Shade trees’ implementation into cocoa agroforestry systems in tropical West Africa has proven to have a high potential in increasing farm resilience and mitigating climate change. However, no studies have yet examined the potential future climate suitability for shade trees in the region, which is important when deciding which shade trees to implement. The aim of this study was to predict the current and future climate suitability of shade trees currently used in cocoa agroforestry systems and give recommendations based on the results. It also aimed to examine how different levels of shadetree diversity can impact a farm’s microclimate and how this can be linked to climate suitability for cocoa. To assess climate suitability, a species distribution model was performed for a baseline scenario with the current climate defined as 1970-2000 and for two future scenarios, SSP126 and SSP585, for the time period 2041-2060. To measure microclimate, a microclimatic study was performed in the Ashanti region of Ghana on 16 farms during mid March to mid May 2023 measuring the maximum temperature above cocoa canopy and 15 cm above the ground for three different treatments. The results showed that three species, Khaya senegalensis, Ceiba pentandra and Albiziazygia, were predicted to have the largest habitat areas of climate suitability in West Africa for several of the scenarios. However, it was also identified that the model could be improved when it comes tothe inclusion of more bioclimatic variables, global circulation models (GCMs) and for which futures cenarios to model for. The microclimatic study showed that both farms with a low and high level of shade tree diversity have a significant possibility to buffer the maximum temperature above cocoa canopy, with low diversity farms having a larger buffering capacity. The study also showed that areaswhere several shade tree species are predicted to have a climatically suitable habitat decreased for future scenarios, which could be a risk for the possibility to mitigate climate change for cocoa with shade tree implementation in the future. / Att implementera skuggträd i kakaoskogslantbruk i tropiska Västafrika har visat sig ha en högpotential att öka resiliens på gårdar såväl som att motverka effekterna av klimatförändringar iregionen. Däremot har inga tidigare studier hittills undersökt vilka skuggträdsarter som väntas varalämpliga i framtida klimat. Syftet med den här studien var att uppskatta nutida och framtidalämplighet för skuggträd som nu används i kakaoskogslantbruk och ge rekommendationer utifrånresultaten. Studien hade också som syfte att undersöka hur olika nivåer av diversitet av skuggträd på gårdarna påverkade deras mikroklimat samt hur detta kan kopplas till klimatlämplighet för kakao. Klimatlämplighet uppskattades med en distributionsmodell (eng: “species distribution model”), med ett basscenario för nutida klimat satt som 1970-2000 samt med två framtida klimatscenarier, SSP126 och SSP585, för tidsperioden 2041-2060. Mikroklimat mättes genom en fältstudie som genomfördes på 16 gårdar i Ashantiregionen i Ghana under perioden mitten av mars till mitten av maj. Mikroklimatet mättes för den maximalt uppnådda temperaturen ovanför kakaoträdens lövverk och 15 cm ovanför marken. Resultaten visade att tre arter, Khaya senegalensis, Ceiba pentandra och Albiziazygia, hade störst område där de var lämpliga för flest klimatscenarier i Västafrika. Dessutom identifierades flera sätt att förbättra modellen, bland annat genom att inkludera fler bioklimatiska variabler, fler globala cirkulationsmodeller (GCMs) och genom att förutspå utvecklingen för fler framtida klimatscenarier. För mikroklimat visade resultaten att både en hög och låg diversitet av skuggträd resulterade i att den maximala temperaturen minskade ovanför skuggträdens lövverk, och den minskade mer där diversiteten var låg. Resultaten visade även att för framtida klimatscenarier minskar områdena där flera skuggträd är lämpliga, vilket kan vara problematiskt för möjligheten att bemöta klimatförändringar i framtiden genom skuggträdsimplementering.

De nombreuses recherches se focalisent sur l'étude des aires de distribution des espèces qui se décalent vers des conditions plus adaptées à leurs besoins physiologiques sous l'effet du changement climatique. Le choix des indices utilisés pour caractériser l'écologie des espèces et définir leur vulnérabilité au réchauffement en cours est souvent conditionné par leur disponibilité, alors qu'il devrait être basé sur les connaissances en écophysiologie qui les concernent. D'autre part, la résolution spatiale parfois grossière utilisée n'est pas toujours pertinente au regard de l'échelle à laquelle les processus biologiques se déroulent. Dans ce cadre, l'objectif de ce travail est de cartographier à fine résolution spatiale les bilans en eau des sols et leurs différentes composantes à l'échelle des forêts de France, et d'évaluer leur intérêt pour modéliser la distribution ou la productivité des espèces au regard des indices traditionnellement utilisés. Dans un premier temps, nous avons modélisé et cartographié les différentes composantes du bilan en eau des sols, et tout particulièrement le rayonnement solaire et la réserve utile maximale en eau (RUM) des sols forestiers à partir des relevés de l'Inventaire Forestier National (IFN). Ces données ont été combinées avec des températures et des précipitations pour spatialiser le bilan en eau des sols forestiers de France. Les principaux résultats montrent l'importance de la nébulosité dans la prise en compte du calcul du rayonnement solaire, et l'inefficacité des indices dérivés de l'exposition pour en simuler les valeurs à l'échelle de la France. Nous avons également déterminé qu'il est possible de réaliser avec des informations simples à collecter une carte des RUM des sols forestiers de France. Elle permet de prédire la croissance des essences avec une efficacité comparable aux valeurs relevées sur des placettes et d'améliorer la modélisation de la distribution de certaines essences. Enfin, nous démontrons que les calculs de bilans en eau qui prennent en compte la réserve en eau des sols sont plus efficaces que les bilans hydriques climatiques ou les pluies, particulièrement pour ce qui concerne les espèces hygrophiles ou xérophiles. Ces résultats laissent penser que l'importance de l'eau a été sous-estimée dans l'analyse de la distribution des espèces et l'étude des conséquences du changement climatique sur les plantes. Les données produites permettent de progresser dans la connaissance de l'écologie des espèces et de mieux caractériser la vulnérabilité des espèces, ouvrant la porte à la création d'outils plus fonctionnels pour aider les gestionnaires à évaluer les impacts du changement de climat et à s'y adapter. / Numerous researches focus on species distribution shifts toward ecological conditions most suited to plants under climate change. Ecological indices used to characterize species ecology and to define their vulnerability over broad areas are often at coarse resolution and are determined by data availability. The aim of this work was to map soil water balance and its different components at a fine spatial resolution, and to evaluate their interest to model plant distribution and growth over the whole French forests. We firstly modeled and mapped the solar radiation and the soil water holding capacity of forest soils. These data were combined with temperatures and precipitation to map the soil water balance. For solar radiation, the main results showed that this parameter is only accurately predicted at the French scale when cloudiness is taken into account. We also showed that soil water holding capacity can be mapped at the French scale using the basic information collected on numerous plots from the French national forest inventory. Values extracted from the soil water holding capacity map allowed predicting tree species growth with efficiency similar to values estimated on plots. We also demonstrated soil water balance is more efficient than climatic water balance or precipitation to model species distribution, mainly for hygrophilous and xerophilous species. These results suggest importance of available water could be underestimated when determining the ecological niche of species. These maps allow to improve species ecology knowledge and to help in the determination of their vulnerability area to climate change.

<p>U radu je izvr&scaron;ena podela 572 registrovane vrste osolikih muva sa područja jugoistočne Evrope na funkcionalne grupe, na osnovu registrovanih funkcionalnih karaktera. Za<br />odabrane 44 endemske i nativne ne-endemske vrste koje su svojim klimatskim ni&scaron;ama ograničene na istraživani prostor, kreirane su mape sada&scaron;nje i buduće potencijalne distribucije (za dva perioda, 2041 -2060 i 2061-2080) upotrebom MAXENT algoritma. Na osnovu dobijenih mapa, izračunata je potenijalna promena areala za odabrane vrste, čime je procenjen efekat klimatskih promena na distribuciju vrsta sirfida. Detektovane su vrste koje će najvi&scaron;e biti pogođene klimatskim promenama, kao i prostori sa najvećim potencijalnim bogatstvom vrsta u budućnosti i područja predviđena da će izgubiti deo diverziteta sirfida. Pokazano je da vrste sa ograničenim arealom neće u budućnosti iskusiti veće smanjenje areala u odnosu na &scaron;iroko rasprostranjene vrste. Takođe, na osnovu mapa potenicijalne sada&scaron;nje distribucije i procenjene retkosti vrsta, izvr&scaron;ena je prioretizacija područja značajnih za opstanak osolikih muva u jugoistočnoj Evropi. Primenom linearnih regresionih i linearnih modela sa me&scaron;ovitim efektom ispitana je međuzavisnost&nbsp; funkcionalnih karaktera vrsta i procenjene promene veličine areala. Definisani su&nbsp; funkcionalni karakteri koji utiču na promenu veličine areala osolikih muva. Ustanovljeni koncept može biti upotrebljen za detekciju vrsta koje se jo&scaron; uvek ne smatraju ugroženim, ali zbog svojih biolo&scaron;ko-ekolo&scaron;kih karakteristika imaju potencijal da to postanu, te stoga zaslužuju pažnju konzervacionista. Dobijeni rezultati mogu u velikoj meri poslužiti za kreiranje sistematskog konzervacionog plana za očuvanje osolikih muva na teritoriji jugoistočne Evrope.</p> / <p>In this paper, the division of 572 species of hoverflies registered in Southeast Europe into functional groups was conducted, based on their functional traits. For 44 selected&nbsp; endemic and native non-endemic species that have their climatic niches limited in the study area, maps of current and future potential distributions were created (for the periods, 2041-2060 and 2061-2080) using MAXENT algorithm. Based on the obtained &nbsp; maps, changes in potential area of occupancy were calculated, in order to estimate the &nbsp; effect of climate change on the distribution of hoverfly species. Species that will be most affected by climate change were detected, as well as the areas with the greatest potential species richness in the future and areas predicted to lose part of the diversity of hoverflies. It is shown that species with limited distribution in the future will not experience a greater reduction of its area in relation to the widespread species. Also, based on the current maps of potential distribution and assessed rarity of species,priority areas important for the survival of hoverflies in Southeast Europe were established. By using linear regression and linear mixed effect models, the interdependence between functional traits and the estimated changes in the range size was tested. Traits affecting the changes in range size of hoverflies were determined. The established concept can be used for the detection of species that are not yet considered endangered, but because of their biological and ecological traits have the potential to become threatened, and therefore deserve the attention of conservationists.&nbsp; This results can largely be used to create a systematic conservation plan for the preservation of hoverflies in Southeast Europe.</p>

A modelagem de distribuição de espécies é uma técnica cuja finalidade é estimar modelos baseados em nichos ecológicos. Esses modelos podem auxiliar nos processos de tomadas de decisões, no planejamento e na realização de ações que visem a conservação e a preservação ambiental. Existem diversas ferramentas projetadas para modelagem de distribuição de espécies, dentre elas o framework openModeller, na qual este trabalho está inserido. Várias técnicas de Inteligência Artificial já foram utilizadas para desenvolver algoritmos de modelagem de distribuição de espécies, como Entropia Máxima. No entanto, as ferramentas estatísticas tradicionais não disponibilizam pacotes com o algoritmo de Entropia Máxima, o que é comum para outras técnicas. Além disso, apesar de existir um software gratuito específico para modelagem de distribuição de espécies com o algoritmo de Entropia Máxima, esse software não possui código aberto. Assim, a base deste trabalho é a investigação acerca da modelagem de distribuição de espécies utilizando Entropia Máxima. Desta forma, o objetivo principal é definir diferentes estratégias para o algoritmo de Entropia Máxima no contexto da modelagem de distribuição de espécies. Para atingir esse objetivo, foram estabelecidos um conjunto de alternativas possíveis a serem exploradas e um conjunto de métricas de avaliação e comparação das diferentes estratégias. Os resultados mais importantes desta pesquisa foram: um algoritmo adaptativo de Entropia Máxima, um algoritmo paralelo de Entropia Máxima, uma análise do parâmetro de regularização e um método de seleção de variáveis baseado no princípio da Descrição com Comprimento Mínimo (MDL Minimum Description Length), que utiliza aprendizagem por compressão de dados. / Species distribution modeling is a technique the purpose of which is to estimate models based on ecological niche. These models can assist decision making processes, planning and carrying out actions aiming at environmental conservation and preservation. There are several tools designed for species distribution modeling, such as the open- Modeller framework, in which this work is inserted. Several Artificial Intelligence techniques have been used to develop algorithms for species distribution modeling, such as Maximum Entropy. However, traditional statistical tools do not offer packages with the Maximum Entropy algorithm, which is common to other techniques. Furthermore, although there is specific free software for species distribution modeling with the Maximum Entropy algorithm, this software is not open source. The basis of this work is the investigation of the species distribution modeling using Maximum Entropy. Thus, its aim is to define different strategies for the Maximum Entropy algorithm in the context of the species distribution modeling. For this, a set of possible alternatives to be explored and a set of metrics for evaluation and comparison of the different strategies were established. The most important results were: an adaptive Maximum Entropy algorithm, a parallel Maximum Entropy algorithm, an analysis of the regularization parameter and a variable selection method based on the Minimum Description Length principle, which uses learning by data compression.

A modelagem de distribuição de espécies é uma técnica cuja finalidade é estimar modelos baseados em nichos ecológicos. Esses modelos podem auxiliar nos processos de tomadas de decisões, no planejamento e na realização de ações que visem a conservação e a preservação ambiental. Existem diversas ferramentas projetadas para modelagem de distribuição de espécies, dentre elas o framework openModeller, na qual este trabalho está inserido. Várias técnicas de Inteligência Artificial já foram utilizadas para desenvolver algoritmos de modelagem de distribuição de espécies, como Entropia Máxima. No entanto, as ferramentas estatísticas tradicionais não disponibilizam pacotes com o algoritmo de Entropia Máxima, o que é comum para outras técnicas. Além disso, apesar de existir um software gratuito específico para modelagem de distribuição de espécies com o algoritmo de Entropia Máxima, esse software não possui código aberto. Assim, a base deste trabalho é a investigação acerca da modelagem de distribuição de espécies utilizando Entropia Máxima. Desta forma, o objetivo principal é definir diferentes estratégias para o algoritmo de Entropia Máxima no contexto da modelagem de distribuição de espécies. Para atingir esse objetivo, foram estabelecidos um conjunto de alternativas possíveis a serem exploradas e um conjunto de métricas de avaliação e comparação das diferentes estratégias. Os resultados mais importantes desta pesquisa foram: um algoritmo adaptativo de Entropia Máxima, um algoritmo paralelo de Entropia Máxima, uma análise do parâmetro de regularização e um método de seleção de variáveis baseado no princípio da Descrição com Comprimento Mínimo (MDL Minimum Description Length), que utiliza aprendizagem por compressão de dados. / Species distribution modeling is a technique the purpose of which is to estimate models based on ecological niche. These models can assist decision making processes, planning and carrying out actions aiming at environmental conservation and preservation. There are several tools designed for species distribution modeling, such as the open- Modeller framework, in which this work is inserted. Several Artificial Intelligence techniques have been used to develop algorithms for species distribution modeling, such as Maximum Entropy. However, traditional statistical tools do not offer packages with the Maximum Entropy algorithm, which is common to other techniques. Furthermore, although there is specific free software for species distribution modeling with the Maximum Entropy algorithm, this software is not open source. The basis of this work is the investigation of the species distribution modeling using Maximum Entropy. Thus, its aim is to define different strategies for the Maximum Entropy algorithm in the context of the species distribution modeling. For this, a set of possible alternatives to be explored and a set of metrics for evaluation and comparison of the different strategies were established. The most important results were: an adaptive Maximum Entropy algorithm, a parallel Maximum Entropy algorithm, an analysis of the regularization parameter and a variable selection method based on the Minimum Description Length principle, which uses learning by data compression.

Beekeepers in Appalachia market varietal honeys derived from particular species of deciduous trees; however, finding places in a mountainous landscape to locate new beeyards is difficult. Site selection is hindered by the high up-front costs of negotiating access to remote areas with limited knowledge of the available forage. Remotely sensed data and species distribution modeling (SDM) of trees important to beekeepers could aid in locating apiary sites at the landscape scale. The objectives of this study are i) using publicly available forest inventory data, to model the spatial distribution of three native tree species that are important to honey producers in eastern Kentucky: American Basswood, Sourwood and Tulip Poplar, and to assess the accuracy of the models, ii) to incorporate a method for discounting the value of a nectar resource as a function of distance based on an energetic model of honeybee foraging, and iii) to provide an example by ranking potential apiary locations around the perimeter of a mine site in the study area based on their proximity to probable species habitat using a GIS model. Logistic regression models were trained using presence-absence records from 1,059 USFS Forest Inventory and Analysis (FIA) sub-plots distributed throughout a 9,000 km2 portion of the Kentucky River watershed. The models were evaluated by applying them to a separate dataset, 950 forest inventory sub-plots distributed over a 40.5 km2 research forest maintained by the University of Kentucky. Weights derived from an energic model of honeybee foraging were then applied to the probabilities of tree species occurrence predicted by the SDM. As an example, 24 potential apiary locations around the perimeter of a reclaimed mine site were selected and then ranked according to a site suitability index. Three tributary areas corresponding to different honeybee flight ranges were considered: 500m, 700m, and 1,200m. Results confirm that rankings are dependent on the foraging range considered, suggesting that the number of colonies at an apiary location would be an important factor to consider when choosing a site. However, the methodology makes assumptions that are only anecdotally supported, notably i) that colonies will forage preferentially at the target species when it is in bloom and, ii) that foragers will exhaust resources closest to the hive first, regardless of patch size. Additional study of how bees deplete the nectar resources surrounding an apiary is needed to verify the usefulness of SDM in site selection for varietal honey production.

Several factors have contributed to impeding the progress of migratory waterfowl spatial modeling, such as (1) waterfowl’s reliance on wetlands, (2) lack of understanding about shifts in distributions through time, and (3) large-scale seasonal migration. This doctoral dissertation provides an array of tools to address each of these concerns in order to better understand and conserve this group of species. The second chapter of this dissertation addresses issues of modeling species dependent on wetlands, a dynamic and often ephemeral habitat type. Correlation models of the relationships between climatic variables and species occurrence will not capture the full habitat constraints of waterfowl. This study introduces a novel data source that explicitly models the depth to water table, which is a simulated long-term measure of the point where climate and geological/topographic water fluxes balance. The inclusion of the depth to water table data contributes significantly to the ability to predict species probability of occurrence. Furthermore, this data source provides advantages over traditional proxies for wetland habitat, because it is not a static measure of wetland location, and is not biased by sampling method. Utilizing the long-term banding bird data again, the third chapter examines the behavior of waterfowl niche selection through time. By using the methods developed in chapter two, probability of occurrence models for the 1950s and the 1990s were developed. It was then possible to detect movements in geographic and environmental space, and how movements in these two spaces are related. This type of analysis provides insight into how different bird species might respond to environment changes and potentially improve climate change forecasts. The final chapter presents a new method for predicting the migratory movement of waterfowl. The method incorporates not only the environmental constraints of stopover habitat, but also includes likely distance and bearing traveled from a source point. This approach uses the USGS’ banding bird database; more specifically, it relies on banding locations, which have multiple recoveries within short time periods. Models made from these banding locations create a framework of migration movement, and allow for predictions to be made from locations where no banding/recovery data are available. / text

Em vista das mudanças ambientais em nível global, disponibilizar informações ecológicas e buscar uma melhor compreensão dos fatores e processos que moldam a distribuição de espécies, é uma iniciativa importante para o planejamento de ações de conservação. Neste contexto, a importância e carência de informações sobre a distribuição geográficas das espécies nos motivaram a predizer a distribuição potencial de arbustos e árvores das famílias Lauraceae e Myrtaceae na Floresta Atlântica, no sul do Brasil. Modelos lineares generalizados (GLM) foram usados para ajustar modelos preditivos com os registros de ocorrência de 88 espécies em função de variáveis ambientais. As variáveis preditoras foram selecionadas com base no menor critério de informação de Akaike corrigido. Nós avaliamos o desempenho dos modelos usando o método de validação cruzada (10-fold) para calcular a habilidade estatística verdadeira (TSS) e a área sob a curva característica do operador receptor (AUC). Nós usamos GLM para testar a influência da área de ocorrência estimada, do número de registros das espécies e da complexidade dos modelos sobre a TSS e a AUC. Nossos resultados mostraram que as variáveis climáticas governam amplamente a distribuição de espécies, mas as variáveis que captam as variações ambientais locais são relativamente importantes na área de estudo. A TSS foi significativamente influenciada pelo número de registros e complexidade dos modelos, enquanto a AUC sofreu com o efeito de todos os três fatores avaliados. A interação entre estes fatores é uma questão importante e a ser considerada em novas avaliações sobre ambas medidas e com diferentes técnicas de modelagem. Nossos resultados também mostraram que as distribuições de algumas espécies foram superestimadas e outras corresponderam bem com a ocorrência por nós conhecida. Efetivamente nossos resultados têm fundamentos para embasar novos levantamentos de campo, a avaliação de áreas prioritárias e planos de conservação, além de inferências dos efeitos de mudanças ambientais sobre as espécies da Mata Atlântica. / In view of environmental change on a global level, providing ecological information and getting a better understanding of the factors and processes that shape species distribution is an important initiative for planning conservation actions. In this context, the importance and lack of information about the geographical distribution of species motivated us to predict the potential species distribution of shrubs and trees of the family Lauraceae and Myrtaceae, in the Atlantic Forest in southern Brazil. Generalized linear models (GLM) were used to fit predictive models with records of occurrence of 88 species according to environmental variables. Predictor variables were selected based on the lowest corrected Akaike information criterion. We evaluate the performance of the models using the method of cross-validation (10-fold) to calculate the true skill statistic (TSS) and area under the receiver operator characteristic curve (AUC). We used GLM to test the influence of the area of occurrence estimated, the number of records of the species and the complexity of the models on the TSS and AUC. Our results show that climatic variables largely govern the distribution of species, but the variables that capture the local environmental variations are relatively important in the study area. The TSS was significantly influenced by the number of records and complexity of models while the AUC suffered from the effect of all three evaluated factors. The interaction between these factors is an important issue and be considered for new reviews on both measures and with different modeling techniques. Our results also showed that the distributions of some species were overestimated and other corresponded well with the occurrence known to us. Indeed our results have foundations to support new field surveys, assessment of priority areas and conservation plans, and inferences of the effects of environmental change on species of the Atlantic Forest.

abstract: This work investigates the effects of non-random sampling on our understanding of species distributions and their niches. In its most general form, bias is systematic error that can obscure interpretation of analytical results by skewing samples away from the average condition of the system they represent. Here I use species distribution modelling (SDM), virtual species, and multiscale geographically weighted regression (MGWR) to explore how sampling bias can alter our perception of broad patterns of biodiversity by distorting spatial predictions of habitat, a key characteristic in biogeographic studies. I use three separate case studies to explore: 1) How methods to account for sampling bias in species distribution modeling may alter estimates of species distributions and species-environment relationships, 2) How accounting for sampling bias in fossil data may change our understanding of paleo-distributions and interpretation of niche stability through time (i.e. niche conservation), and 3) How a novel use of MGWR can account for environmental sampling bias to reveal landscape patterns of local niche differences among proximal, but non-overlapping sister taxa. Broadly, my work shows that sampling bias present in commonly used federated global biodiversity observations is more than enough to degrade model performance of spatial predictions and niche characteristics. Measures commonly used to account for this bias can negate much loss, but only in certain conditions, and did not improve the ability to correctly identify explanatory variables or recreate species-environment relationships. Paleo-distributions calibrated on biased fossil records were improved with the use of a novel method to directly estimate the biased sampling distribution, which can be generalized to finer time slices for further paleontological studies. Finally, I show how a novel coupling of SDM and MGWR can illuminate local differences in niche separation that more closely match landscape genotypic variability in the two North American desert tortoise species than does their current taxonomic delineation. / Dissertation/Thesis / Doctoral Dissertation Geography 2018

Em vista das mudanças ambientais em nível global, disponibilizar informações ecológicas e buscar uma melhor compreensão dos fatores e processos que moldam a distribuição de espécies, é uma iniciativa importante para o planejamento de ações de conservação. Neste contexto, a importância e carência de informações sobre a distribuição geográficas das espécies nos motivaram a predizer a distribuição potencial de arbustos e árvores das famílias Lauraceae e Myrtaceae na Floresta Atlântica, no sul do Brasil. Modelos lineares generalizados (GLM) foram usados para ajustar modelos preditivos com os registros de ocorrência de 88 espécies em função de variáveis ambientais. As variáveis preditoras foram selecionadas com base no menor critério de informação de Akaike corrigido. Nós avaliamos o desempenho dos modelos usando o método de validação cruzada (10-fold) para calcular a habilidade estatística verdadeira (TSS) e a área sob a curva característica do operador receptor (AUC). Nós usamos GLM para testar a influência da área de ocorrência estimada, do número de registros das espécies e da complexidade dos modelos sobre a TSS e a AUC. Nossos resultados mostraram que as variáveis climáticas governam amplamente a distribuição de espécies, mas as variáveis que captam as variações ambientais locais são relativamente importantes na área de estudo. A TSS foi significativamente influenciada pelo número de registros e complexidade dos modelos, enquanto a AUC sofreu com o efeito de todos os três fatores avaliados. A interação entre estes fatores é uma questão importante e a ser considerada em novas avaliações sobre ambas medidas e com diferentes técnicas de modelagem. Nossos resultados também mostraram que as distribuições de algumas espécies foram superestimadas e outras corresponderam bem com a ocorrência por nós conhecida. Efetivamente nossos resultados têm fundamentos para embasar novos levantamentos de campo, a avaliação de áreas prioritárias e planos de conservação, além de inferências dos efeitos de mudanças ambientais sobre as espécies da Mata Atlântica. / In view of environmental change on a global level, providing ecological information and getting a better understanding of the factors and processes that shape species distribution is an important initiative for planning conservation actions. In this context, the importance and lack of information about the geographical distribution of species motivated us to predict the potential species distribution of shrubs and trees of the family Lauraceae and Myrtaceae, in the Atlantic Forest in southern Brazil. Generalized linear models (GLM) were used to fit predictive models with records of occurrence of 88 species according to environmental variables. Predictor variables were selected based on the lowest corrected Akaike information criterion. We evaluate the performance of the models using the method of cross-validation (10-fold) to calculate the true skill statistic (TSS) and area under the receiver operator characteristic curve (AUC). We used GLM to test the influence of the area of occurrence estimated, the number of records of the species and the complexity of the models on the TSS and AUC. Our results show that climatic variables largely govern the distribution of species, but the variables that capture the local environmental variations are relatively important in the study area. The TSS was significantly influenced by the number of records and complexity of models while the AUC suffered from the effect of all three evaluated factors. The interaction between these factors is an important issue and be considered for new reviews on both measures and with different modeling techniques. Our results also showed that the distributions of some species were overestimated and other corresponded well with the occurrence known to us. Indeed our results have foundations to support new field surveys, assessment of priority areas and conservation plans, and inferences of the effects of environmental change on species of the Atlantic Forest.