on the traditional rank‐based codes to name species and infraspecific taxa; thus, little will change regarding the main targets of the ESA because they will retain rank labels. For this reason, and because knowledge of evolutionary relationships is of greater importance than nomenclatural procedures for initial protection of endangered taxa under the ESA, I conclude that PN under the PhyloCode will have little impact on implementation of the ESA. Impactos de la Nomenclatura Filogenética sobre la Eficiencia del Acta Estadunidense para las Especies en Peligro     

Sympatric species of threespine stickleback differ in their performance in a spatial learning task

Increasing evidence suggests that cognitive function is under selection in diverse taxa and that this results in different cognitive abilities in taxa experiencing different selective environments. For example, species inhabiting spatially complex environments might be expected to have good spatial learning ability. We investigated whether local habitat conditions influence learning by comparing the performance of two recently diverged species of threespine stickleback (Gasterosteus aculeatus complex) in a spatial learning task. The two species reside sympatrically in several lakes. Benthics occupy the spatially structured vegetated littoral zone, whereas limnetics occupy the spatially homogenous open-water pelagic zone. We trained fish to locate a hidden reward in a T-maze and asked whether they did so by learning a body-centred pattern of movement (turn left or right) or by using plant landmarks. Both benthics and limnetics used turn and landmark cues, but benthics learnt the task almost twice as quickly as limnetics. This difference is consistent with the hypothesis that benthic and limnetic sticklebacks are equipped with spatial learning abilities well suited to the spatial complexity of their littoral and pelagic habitats. Our findings add to the understanding of the evolution of learning.     

Spatial elements of mortality risk in old-growth forests

For many species of long-lived organisms, such as trees, survival appears to be the most critical vital rate affecting population persistence. However, methods commonly used to quantify tree death, such as relating tree mortality risk solely to diameter growth, almost certainly do not account for important spatial processes. Our goal in this study was to detect and, if present, to quantify the relevance of such processes. For this purpose, we examined purely spatial aspects of mortality for four species, Abies concolor, Abies magnifica, Calocedrus decurrens, and Pinus lambertiana, in an old-growth conifer forest in the Sierra Nevada of California, USA. The analysis was performed using data from nine fully mapped long-term monitoring plots. In three cases, the results unequivocally supported the inclusion of spatial information in models used to predict mortality. For Abies concolor, our results suggested that growth rate may not always adequately capture increased mortality risk due to competition. We also found evidence of a facilitative effect for this species, with mortality risk decreasing with proximity to conspecific neighbors. For Pinus lambertiana, mortality risk increased with density of conspecific neighbors, in keeping with a mechanism of increased pathogen or insect pressure (i.e., a Janzen-Connell type effect). Finally, we found that models estimating risk of being crushed were strongly improved by the inclusion of a simple index of spatial proximity. Not only did spatial indices improve models, those improvements were relevant for mortality prediction. For P. lambertiana, spatial factors were important for estimation of mortality risk regardless of growth rate. For A. concolor, although most of the population fell within spatial conditions in which mortality risk was well described by growth, trees that died occurred outside those conditions in a disproportionate fashion. Furthermore, as stands of A. concolor become increasingly dense, such spatial factors are likely to become increasingly important. In general, models that fail to account for spatial pattern are at risk of failure as conditions change.     

Modeling the effect of photosynthetic vegetation properties on the NDVI--LAI relationship

Developing a relationship between the normalized difference vegetation index (NDVI) and the leaf area index (LAI) is essential to describe the pattern of spatial or temporal variation in LAI that controls carbon, water, and energy exchange in many ecosystem process models. Photosynthetic vegetation (PV) properties can affect the estimation of LAI, but no models integrate the effects of multiple species. We developed four alternative NDVI-LAI models, three of which integrate PV effects: no PV effects, leaf-level effects, canopy-level effects, and effects at both levels. The models were fit to data across the natural range of variation in NDVI for a widespread High Arctic ecosystem. The weight of evidence supported the canopy-level model (Akaike weight, wr = 0.98), which includes species-specific canopy coefficients that primarily scale fractional PV cover to LAI by accounting for the area of unexposed PV. Modeling the canopy-level effects improved prediction of LAI (R2 = 0.82) over the model with no PV effect (R2 = 0.71) across the natural range of variation in NDVI but did not affect the site-level estimate of LAI. Satellite-based methods to estimate species composition, a variable in the model, will need to be developed. We expect that including the effects of PV properties in NDVI-LAI models will improve prediction of LAI where species composition varies across space or changes over time.     

Null models reveal preferential sampling, spatial autocorrelation and overfitting in habitat suitability modelling

Nowadays, species are driven to extinction at a high rate. To reduce this rate it is important to delineate suitable habitats for these species in such a way that these areas can be suggested as conservation areas. The use of habitat suitability models (HSMs) can be of great importance for the delineation of such areas. In this study MaxEnt, a presence-only modelling technique, is used to develop HSMs for 223 nematode species of the Southern Bight of the North Sea. However, it is essential that these models are beyond discussion and they should be checked for potential errors. In this study we focused on two categories (1) errors which can be attributed to the database such as preferential sampling and spatial autocorrelation and (2) errors induced by the modelling technique such as overfitting, In order to quantify these adverse effects thousands of nulls models were created. The effect of preferential sampling (i.e. some areas where visited more frequenty than others) was investigated by comparing model outcomes based from null models sampling the actual sampling stations and null models sampling the entire mapping area (Raes and ter Steege, 2007). Overfitting is exposed by a fivefold cross-validation and the influence of spatial autocorrelation is assessed by separating test and training sets in space. Our results clearly show that all these effects are present: preferential sampling has a strong effect on the selection of non-random species models. Crossvalidation seems to have less influence on the model selection and spatial autocorrelation is also strongly present. It is clear from this study that predefined thresholds are not readily applicable to all datasets and additional tests are needed in model selection.