Predicting Species Distributions Research Articles

A semi-automated workflow for biodiversity data retrieval, cleaning, and quality control.

The compilation and cleaning of data needed for analyses and prediction of species distributions is a time consuming process requiring a solid understanding of data formats and service APIs provided by biodiversity informatics infrastructures. We designed and implemented a Taverna-based Data Refinement Workflow which integrates taxonomic data retrieval, data cleaning, and data selection into a consistent, standards-based, and effective system hiding the complexity of underlying service infrastructures. The workflow can be freely used both locally and through a web-portal which does not require additional software installations by users.

Predicting bottlenose dolphin distribution along Liguria coast (northwestern Mediterranean Sea) through different modeling techniques and indirect predictors

Habitat modeling is an important tool to investigate the quality of the habitat for a species within a certain area, to predict species distribution and to understand the ecological processes behind it. Many species have been investigated by means of habitat modeling techniques mainly to address effective management and protection policies and cetaceans play an important role in this context. The bottlenose dolphin (Tursiops truncatus) has been investigated with habitat modeling techniques since 1997. The objectives of this work were to predict the distribution of bottlenose dolphin in a coastal area through the use of static morphological features and to compare the prediction performances of three different modeling techniques: Generalized Linear Model (GLM), Generalized Additive Model (GAM) and Random Forest (RF). Four static variables were tested: depth, bottom slope, distance from 100 m bathymetric contour and distance from coast. RF revealed itself both the most accurate and the most precise modeling technique with very high distribution probabilities predicted in presence cells (90.4% of mean predicted probabilities) and with 66.7% of presence cells with a predicted probability comprised between 90% and 100%. The bottlenose distribution obtained with RF allowed the identification of specific areas with particularly high presence probability along the coastal zone; the recognition of these core areas may be the starting point to develop effective management practices to improve T. truncatus protection.

Modeling spatially varying landscape change points in species occurrence thresholds

Predicting species distributions at scales of regions to continents is often necessary, as large‐scale phenomena influence the distributions of spatially structured populations. Land use and land cover are important large‐scale drivers of species distributions, and landscapes are known to create species occurrence thresholds, where small changes in a landscape characteristic results in abrupt changes in occurrence. The value of the landscape characteristic at which this change occurs is referred to as a change point. We present a hierarchical Bayesian threshold model (HBTM) that allows for estimating spatially varying parameters, including change points. Our model also allows for modeling estimated parameters in an effort to understand large‐scale drivers of variability in land use and land cover on species occurrence thresholds. We use range‐wide detection/nondetection data for the eastern brook trout (Salvelinus fontinalis), a stream‐dwelling salmonid, to illustrate our HBTM for estimating and modeling spatially varying threshold parameters in species occurrence. We parameterized the model for investigating thresholds in landscape predictor variables that are measured as proportions, and which are therefore restricted to values between 0 and 1. Our HBTM estimated spatially varying thresholds in brook trout occurrence for both the proportion agricultural and urban land uses. There was relatively little spatial variation in change point estimates, although there was spatial variability in the overall shape of the threshold response and associated uncertainty. In addition, regional mean stream water temperature was correlated to the change point parameters for the proportion of urban land use, with the change point value increasing with increasing mean stream water temperature. We present a framework for quantify macrosystem variability in spatially varying threshold model parameters in relation to important large‐scale drivers such as land use and land cover. Although the model presented is a logistic HBTM, it can easily be extended to accommodate other statistical distributions for modeling species richness or abundance.

Understanding human--coyote encounters in urban ecosystems using citizen science data: what do socioeconomics tell us?

The coyote (Canis latrans) has dramatically expanded its range to include the cities and suburbs of the western US and those of the Eastern Seaboard. Highly adaptable, this newcomer's success causes conflicts with residents, necessitating research to understand the distribution of coyotes in urban landscapes. Citizen science can be a powerful approach toward this aim. However, to date, the few studies that have used publicly reported coyote sighting data have lacked an in-depth consideration of human socioeconomic variables, which we suggest are an important source of overlooked variation in data that describe the simultaneous occurrence of coyotes and humans. We explored the relative importance of socioeconomic variables compared to those describing coyote habitat in predicting human-coyote encounters in highly-urbanized Mecklenburg County, North Carolina, USA using 707 public reports of coyote sightings, high-resolution land cover, US Census data, and an autologistic multi-model inference approach. Three of the four socioeconomic variables which we hypothesized would have an important influence on encounter probability, namely building density, household income, and occupation, had effects at least as large as or larger than coyote habitat variables. Our results indicate that the consideration of readily available socioeconomic variables in the analysis of citizen science data improves the prediction of species distributions by providing insight into the effects of important factors for which data are often lacking, such as resource availability for coyotes on private property and observer experience. Managers should take advantage of citizen scientists in human-dominated landscapes to monitor coyotes in order to understand their interactions with humans.

Environmental niche modelling fails to predict Last Glacial Maximum refugia: niche shifts, microrefugia or incorrect palaeoclimate estimates?

AbstractAimMany predictions of responses to future climate change utilize ecological niche models (ENMs). We assess the capacity of these models to predict species distributions under conditions that differ from the current environment by testing whether they can predict past distributions of species.LocationFrom 43° S to 31° S in south‐eastern Australia (including Tasmania).MethodsWe studied three dominant tree species of temperate Australian mesic forests, Atherosperma moschatum, Eucalyptus regnans and Nothofagus cunninghamii. Phylogeographic evidence indicates that these species each survived the Last Glacial Maximum (LGM) in multiple refugia. We modelled the current distribution of each species and projected those models onto LGM climates under six palaeoclimatic scenarios. The support for phylogeographic‐based glacial refugia was estimated under each scenario using three different thresholds for inferring species presence/absence.ResultsThe LGM models under scenarios that allowed for a realistic level of rainfall failed to predict survival of the study species in refugia identified from genetic evidence, apart from those in perhumid western Tasmania.Main conclusionsCorrect prediction of nearly all modern occurrences of the species suggests that this failure of ENMs to predict refugial survival was not methodological. Rather we conclude that the existing realized niches of these species may have changed since the LGM. Such niche changes may have involved the occurrence of non‐analogue climates in the LGM and some significant alteration of fundamental niche (for at least E. regnans). Our results emphasize that predictions of future impacts of climate change on biodiversity will benefit from awareness of the limitations of ENMs in predicting the extinction of populations/species. Greater knowledge of how niches have changed through time and how this relates to the characteristics of species is needed to improve the reliability of ENMs. Niche changes in plants may also affect palaeoclimatic estimates based on fossil pollen.

Field validation shows bias‐corrected pseudo‐absence selection is the best method for predictive species‐distribution modelling

AbstractAimTo determine the performance of different pseudo‐absence selection strategies on the prediction of species‐distribution models after 30 years of regional climatic and land use changes.LocationContinental France and the Iberian Peninsula.MethodsIn this study, we used a large database of Coprophagous Scarabaeidae beetle records collected between 1970 and 1980 in continental France and the Iberian Peninsula to assess the relative performance of different modelling methods in predicting species distributions using current climate and land use information. We usedmaxentwith standard settings and boosted regression trees with three different approaches to generate pseudo‐absences. We used historical data to model species distribution and then projected the models into the present. Each method's performance was then assessed by specific field sampling conducted at 20 different sites.ResultsField validation demonstrated that model predictions were more accurate when pseudo‐absence data were selected from a sampling bias grid and that model evaluations based on test datasets can lead to false conclusions if not correctly calibrated. The study also demonstrated that the method in which pseudo‐absences are dealt with has a major impact on ecological conclusions.Main conclusionCorrecting for spatial bias in collections datasets is of great importance for predicting future trends in species distributions. Uncorrected models showed a strong bias in their predicted species richness patterns.

Local-scale biotic interactions embedded in macroscale climate drivers suggest Eltonian noise hypothesis distribution patterns for an invasive grass.

A hierarchical view of niche relations reconciles the scale-dependent effects of abiotic and biotic processes on species distribution patterns and underlies most current approaches to distribution modeling. A key prediction of this framework is that the effects of biotic interactions will be averaged out at macroscales - an idea termed the Eltonian noise hypothesis (ENH). We test this prediction by quantifying regional variation in local abiotic and biotic niche relations and assess the role of macroclimate in structuring biotic interactions, using a non-native invasive grass, Microstegium vimineum, in its introduced range. Consistent with hierarchical niche relations and the ENH, macroclimate structures local biotic interactions, while local abiotic relations are regionally conserved. Biotic interactions suppress M. vimineum in drier climates but have little effect in wetter climates. A similar approach could be used to identify the macroclimatic conditions under which biotic interactions affect the accuracy of local predictions of species distributions.

Sapling performance along resource gradients drives tree species distributions within and across tropical forests

Niche differentiation is a major hypothesized determinant of species distributions, but its practical importance is heavily debated and its underlying mechanisms are poorly understood. Trait‐based approaches have been used to infer niche differentiation and predict species distributions. For understanding underlying mechanisms, individual traits should be scaled up to whole‐plant performance, which has rarely been done.We measured seven key traits that are important for carbon and water balance for 37 tropical tree species. We used a process‐based plant physiological model to simulate the carbon budget of saplings along gradients of light and water availability, and quantified the performance of the species in terms of their light compensation points (a proxy for shade tolerance), water compensation points (proxy for drought tolerance), and maximum carbon gain rates (proxy for potential growth rate). We linked species performances to their observed distributions (the realized niches) at two spatial scales in Bolivian lowland forests: along a canopy openness gradient at local scale (∼1 km2) and along a rainfall gradient (1100–2200 mm/yr) at regional (∼1000 km) scale.We show that the water compensation point was the best predictor of species distributions along water and light resource gradients within and across tropical forests. A sensitivity analysis suggests that the stomatal regulation of minimum leaf water potentials, rather than stem hydraulic traits (sapwood area and specific conductivity), contributed to the species differences in the water compensation point of saplings. The light compensation point and maximum carbon gain, both driven by leaf area index and leaf nitrogen concentration, also contributed to differential species distributions at the local scale, but not or only marginally at the regional scale. Trait‐and‐physiology‐based simulations of whole‐plant performance thus help to evaluate the possible roles of individual traits in physiological processes underlying species performance along environmental gradients. The development of such whole‐plant concepts will improve our ability to understand responses of plant communities to shifts in resource availability and stress under global change.

Using vegetation data within presettlement land survey records for species distribution modeling: A tale of two datasets

Researchers have long utilized the vegetation data within presettlement land survey records (PLSRs) to understand past forest composition in North America. PLSRs typically contain two datasets: bearing-tree (BT) data and line-description (LD) data. BT data are records of the trees that surveyors blazed adjacent to survey monuments, whereas LD data provide descriptions of tree species that surveyors observed along survey lines. Recently, studies have applied BT data to develop species distribution models (SDMs). SDMs create predictions of species distributions, based upon the modeled relationship between species presence and absence records, and environmental variables. Despite the applications of BT data in SDMs, the value of LD data for developing SDMs has not been explored. This study compares SDMs trained from LD data versus BT data, using PLSRs that were created ca. 1799–1814 CE in Chautauqua County, New York State. Using consensus modeling techniques, this study finds that despite positional uncertainty issues, LD data produce SDMs with better predictive performance than BT data, and more adequately generalize to independent datasets. Moreover, a comparable amount of data can be collected from LD data as from BT data, in order to develop models with greater predictive ability. This study challenges the use of BT data in SDMs, and suggests that modeling past species distributions can be accomplished more effectively using LD data.

The use of species distribution models to predict the spatial distribution of deforestation in the western Brazilian Amazon

The prevention of deforestation in rainforests requires the identification of where facilitating and mitigating factors will combine and increase the likelihood of deforestation. This approach, which relates a geographic space with an environmental space of factors to predict where new deforestation will occur, is very similar to the approaches used to predict species distributions. Thus, we believe that deforestation can be treated as a “species” and that its future occurrence can be determined using species distribution models.The objective of this work is to test the efficiency of species distribution models in predicting the potential areas of deforestation in a region of the western Amazon. We analyzed five different areas in the arc of deforestation. For each area, we ran the MaxEnt model in six different experiments to determine the boundaries of the probability distributions.Potential areas identified using the different models of MaxEnt were very effective in predicting deforestation areas. The models that used only previous deforestation density were less effective than the models that included functional variables.In four of the five areas, 80% of the new deforestation occurred in the area predicted by the models. These models were more effective than the business-as-usual (BAU) and governance (GOV) model scenarios described using the DINAMICA-EGO platform by Soares-Filho et al. (2006).Species distribution models are a valuable tool for determining potential areas of future Amazon deforestation. The use of these models arises as support to efforts to conserve tropical forests and identify critical locations where command and control actions against deforestation can be most efficient.

Mutualist-mediated effects on species' range limits across large geographic scales.

Understanding the processes determining species range limits is central to predicting species distributions under climate change. Projected future ranges are extrapolated from distribution models based on climate layers, and few models incorporate the effects of biotic interactions on species' distributions. Here, we show that a positive species interaction ameliorates abiotic stress, and has a profound effect on a species' range limits. Combining field surveys of 92 populations, 10 common garden experiments throughout the range, species distribution models and greenhouse experiments, we show that mutualistic fungal endophytes ameliorate drought stress and broaden the geographic range of their native grass host Bromus laevipes by thousands of square kilometres (~ 20% larger) into drier habitats. Range differentiation between fungal-associated and fungal-free grasses was comparable to species-level range divergence of congeners, indicating large impacts on range limits. Positive biotic interactions may be underappreciated in determining species' ranges and species' responses to future climates across large geographic scales.

Can fire atlas data improve species distribution model projections?

Correlative species distribution models (SDMs) are widely used in studies of climate change impacts, yet are often criticized for failing to incorporate disturbance processes that can influence species distributions. Here we use two temporally independent data sets of vascular plant distributions, climate data, and fire atlas data to examine the influence of disturbance history on SDM projection accuracy through time in the mountain ranges of California, USA. We used hierarchical partitioning to examine the influence of fire occurrence on the distribution of 144 vascular plant species and built a suite of SDMs to examine how the inclusion of fire-related predictors (fire occurrence and departure from historical fire return intervals) affects SDM projection accuracy. Fire occurrence provided the least explanatory power among predictor variables for predicting species' distributions, but provided improved explanatory power for species whose regeneration is tied closely to fire. A measure of the departure from historic fire return interval had greater explanatory power for calibrating modern SDMs than fire occurrence. This variable did not improve internal model accuracy for most species, although it did provide marginal improvement to models for species adapted to high-frequency fire regimes. Fire occurrence and fire return interval departure were strongly related to the climatic covariates used in SDM development, suggesting that improvements in model accuracy may not be expected due to limited additional explanatory power. Our results suggest that the inclusion of coarse-scale measures of disturbance in SDMs may not be necessary to predict species distributions under climate change, particularly for disturbance processes that are largely mediated by climate.

Initial temperature of seed germination as related to species occurrence along a temperature gradient

SummaryReproductive stages of life cycle are important for the explanation of distribution patterns of plant species at different scales, due to their extreme vulnerability to environmental conditions. Despite reported evidences that seed germination is related to habitat macroclimatic characteristics such as mean annual temperature (MAT) and precipitation, the role of this trait in controlling plant species distribution is not systematically and quantitatively evaluated yet.Using the data on seed germination along a temperature gradient for 49 species originating from contrasting climatic conditions, we test here whether initial temperature of seed germination (Tmin) is a direct correlate for predicting species distribution ranges along the temperature gradient.Our study reveals thatTminis strongly negatively correlated with habitat temperature; among the studied species,Tminclearly increased with decreasingMAT(r2 = 0·57,P < 0·001). Considering phylogenetic biases, co‐evolution of seed traits as well as precipitation along with microclimatic factors did not affect the strength of this relationship.The results suggest that theTmin–MATrelation can provide insights particularly into species distribution patterns, vegetation dynamics and community assembly rules along altitudinal and latitudinal gradients. We argue that including theTminin species distribution models may help to improve the accuracy and specificity of predictions of vegetation shifts under global change scenarios.

Freshwater conservation planning under climate change: demonstrating proactive approaches for Australian Odonata

SummaryClimate change represents a major challenge for conservation in the future and undermines protection within reserve boundaries. Freshwater biodiversity is still under‐represented within reserves world‐wide, and connectivity among reserves will become increasingly crucial if species are to persist under climate change.We tested the likely benefits of including predicted species distributions in systematic reserve design for rivers under climate change and the impact of varying connectivity requirements on future representation.We used the modelled distribution of 126 east Australian Odonata to identify reserve networks using data for current or future (2055 and 2085) distributions either by filling gaps additively, or as separate targets in a single solution. We then assessed the potential improvements to species representation in the future using different types of connectivity penalties that emphasized either longitudinal riverine connections or connections to all neighbouring subcatchments.Solutions that did not include future distributions in the planning stages were 16 to 30% less likely to protect the same species by 2055 and 2085, respectively. Inclusion of species' future distributions in the design phase leads to short‐term increases in cost, but in the longer term fewer additional areas are required to meet targets and this strategy is likely to be significantly more efficient than implementing systematic design in stages. In addition, solely targeting riverine connectivity was significantly less likely to protect current species in the future than if cross‐catchment connections were included.Synthesis and applications. Where protected areas can be expanded to assist species adaptation to climate change, significant gains in efficiency are possible if longer term goals are considered when selecting sites. Furthermore, to improve the representation of species under future climates, reserve selection should consider inter‐catchment connectivity, although the nature of optimal solutions will depend heavily on the range of taxa included, their dispersal capacity, and the availability of climatic refugia.

Vulnerability of breeding waterbirds to climate change in the Prairie Pothole Region, U.S.A.

The Prairie Pothole Region (PPR) of the north-central U.S. and south-central Canada contains millions of small prairie wetlands that provide critical habitat to many migrating and breeding waterbirds. Due to their small size and the relatively dry climate of the region, these wetlands are considered at high risk for negative climate change effects as temperatures increase. To estimate the potential impacts of climate change on breeding waterbirds, we predicted current and future distributions of species common in the PPR using species distribution models (SDMs). We created regional-scale SDMs for the U.S. PPR using Breeding Bird Survey occurrence records for 1971–2011 and wetland, upland, and climate variables. For each species, we predicted current distribution based on climate records for 1981–2000 and projected future distributions to climate scenarios for 2040–2049. Species were projected to, on average, lose almost half their current habitat (-46%). However, individual species projections varied widely, from +8% (Upland Sandpiper) to -100% (Wilson's Snipe). Variable importance ranks indicated that land cover (wetland and upland) variables were generally more important than climate variables in predicting species distributions. However, climate variables were relatively more important during a drought period. Projected distributions of species responses to climate change contracted within current areas of distribution rather than shifting. Given the large variation in species-level impacts, we suggest that climate change mitigation efforts focus on species projected to be the most vulnerable by enacting targeted wetland management, easement acquisition, and restoration efforts.

Assessing landscape constraints on species abundance: does the neighborhood limit species response to local habitat conservation programs?

Landscapes in agricultural systems continue to undergo significant change, and the loss of biodiversity is an ever-increasing threat. Although habitat restoration is beneficial, management actions do not always result in the desired outcome. Managers must understand why management actions fail; yet, past studies have focused on assessing habitat attributes at a single spatial scale, and often fail to consider the importance of ecological mechanisms that act across spatial scales. We located survey sites across southern Nebraska, USA and conducted point counts to estimate Ring-necked Pheasant abundance, an economically important species to the region, while simultaneously quantifying landscape effects using a geographic information system. To identify suitable areas for allocating limited management resources, we assessed land cover relationships to our counts using a Bayesian binomial-Poisson hierarchical model to construct predictive Species Distribution Models of relative abundance. Our results indicated that landscape scale land cover variables severely constrained or, alternatively, facilitated the positive effects of local land management for Ring-necked Pheasants.

Predicting species distributions in new areas or time periods with alpha-shapes

Statistical models relating species distributions to environmental data are now commonly applied to predict where invasive species may become established or how range limits may shift under climate change. As species absences can originate from factors other than an unsuitable environment (e.g. dispersal constraints), the models that discriminate between occupied and unoccupied environments are likely to underestimate potential ranges. However, the techniques that “envelope” the occupied environments (i.e. profile techniques) usually rely on simple convex estimators (e.g. elliptical or rectangular shapes), which tend to overestimate these ranges. Here we describe alpha-shapes, a profile-type technique that relaxes the assumption of convexity. By using native range data for the invasive African clawed frog, we demonstrate how this technique can be used to model climatic envelopes of variable complexity. In particular, we compared predictions from an envelope maximizing discrimination between presences and absences, an envelope tightly enclosing all occupied climatic combinations (i.e. the minimum bounding envelope) and an “expert-based” generalization of the previous. In addition, we also use this technique to identify climatic combinations that are outside the climatic space of the study area (i.e. non-analog climates). The envelope accounting for the absences of the African clawed frog achieved a high discrimination ability (true skill statistics=0.71), but failed to predict many of the areas in which the species occurs. Predictions based on the minimum bounding envelope encompassed all species occurrences while still providing a sharp delineation of its distribution range. The generalized version of the previous envelope also captured all occurrences, but predicted a wider extent of suitable areas. We also found that most parts of the world present climatic conditions that are non-analog to those of our study area. Although conceptually more suitable for predicting species distributions across space and time than presence–absence models, profile techniques are frequently overlooked because of their inability to fit flexible envelopes. Here, we demonstrate that alpha-shapes are a transparent and intuitive profile-type technique that has this flexibility.

Local Hydrological Conditions Explain Floristic Composition in Lowland Amazonian Forests

AbstractAmazonian forests harbor a large variety of understory herbs adapted to areas with different hydrological conditions, ranging from well‐drained to seasonally flooded forests. The presence versus absence of flooding forms the extremes of a hydrological gradient, with various intermediate conditions, such as seasonal soil waterlogged areas, in between. We investigated the relationship between understory herbs and hydrological conditions in Central Amazonian forests using eighty‐eight 250 × 2 m plots distributed along a 600‐km transect. Hydrological conditions were determined regionally by precipitation and locally by topographic conditions based on drainage potential, flooding height and soil permeability (sand content). Soil cation concentration was used as a proxy for soil fertility. The floristic dissimilarities among plots were visualized by Nonmetric Multidimensional Scaling, and simple and multiple regressions were used to identify the best predictor of herb species composition. Local drainage potential was more important in determining herb species composition than soil fertility or precipitation at non‐flooded and flooded sites. Flooded sites comprised a very distinctive herb species composition even when the flood height was low (0.3 m). We conclude that hydrological conditions are the primary constraint of herb distribution within this flat regional landscape with moderate amounts of soil fertility variation (0.09–2.280 cmol(+)/kg). Hydrological models that consider local water conditions explained the largest part of herb species composition. Therefore, predictions of species distribution based on large‐scale climatic variables may underestimate the favorable area for understory herbs if the variation on local hydrological conditions is not considered.

Climate-driven range shifts explain the distribution of extant gene pools and predict future loss of unique lineages in a marine brown alga.

The climate-driven dynamics of species ranges is a critical research question in evolutionary ecology. We ask whether present intraspecific diversity is determined by the imprint of past climate. This is an ongoing debate requiring interdisciplinary examination of population genetic pools and persistence patterns across global ranges. Previously, contrasting inferences and predictions have resulted from distinct genomic coverage and/or geographical information. We aim to describe and explain the causes of geographical contrasts in genetic diversity and their consequences for the future baseline of the global genetic pool, by comparing present geographical distribution of genetic diversity and differentiation with predictive species distribution modelling (SDM) during past extremes, present time and future climate scenarios for a brown alga, Fucus vesiculosus. SDM showed that both atmospheric and oceanic variables shape the global distribution of intertidal species, revealing regions of persistence, extinction and expansion during glacial and postglacial periods. These explained the distribution and structure of present genetic diversity, consisting of differentiated genetic pools with maximal diversity in areas of long-term persistence. Most of the present species range comprises postglacial expansion zones and, in contrast to highly dispersive marine organisms, expansions involved only local fronts, leaving distinct genetic pools at rear edges. Besides unravelling a complex phylogeographical history and showing congruence between genetic diversity and persistent distribution zones, supporting the hypothesis of niche conservatism, range shifts and loss of unique genetic diversity at the rear edge were predicted for future climate scenarios, impoverishing the global gene pool.

Optimising long‐term monitoring projects for species distribution modelling: how atlas data may help

Long‐term biodiversity monitoring data are mainly used to estimate changes in species occupancy or abundance over time, but they may also be incorporated into predictive models to document species distributions in space. Although changes in occupancy or abundance may be estimated from a relatively limited number of sampling units, small sample size may lead to inaccurate spatial models and maps of predicted species distributions. We provide a methodological approach to estimate the minimum sample size needed in monitoring projects to produce accurate species distribution models and maps. The method assumes that monitoring data are not yet available when sampling strategies are to be designed and is based on external distribution data from atlas projects. Atlas data are typically collected in a large number of sampling units during a restricted timeframe and are often similar in nature to the information gathered from long‐term monitoring projects. The large number of sampling units in atlas projects makes it possible to simulate a broad gradient of sample sizes in monitoring data and to examine how the number of sampling units influences the accuracy of the models. We apply the method to several bird species using data from a regional breeding bird atlas. We explore the effect of prevalence, range size and habitat specialization of the species on the sample size needed to generate accurate models. Model accuracy is sensitive to particularly small sample sizes and levels off beyond a sufficiently large number of sampling units that varies among species depending mainly on their prevalence. The integration of spatial modelling techniques into monitoring projects is a cost‐effective approach as it offers the possibility to estimate the dynamics of species distributions in space and over time. We believe our innovative method will help in the sampling design of future monitoring projects aiming to achieve such integration.