No full-text available
To read the full-text of this research,
you can request a copy directly from the authors.
Comparing models for predicting species' potential distributions: A case study using correlative and mechanistic predictive modelling techniques
Abstract
Models used to predict species’ potential distributions have been described as either correlative or mechanistic. We attempted to determine whether correlative models could perform as well as mechanistic models for predicting species potential distributions, using a case study. We compared potential distribution predictions made for a coastal dune plant (Scaevola plumieri) along the coast of South Africa, using a mechanistic model based on summer water balance (SWB), and two correlative models (a profile and a group discrimination technique). The profile technique was based on principal components analysis (PCA) and the group-discrimination technique was based on multiple logistic regression (LR). Kappa (κ) statistics were used to objectively assess model performance and model agreement. Model performance was calculated by measuring the levels of agreement (using κ) between a set of testing localities (distribution records not used for model building) and each of the model predictions. Using published interpretive guidelines for the kappa statistic, model performance was “excellent” for the SWB model (κ=0.852), perfect for the LR model (κ=1.000), and “very good” for the PCA model (κ=0.721). Model agreement was calculated by measuring the level of agreement between the mechanistic model and the two correlative models. There was “good” model agreement between the SWB and PCA models (κ=0.679) and “very good” agreement between the SWB and LR models (κ=0.786). The results suggest that correlative models can perform as well as or better than simple mechanistic models. The predictions generated from these three modelling designs are likely to generate different insights into the potential distribution and biology of the target organism and may be appropriate in different situations. The choice of model is likely to be influenced by the aims of the study, the biology of the target organism, the level of knowledge the target organism’s biology, and data quality.
... The versatility of SDMs has resulted in a plethora of modeling frameworks and parameterization choices. SDMs are broadly classified into correlative and mechanistic techniques (Connolly et al., 2017;Dormann et al., 2012;Robertson et al., 2003). Parameter responses in correlative models are not predefined and are instead modeled implicitly, resulting in responses that are not necessarily ecologically appropriate. ...
... Mechanistic models, on the other hand, employ explicit functions to characterize interactions between different components of the ecosystem (Connolly et al., 2017;Dormann et al., 2012). Correlative models, in contrast to mechanistic models, are frequently conceptually simple, capable of performing as well as or better than simple mechanistic models in estimation or short-term forecasting applications (Muhling et al., 2017;Robertson et al., 2003), and their independence from explicit assumptions can avoid confirmation biases (Connolly et al., 2017). Correlative models, on the other hand, are hampered in their ability to estimate conditions when system states change or display nonlinearity, such as under climate change (Lurgi et al., 2012;Plagányi et al., 2011Plagányi et al., , 2014. ...
... The versatility of SDMs has resulted in a plethora of modeling frameworks and parameterization choices. SDMs are broadly classified into correlative and mechanistic techniques (Connolly et al., 2017;Dormann et al., 2012;Robertson et al., 2003). Parameter responses in correlative models are not predefined and are instead modeled implicitly, resulting in responses that are not necessarily ecologically appropriate. ...
... Mechanistic models, on the other hand, employ explicit functions to characterize interactions between different components of the ecosystem (Connolly et al., 2017;Dormann et al., 2012). Correlative models, in contrast to mechanistic models, are frequently conceptually simple, capable of performing as well as or better than simple mechanistic models in estimation or short-term forecasting applications (Muhling et al., 2017;Robertson et al., 2003), and their independence from explicit assumptions can avoid confirmation biases (Connolly et al., 2017). Correlative models, on the other hand, are hampered in their ability to estimate conditions when system states change or display nonlinearity, such as under climate change (Lurgi et al., 2012;Plagányi et al., 2011Plagányi et al., , 2014. ...
Climate change is linked to the increase of natural disasters occurrence over the world. The increasing frequency of wildfire events constitutes a threat toward ecology, economy, and human lives. Hence, accurate information extracted from detailed burnt area cartography plays a primary role in the ecosystems’ preparation. Geoinformation has enabled rapid and low-cost Earth Observation data acquisition and their processing in cloud-based platforms such as Google Earth Engine (GEE). In the present study, a GEE-based approach that exploits machine learning (ML) techniques and ESA’s Sentinel-2 imagery is developed with the purpose of automating the mapping of burnt areas. As a case study is used an area located in the outskirts of Athens in Greece, on which it was occurred one of the largest wildfires in the summer of 2021. A Sentinel-2 image, obtained from GEE immediately after the fire event, was combined with ML classifiers for the purpose of mapping the burnt area at the fire-affected site. Validation of the derived products was based on the error matrix and the Copernicus Rapid Mapping operational product available for this fire event. Our study results clearly demonstrated the importance of rapid and accurate burnt area mapping exploiting sophisticated algorithms such as ML, when applied to Sentinel-2 imagery. Our findings provide valuable information regarding random forests and support vector machines classifiers through comparisons during their implementation as well as the potential of automated mapping in cloud-based platforms such as GEE. All in all, burnt area mapping is very valuable toward preparation activities regarding postfire management in future fire incidents.
... The most important models include GLM (Generalized Linear Models), GAM (Generalized Additive Models), ANNs (Artificial Neural Networks), PCA (Principal Components Analysis), and CCA (Canonical Correspondence Analysis) [10][11][12][13]. Some models such as GLM and GAM are used to determine the likeliness of species presence, and others such as CCA, ANNS, and PCA are used to examine the factors affecting species distribution and the spatial prediction of the habitat suitable for the establishment of the target species [14,15]. CCA and GAM are among the methods most used for analyzing the reaction of plant species to environmental factors [3,16]. ...
... CCA and GAM are among the methods most used for analyzing the reaction of plant species to environmental factors [3,16]. ers such as CCA, ANNS, and PCA are used to examine the factors affecting species distribution and the spatial prediction of the habitat suitable for the establishment of the target species [14,15]. CCA and GAM are among the methods most used for analyzing the reaction of plant species to environmental factors [3,16]. ...
The present study investigates the ecological requirements of Astragalus curvirostris Boiss., with emphasis on determining the ecological factors that affect the distribution of plant species, and the species’ response to changes in ecological factors using a Generalized Additive Model (GAM) in the Iranian Province of Zanjan from 2017 to 2019. Randomized-systematic sampling was used to collect vegetation data. Data analysis was performed using SPSS17 and CANOC4.5 software. The results showed that the growth and development of A. curvirostris change according to environmental factors linked to the composition of the soil and the variety of the other species present. This model is indicative of a competitive limitation along the environmental gradient. By understanding all environmental parameters, the necessary steps could be taken towards planning proper management programs, including rangeland grazing management and determining the proper moment for seed collection, which will result in the conservation, improvement, and restoration of rangelands.
... There is some debate over the use of correlative models instead of mechanistic models to study the potential distribution of species, because some researchers argue that the correlation between the species and environment may cease to exist or may change in future decades. It has also been suggested that, for this reason, correlative models will provide more robust results compared to mechanistic models under future climate scenarios [31,32]. However, despite their usefulness, mechanistic models are hard to build, time consuming, and they require a good knowledge and background information about target species [31,32]. ...
... It has also been suggested that, for this reason, correlative models will provide more robust results compared to mechanistic models under future climate scenarios [31,32]. However, despite their usefulness, mechanistic models are hard to build, time consuming, and they require a good knowledge and background information about target species [31,32]. Many SDMs have been established in the world to predict species distribution areas, but few models can be effectively used for endangered species and rare species with limited distribution data [33]. ...
Future climate change will have serious impacts on species survival and distribution and will likely lead to the extinction of some species classified as endangered. Carpinus tientaiensis (Betulaceae), a unique and endangered species in China, has restricted distribution and a small population, indicating an urgent need for its protection. However, research on its current distribution or the influence that climate change will have on its future survival and distribution is limited. We used a MaxEnt model and ArcGIS software to predict the current and future niches of C. tientaiensis. The current suitable distribution area of C. tientaiensis is small, mainly in east China, south Zhejiang and Anhui, and central and southern mountainous areas of Taiwan province. The core suitable areas are concentrated in the Xianxialing and Kuocang mountains in south Zhejiang, the southern mountains of Taiwan, and the Dabie, Huangshan and Jiuhua mountains in south Anhui. Among the 15 BIOCLIM variables examined, the precipitation of the driest quarter (bio17) was found to be the most important factor limiting C. tientaiensis survival and distribution. Future field investigations will focus on the Xianxialing and Kuocang mountains, as they may have unidentified wild C. tientaiensis communities. In the future, the Kuocang, Dapan and Tiantai mountains in east Zhejiang, and the high-altitude areas of Dabie and Jiuhua mountains in south Anhui, will be suitable for C. tientaiensis ex situ conservation and cultivation. However, the suitable distribution and core suitable areas for C. tientaiensis will decrease sharply as they are susceptible to climate shocks. Moreover, the suitable distribution area of C. tientaiensis is predicted to move slightly north and obviously eastward. Therefore, we suggest that strengthen conservation and management efforts for C. tientaiensis in its original habitats, and actively carry out ex situ conservation and artificial breeding in botanical gardens.
... Correlative modelling of vegetation patterns enables the examination of the probable importance of environmental drivers and their effects on vegetation with static survey data (e.g. [1,2]). In addition, the models can be used to project vegetation patterns in future conditions ( [1,2]). ...
... [1,2]). In addition, the models can be used to project vegetation patterns in future conditions ( [1,2]). Recent research highlights the importance of the incorporation of all relevant factors into the models and measuring them at the considered geographical scale (e.g. ...
This study aimed at illustrating how direct measurements, mobile laser scanning and hydraulic modelling can be combined to quantify environmental drivers, improve vegetation models and increase our understanding of vegetation patterns in a sub-arctic river valley. Our results indicate that the resultant vegetation models successfully predict riparian vegetation patterns (Rho = 0.8 for total species richness, AUC = 0.97 for distribution) and highlight differences between eight functional species groups (Rho 0.46–0.84; AUC 0.79–0.93; functional group-specific effects). In our study setting, replacing the laser scanning-based and hydraulic modelling-based variables with a proxy variable elevation did not significantly weaken the models. However, using directly measured and modelled variables allows relating species patterns to e.g. stream power or the length of the flood-free period. Substituting these biologically relevant variables with proxies mask important processes and may reduce the transferability of the results into other sites. At the local scale, the amount of litter is a highly important driver of total species richness, distribution and abundance patterns (relative influences 49, 72 and 83%, respectively) and across all functional groups (13–57%; excluding lichen species richness) in the sub-arctic river valley. Moreover, soil organic matter and soil water content shape vegetation patterns (on average 16 and 7%, respectively). Fluvial disturbance is a key limiting factor only for lichen, bryophyte and dwarf shrub species in this environment (on average 37, 6 and 10%, respectively). Fluvial disturbance intensity is the most important component of disturbance for most functional groups while the length of the disturbance-free period is more relevant for lichens. We conclude that striving for as accurate quantifications of environmental drivers as possible may reveal important processes and functional group differences and help anticipate future changes in vegetation. Mobile laser scanning, high-resolution digital elevation models and hydraulic modelling offer useful methodology for improving correlative vegetation models.
... SDMs can broadly be categorized into correlative and mechanistic approaches (Robertson et al. 2003, Dormann et al. 2012, Connolly et al. 2017. In correlative models, parameter responses are not pre-defined and are instead modeled implicitly, resulting in responses that are not always ecologically reasonable. ...
... Whereas, mechanistic models use explicit functions to characterize relationships among different components of the ecosystem and are usually defined a priori based on ecological theory (Dormann et al. 2012, Connolly et al. 2017. Relative to mechanistic models, correlative models are often conceptually simple, are capable of performing as well or better than simple mechanistic models in estimation or near-term forecasting applications (Robertson et al. 2003, Muhling et al. 2016, and their independence from explicit assumptions can avoid confirmation biases (Connolly et al. 2017). However, correlative models can be limited in their capacity to estimate conditions when system states change or exhibit non-linearity, such as under climate-driven change (Plagányi et al. 2011, Lurgi et al. 2012). ...
Species distribution models (SDMs) are a common approach to describing species’ space‐use and spatially‐explicit abundance. With a myriad of model types, methods and parameterization options available, it is challenging to make informed decisions about how to build robust SDMs appropriate for a given purpose. One key component of SDM development is the appropriate parameterization of covariates, such as the inclusion of covariates that reflect underlying processes (e.g. abiotic and biotic covariates) and covariates that act as proxies for unobserved processes (e.g. space and time covariates). It is unclear how different SDMs apportion variance among a suite of covariates, and how parameterization decisions influence model accuracy and performance. To examine trade‐offs in covariation parameterization in SDMs, we explore the attribution of spatiotemporal and environmental variation across a suite of SDMs. We first used simulated species distributions with known environmental preferences to compare three types of SDM: a machine learning model (boosted regression tree), a semi‐parametric model (generalized additive model) and a spatiotemporal mixed‐effects model (vector autoregressive spatiotemporal model, VAST). We then applied the same comparative framework to a case study with three fish species (arrowtooth flounder, pacific cod and walleye pollock) in the eastern Bering Sea, USA. Model type and covariate parameterization both had significant effects on model accuracy and performance. We found that including either spatiotemporal or environmental covariates typically reproduced patterns of species distribution and abundance across the three models tested, but model accuracy and performance was maximized when including both spatiotemporal and environmental covariates in the same model framework. Our results reveal trade‐offs in the current generation of SDM tools between accurately estimating species abundance, accurately estimating spatial patterns, and accurately quantifying underlying species–environment relationships. These comparisons between model types and parameterization options can help SDM users better understand sources of model bias and estimate error.
... Various statistical techniques are available for modelling species distribution. Correlative models are particularly applied to cases where an initial prediction of the potential distribution of a given species is required, especially when the biology of the species is not well known [15]. Those models that use both presence and absence locality records to make predictions have been referred to as group discrimination techniques. ...
... Those models that use both presence and absence locality records to make predictions have been referred to as group discrimination techniques. Examples of group-discrimination techniques include those models based on discriminant analysis (DA), logistic regression analysis (LR), classification tree technique (CT) and generalised additive model (GAM) [15]. ...
... Such data could be used to build mechanistic models, which could then be examined in combination with correlative models like SDM (e.g. Robertson et al. 2003;Tourinho and Vale 2023). ...
Determining species' distributions is challenging for cryptic species that are difficult to detect using standard techniques. The mallee worm-lizard (Aprasia inaurita Kluge, 1974) is a cryptic reptile in the family Pygopodidae, listed as Endangered in New South Wales. We modelled the species' potential distribution (Maxent) to improve understanding of the species' distribution and surveyed potential habitat in the Scotia Mallee region (an area with suitable habitat) from 2018 to 2022, with pitfall traps and artificial refuges (terracotta roof tiles). We completed 11 587 pitfall trap-nights and 3200 tile checks over eight monitoring sessions. Over this period, we detected six vertebrate species (all lizards) using roof tiles and 40 species with pitfall traps, but no mallee worm-lizards. Evaluation of existing records of the mallee worm-lizard from NSW suggested that the state constitutes the northeastern edge of its continental range, with the species apparently present in low numbers across a wide swathe of southwestern NSW. Most records were located within or near to spinifex or porcupine grass (Triodia spp.) communities, on private land. Species distribution modelling provided outputs that are useful for spatial prioritisation of conservation efforts for the species, with region-wide maps showing that much of the Scotia Mallee study area contains potentially suitable habitat for the mallee worm-lizard. However, habitat suitability scores for individual cells in this area were low, in some instances, because of high maximum summer temperatures and soil available water capacity. We anticipate that increasing temperatures associated with climate change may further reduce the suitability of habitat in this area in the future.
... Correlative models evaluate the relationship between biological and environmental variables, often working with species distribution and abundance data across spatial and temporal gradients (ecological niche modeling - Soberon and Nakamura, 2009). While correlative models can offer greater precision, due to their simplicity (e. g. evapotransmission of coastal dune plants - Robertson et al., 2003), models that include mechanistic linkages between the functional traits of organisms and their environments (Kearney et al., 2010) may offer insights into biological patterns that may be more generalizable to novel future scenarios (Buckley et al., 2010;Kearney and Porter, 2009). Integrating these methodologies may better inform climate predictions, the uncertainty surrounding these predictions, and future key research areas necessary for gaining a better understanding what to expect in future novel scenarios. ...
Anthropogenic warming of the ocean and atmosphere, concurrent with ocean acidification and deoxygenation, has made it even more pressing to quantify the link between environmental stressors and marine organism population dynamics. In marine environments, low food availability, low feeding rates, and/or increased metabolic costs can cause energetic limitation. Energetic limitation affects some functional traits, such as growth rates of body tissue and reproductive output. Other functional traits that are linked with short-term survival are often prioritized in conditions of energy limitation, despite their energetic cost. Mussels are ecosystem engineers in rocky shore ecosystems, and they produce byssal threads to attach to hard substrate and aquaculture line. Previous studies of mytilid mussel bioenergetics suggest tissue and shell growth are energetically-constrained, while production of byssal threads presents a fitness trade-off and could potentially be a fixed or 'constitutive' response regardless of energetic state. In this study, we conduct a field test with two congener mussel species, Mytilus trossulus and Mytilus galloprovincialis to determine whether an index of energetic availability, scope for growth (SFG), correlates with growth and byssal thread production, and the extent to which other potential stressors (hypoxia, low pH, low salinity and high temperature) modulate this response. We find a positive correlation between SFG and growth (both tissue and shell) but not the number of byssal threads produced. We also find low pH or low DO, two co-varying physiological stressors, negatively affect tissue growth of both species, but only marginally affect byssal thread production. We also observed mortality in the late summer/early autumn that coincides with the period of greater hypoxia and low pH. Overall, this work suggests that some functional traits, such as shell and tissue growth, are energetically-constrained while other functional traits, such as mussel byssal thread production, may be best described as a fitness trade-off.
... The second model is process based mechanistic approach which are based on detailed knowledge of parasite physiology, and it has attempted to copy the basic mechanisms that drive the parasite's response to environmental components (Robertson et al., 2003). In practice, it is very hard to find distinction between correlative and mechanistic modelling techniques and therefore, an integration of both techniques may be needed for future improvements in prediction. ...
... SDMs are valuable tools for the evaluation and protection of regions degrading and losing their biodiversity due to various factors (Kosanic et al., 2018). Robertson (2003) suggested that the prediction provided by each model may present different conceptions of the potential distribution and biology of the target species. Despite that, it is essential to perfectly understand restrictions and ambiguities embedded in species distribution modeling to produce suitable and precise models (Zimmermann et al., 2010;Kumar, 2012;Zomer et al., 2015;Akhter et al., 2017). ...
Various statistical techniques have been used for species distribution modeling that attempt to predict the occurrence of a given species with respect to environmental conditions. The current study was conducted to compare the performance of three regression-based models (multivariate adaptive regression splines, generalized additive models, and generalized linear models) with three machine-learning algorithms (random forest, artificial neural networks, and generalized boosted models). Also in this study, three sets of explanatory variables (climate-only, topography-only and combined topography-climate) for each species (i.e. Achillea millefolium, Festuca rupicola, and Centaurea jacea) were quantified and the effect of the interaction of the predictor variables with the modeling approaches on determining the accuracy of the predictions was tested. Model accuracy was evaluated using the area under the curve (AUC) of the receiver operating characteristics and true skill statistics (TSS). It was found that regression-based approaches, especially generalized additive model, performed better than those of machine-learning. The results showed that the topography-climate variables were the most important for mapping potentially suitable habitats of target species. The response curves associated with these variables indicate that there are ecological thresholds for favorable growth of all plant species studied.
... Segundo De Marco Júnior e Siqueira (2009), a qualidade do dado usado na modelagem é fundamental para o sucesso do resultado final. A qualidade e a quantidade dos dados de distribuição afetam fortemente os resultados da modelagem (Suarez-Seoane et al., 2002), assim como a resolução e escolha das variáveis ambientais (Robertson et al., 2003). O aumento do número de pontos aumenta a acurácia dos modelos (Stockwell;Peterson, 2002) e a seleção de um adequado conjunto de pontos pode afetar mais o resultado dos modelos do que a seleção do melhor algoritmo. ...
Este capítulo de livro aborda o uso de modelos de nicho na predição de ocorrência de espécies arbóreas, entre as quais a araucária, evidenciando a necessidade de sua conservação. A mudança de uso da terra e as mudanças climáticas representam forte ameaça à sobrevivência da espécie, típica de clima frio do sul do Brasil, e que já vem sofrendo com o aumento de temperatura e alteração do regime hídrico, correndo o risco de perder grande parte de sua diversidade genética, principalmente nas áreas marginais, de transição climática. Estas serão as primeiras áreas a serem afetadas com as ameaças do clima, onde a espécie concentra maior diversidade genética e onde existem populações únicas, sem similares em outras regiões, ocorrendo perdas inestimáveis.
... Só são usados quando se tem certeza de que a espécie não ocorre em uma dada região. , 2002), assim como a resolução e escolha das variáveis ambientais (Robertson et al., 2003). O aumento do número de pontos aumenta a acurácia dos modelos (Stockwell;Peterson, 2002) e a seleção de um adequado conjunto de pontos pode afetar mais o resultado dos modelos do que a seleção do melhor algoritmo. ...
O livro trata da pesquisa e desenvolvimento da Araucaria angustifolia no Brasil, com 17 capítulos referentes aos mais diversos temas do pinheiro do Paraná. Nosso capítulo refere-se à morfologia da raiz, caule, madeira, folhas, frutos e sementes de Araucária. Nesse capítulo consideramos o pinhão de Araucária como fruto, concordando com o pesquisador Hertel (in memorian) da Universidade Federal do Paraná.
... We have utilised and presented different habitat suitability models that: promotes consistency and transparency in model development (Elith and Graham, 2009); allows for modifications in invasiveness and filters in space and time (Elith and Leathwick, 2009); assimilated mechanistic and correlative modelling techniques (Pertierra et al., 2020;Robertson et al., 2003). The lack of ecologically applicable predictors hinders the accurate predictions of the species distribution models in delivering accurate and robust outputs (Regos et al., 2019). ...
Conservation strategies need reliable information on species distribution and habitat suitability. Modelling the invasion probability of species is required for invasive alien species management. The present study aims in predicting the suitable habitat and potential distribution of invasive plant species - Chromolaena odorata and Lantana camara. We used the maximum entropy (MaxEnt), Random forest, Surface range envelope, and Boosted regression tree analysis to identify the suitable habitats for potential environmental distribution of invasive alien species in the Eastern Ghats Visakhapatnam district, Andhra Pradesh, India. The Area under Curve (AUC) for each model has been estimated for predicting the potentially suitable habitats of invasion. The results suggested that the MaxEnt model, Random forest, boosted regression tree, and Surface range envelope were suitable approaches in identifying areas of potential habitats of selected invasive alien species. This study aids in developing long-term conservation practices for the management of invasive alien species.
... Esta prueba es aplicable cuando se encuentran las mismas clases de datos en los mapas por analizar y puede ser usada objetivamente para estimar el nivel de concordancia entre los datos observados y los predichos (Robertson et al, 2003). ...
The study characterized the habitat of five threatened cacti species in Aguascalientes, Mexico, and determined their potential distribution. Geographic information systems and modeling were used for these purposes.
... La inclusión de las interacciones bióticas (depredación, competencia, polinización, parasitismo, simbiosis, etc.) y las características ecológicas de la especie a modelar (McPherson y Jetz, 2007) mejoran el nivel predictivo de los modelos, principalmente a escala local (Pearson et al., 2004;Elith y Leathwick, 2009), aunque esta información es conocida para unas pocas especies. En el caso de especies con biología poco conocida se pueden realizar análisis de correlación entre las especies y las variables de su hábitat para identificar el hábitat esencial o de alta calidad para la especie (Robertson et al., 2003). Sin embargo, hay que tener en cuenta algunas de las limitantes de la correlación, como la reducción del número de variables, la escala (resolución y extensión) a la cual se mide la respuesta de la especie y que puede variar (Wiens, 1989), además la correlación no necesariamente explica la causa o el proceso. ...
En esta investigación se actualiza el conocimiento de Chalcididae en México, se evalúa la
distribución geográfica de los registros, se estima su riqueza espacial y distribución
potencial de la familia y géneros. Se realizó una revisión de material biológico en
colecciones, museos, bases de datos y literatura, con lo que se elaboró un catálogo y una
base de datos georeferenciada con registros de subfamilias, géneros y especies. Ésta
información se utilizó para evaluar la distribución geográfica de los registros de la familia y estimar la riqueza genérica y específica en las entidades federativas, provincias
biogeográficas y ecoregiones de la República Mexicana. Se realizó una revisión del marco
conceptual y aspectos a considerar en las etapas básicas de la construcción de modelos
predictivos de especies y su posible aplicación con avispas parasitoides en el contexto del control biológico. Se aplicaron tres modelos predictivos de especies con la información de la base de datos georeferenciada. Los principales resultados son el registro de 3 subfamilias, 15 géneros y 173 especies, de los cuales 3 géneros y 38 especies son primeros reportes para México. Los estados de Yucatán, Morelos, Jalisco, Veracruz, Baja California Sur y Chiapas registran el mayor número de especies. La mayor riqueza genérica y específica se registró en las provincias biogeográficas Costa Pacífico y Yucatán, y las ecoregiones Planicie Noroccidental de Yucatán con Selva Caducifolia, la Planicie Occidental Yucateca con Selva Caducifolia y Lomeríos y Piedemontes del Pacífico Sur Mexicano con Selva Espinosa. La distribución potencial de Chalcididae sugiere que la familia tenga una distribución amplia en la región Neotropical, en la vegetación de selvas, principalmente a bajas altitudes (≈ 300 m), aunque los géneros tienen distinto grado de distribución y en algunos casos coincide con su distribución conocida o se amplía. Para mejorar el nivel de representatividad de los modelos se necesita incluir variables distintas a las climáticas, utilizar datos independientes a los usados para generarlos y corroborar en campo la exactitud de predicciones de los modelos.
Palabras clave: Chalcididae, distribución geográfica, riqueza espacial, distribución
potencial, provincias biogeográficas, ecoregiones, México.
... topographic, landscape and climatic factors) is often quantified using predictive modelling (Guisan & Zimmermann 2000) and is essential to guide appropriate conservation decisions (Larson et al. 2004, Nams et al. 2006. In recent decades, species distribution modelling has been commonly used to identify suitable habitat and to predict a potential distribution of target taxa (Robertson et al. 2003, Rushton et al. 2004, Matyukhina et al. 2015, with particular emphasis on threatened species (Guisan & Zimmermann 2000). ...
Capsule: Golden Eagles Aquila chrysaetos in Sardinia are clustered across the main mountain
ranges of the island, with a preference for undisturbed and homogeneous inland habitats.
Aims: To analyse habitat preferences of the Golden Eagle in Sardinia, Italy, at the landscape and
home range spatial scales.
Methods: Landscape scale habitat preferences were analysed using the 10 × 10 km Universal
Transverse Mercator grid and the home range scale was based on the spatial distribution of
breeding territories. Generalized linear models were fitted with three different sets of
environmental predictors (topographic, bioclimatic and land use variables) to analyse the spatial
distribution of Golden Eagles with a case–control design.
Results: Eagles showed a preference for rugged and elevated areas, characterized by a certain
degree of humidity and surrounded by areas of forest. The distribution of Golden Eagles on this
Mediterranean island was negatively affected by the occurrence of arable farmland and coastal
areas, as well as by the effects of habitat fragmentation.
Conclusions: The results of this study could contribute to future management strategies and
conservation projects aimed to protect this species, and may be used to identify the most
suitable conservation areas for this and other competing species, such as the Bonelli’s Eagle
Aquila fasciata, which is currently the subject of a reintroduction project in Sardinia.
... Analysis of interrelationship between plants and environmental factors has always been considered as a primary issue in ecological studies (Guisan and Zimmermann, 2000). Nowadays, various methods have been initiated to study the ecological factors and the relationship between plant species dispersion and environmental factors, some for the possibility of species presence and others to investigate factors affecting species distribution and spatial prediction of suitable habitat for species establishment (Bakkenes, 2002;Peterson 2001;Berg et al., 2004;Guisan et al., 2002;Robertson, 2003;Engler et al., 2004). ...
Environmental factors have major impact on the distribution and yield of plant species. For this purpose, the responses of Khorasanian sainfoin (Onobrychis chorassanica Bunge.) were evaluated with regard to some environmental factors in habitats of Khorasan Razavi province in 2018-2020. Generalized Additive Model (GAM) was used to investigate the response of this species to soil and topographic factors. The results indicated that O. chorassanica exhibited a substantial response to some environmental factors in its habitat. The response pattern of this species includes the gradient of soil, Total Neutralizing Value (TNV) as well as soil clay percent, followed by the monotonic increase model. Therefore, with the increase in the values of these factors, its vegetation cover percent increased. In contrast, the response of this species along with the gradient of soil sand percent followed the monotonic decrease model and with enhancement of the factor amount, the presence of O. chorassanica decreased. Soil studies have revealed that this species is mainly distributed on loamy to sandy loam soils. The response pattern of O. chorassanica along with the gradient of Organic Carbon content (OC%) and soil litter percent complied with unimodal model and its optimal growth levels for these factors were 0.4% and 30%, respectively. The geographical response of this plant also displayed that the vegetation cover percent of the studied species increased in the western and north aspects and is rarely apparent in the southern and southeastern aspects. Investigation of O. chorassanica response on gradient of topographic and soil factors provided valuable information for determining ecological needs of this species, which can be considered in vegetation management and range improvement operations in similar areas.
... The development of insects is a function of temperature over time (Jarošík, Honěk, Magarey, & Skuhrovec, 2011). Under climate change, temperatures in Europe, particularly at higher latitudes, Newman, 2005;Robertson, Peter, Villet, & Ripley, 2003). ...
Climate change and globalization affect the suitable conditions for agricultural crops and insect pests, threatening future food security. It remains unknown whether shifts in species’ climatic suitability will be linear or rather non‐linear, with crop exposure to pests suddenly increasing when a critical temperature threshold is crossed. Moreover, uncertainty of forecasts can arise because of the modelling approach based either on species distribution data or on physiological measurements. Here, we compared the predictions of two modelling approaches (physiological models and species distribution models) for forecasting the potential distribution of agricultural insect pests in Europe. Despite conceptual differences, we found good agreement overall between the two approaches. We further identified a potential regime change in pest pressure along a temperature gradient. With both modelling approaches, we found an inflection point in the number of pest species with suitable climatic conditions around a minimum temperature of the coldest month of ‐3°C. Our results could help decision‐makers anticipate the onset of rising pest pressure and provide support for intensifying surveillance measures, particularly in regions where temperatures are already beyond the inflection point.
... When running the model, random data partitioning 75% of the training data and 25% of the testing data were simulated and repeated three times. Meanwhile, the Kappa statistic (K), another most frequently used prediction accuracy index (Robertson et al. 2003), was also applied. Both AUC and Kappa have values range from 0 to 1; the higher the values, the closer the model agrees with the data. ...
Subtropical forest in China has received much attention due to its complex geologic environment and bioclimatic heterogeneity. There have been very few studies addressing which climatic factors have shaped both distribution patterns and niche differentiation of species from this region. It also remains unclear whether phylogenetic niche conservatism retains in plant species from this biodiversity-rich subtropical region in China. In this study, we used geographic occurrence records and bioclimatic factors of Prunus dielsiana (Rosaceae), a wild cherry species, combined with the classical ENM-based DIVA-GIS software to access contemporary distribution and richness patterns of its natural populations. The current distribution of P. dielsiana occupied a relatively wide range but exhibited an uneven pattern eastward in general, and the core distribution zone of its populations are projected to concentrate in the Wushan and Wuling Mountain ranges of western China. Hydrothermic variables, particularly the Temperature Seasonality (bio4) are screened out quantitatively to be the most influential factors that have shaped the current geographical patterns of P. dielsiana. By comparison with other sympatric families, climatic niche at regional scale showed a pattern of phylogenetic niche conservatism within cherry species of Rosaceae. The effect of habitat filtering from altitude is more significant than those of longitude and latitude. We conclude that habitat filtering dominated by limiting hydrothermic factors is the primary driving process of the diversity pattern of P. dielsiana in subtropical China.
... Both correlative and mechanistic modelling approaches have been applied in exploring organism-microclimate interactions (Beerling et al., 1995;Kearney et al., 2010b;Robertson et al., 2003). Correlative models statistically link macroclimate variables (such as ambient temperature and sea surface temperature) and other environmental variables (e.g. ...
The continual development of ecological models and availability of high-resolution gridded climate surfaces have stimulated studies that link climate variables to functional traits of organisms. A primary constraint of these studies is the ability to reliably predict the microclimate that an organism experiences using macroscale climate inputs. This is particularly important in regions where access to empirical information is limited. Here, we contrast correlative models based on both ambient and sea surface temperatures to mechanistic modelling approaches to predict beach sand temperatures at depths relevant to sea turtle nesting. We show that mechanistic models are congruent with correlative models at predicting sand temperatures. We used these predictions to explore thermal variation across 46 mainland and island beaches that span the geographical range of sea turtle nesting in Western Australia. Using high resolution gridded climate surfaces and site-specific soil reflectance, we predict almost 9 °C variation in average annual temperatures between beaches, and nearly 10 °C variation in average temperatures during turtle nesting seasons. Validation of models demonstrated that predictions were typically within 2 °C of observations and, although most sites had high correlations (r2 > 0.7), predictive capacity varied between sites. An advantage of the mechanistic model demonstrated here is that it can be used to explore the impacts of climate change on sea turtle nesting beach temperatures as, unlike correlative models, it can be forced with novel combinations of environmental variables.
... This would be helpful in expanding the vulture conservation area to be used by future populations after facilitative intervention, to increase the population from the current level. The species distribution modelling (a tool for conservation planning and resource management), owing to the broad use of geographical information systems in recent years, could be used for identifying suitable habitat and to predict potential distributions of vulture populations ( Robertson et al., 2003, Rushton et al., 2004Vaz et al., 2008). ...
Conflicts between people and elephants are common in some parts of tropical Asia.
This may occur when the elephant population is concentrated or increases, in regions
with high human population densities or intensive land use. Dalma Wildlife Sanctuary,
located near Jamshedpur city in the Jharkhand State of India is known for its elephant
(Elephas maximus, Linnaeus, 1758) habitat and large elephant population. The Dalma
elephants migrate between the forested areas located in the border regions of Jharkhand,
West Bengal, Odisha as per food and water availability, and climatic conditions. The
present study in Dalma Wildlife Sanctuary and surroundings examines the migration
patterns, recent migration systems, land use and land cover change for the period 1989 to
2014 and assesses suitable areas for Dalma elephants. The study is based on field
surveys, published records, and image-based analyses. The objective is to identify
potential conflict spots which will support appropriate planning to avoid and reduce
human-elephant conflict in the study area. Geomatics-based analyses of such issues are
vital for the study of the conservation biology of elephants in India and for understanding
the general issues of human-large wildlife conflict.
... Two main approaches are used for modeling the structure and dynamics of the geographic ranges of invasive species (Robertson, Peter, Villet, & Ripley, 2003). Mechanistic-based distribution models use inherent physiological and/or demographic characteristics to better capture the processes underpinning species distributions (Fordham, Akçakaya, Araújo, Keith, & Brook, 2013;Kearney & Porter, 2004). ...
In its invasive range in Australia, the European rabbit threatens the persistence of native flora and fauna and damages agricultural production. Understanding its distribution and ecological niche is critical for developing management plans to reduce populations and avoid further biodiversity and economic losses.
We developed an ensemble of species distribution models (SDMs) to determine the geographic range limits and habitat suitability of the rabbit in Australia. We examined the advantage of incorporating data collected by citizens (separately and jointly with expert data) and explored issues of spatial biases in occurrence data by implementing different approaches to generate pseudo‐absences. We evaluated the skill of our model using three approaches: cross‐validation, out‐of‐region validation, and evaluation of the covariate response curves according to expert knowledge of rabbit ecology.
Combining citizen and expert occurrence data improved model skill based on cross‐validation, spatially reproduced important aspects of rabbit ecology, and reduced the need to extrapolate results beyond the studied areas.
Our ensemble model projects that rabbits are distributed across approximately two thirds of Australia. Annual maximum temperatures >25°C and annual minimum temperatures >10°C define, respectively, the southern and northern most range limits of its distribution. In the arid and central regions, close access to permanent water (≤~ 0.4 km) and reduced clay soil composition (~20%–50%) were the major factors influencing the probability of occurrence of rabbits.
Synthesis and applications. Our results show that citizen science data can play an important role in managing invasive species by providing missing information on occurrences in regions not surveyed by experts because of logistics or financial constraints. The additional sampling effort provided by citizens can improve the capacity of SDMs to capture important elements of a species ecological niche, improving the capacity of statistical models to accurately predict the geographic range of invasive species.
... Devido à complexidade do processo de modelagem, no que tange previsões confiáveis, diferentes abordagens de algoritmos e métodos têm sido aplicadas, como exemplo, os métodos estatísticos (modelos lineares generalizados -GLM e modelos aditivos generalizados -GAM) e de aprendizagem de máquina (redes neurais artificiais -RNA, suport vector machine -SVM, árvores de decisão -CART, random forest -RF e entropia máxima -MAXENT), conforme observado nos trabalhos de Paglia et al. (2012); Merow et al. (2014);García-Callejas;Araújo (2016). Pesquisas comprovam que o desempenho dos algoritmos varia de acordo com os dados referentes às espécies e sua distribuição espacial (ROBERTSON et al., 2003;CARVALHO et al., 2017). Não há um consenso de qual o melhor método, já que existem variações dessa natureza, o que decorre em uma lacuna sobre qual a melhor técnica de modelagem e qual algoritmo possui desempenho superior. ...
O estudo teve como objetivo avaliar três métodos de aprendizagem de máquina (árvore de decisão-J48, random forest e redes neurais artificias), na modelagem da distribuição de dez espécies arbóreas mais abundantes em uma sub-bacia do rio São Francisco (MG). Utilizaram-se dados provenientes do Inventário Florestal de Minas, com total de 77 fragmentos amostrados e 2.234 parcelas, nas quais foram computadas a presença/ausência de cada espécie. Empregaram-se 12 variáveis ambientais categóricas procedentes do Zoneamento Ecológico Econômico de Minas Gerais (ZEE/MG), além de variáveis relacionadas ao balanço hídrico do solo (evapotranspiração atual e potencial, aridez e índice alpha). A parametrização dos três algoritmos para as dez espécies selecionadas foi feita com o auxílio do algoritmo cv parameter do software WEKA. Os resultados mostram que os algoritmos testados apresentaram desempenhos estatisticamente iguais em 60% das espécies arbóreas. Os algoritmos random forest e multilayer perceptron foram estatisticamente iguais para a espécie Eugenia dysenterica, sendo superiores ao algoritmo J48. Contudo, o algoritmo random forest foi superior aos demais para as três espécies do gênero Qualea. Conclui-se que o algoritmo random forest apresentou-se como o mais robusto para a modelagem da distribuição potencial de habitat de espécies arbóreas.
Palavras-chave: inteligência artificial; árvore de decisão; random forest; redes neurais artificiais.
MACHINE LEARNING ALGORITHMS FOR MODELING THE POTENTIAL DISTRIBUTION HABITAT OF TREE SPECIES
ABSTRACT:
The aim of the present study was to evaluate three methods of machine learning (decision tree-J48, random forest and artificial neural networks) to model the potential habitat distribution of the ten most abundant tree species of the São Francisco river watershed. The presence/absence tree species data were from 77 fragments sampled with 2,234 plots. We used 12 categorical environmental variables from the Economic Ecological Zoning of Minas Gerais (ZEE/MG), as well as variables related to soil water balance (current and potential evapotranspiration, aridity and alpha index). The parameterization of the three algorithms was done with cv parameter algorithm of the WEKA software. The results showed the applied algorithms were statistically similar for 60% of the tree species. The random forest and multilayer perceptron algorithms were statistically similar considering the Eugenia dysenterica and superior to J48 algorithm. However, the random forest algorithm was superior to the other for the three species of Qualea genera. The conclusion is the random forest was the most robust model for the potential distribution habitat of tree species.
Keywords: artificial intelligence; decision trees; random forest; artificial neural networks.
... This would be helpful in expanding the vulture conservation area to be used by future populations after facilitative intervention, to increase the population from the current level. The species distribution modelling (a tool for conservation planning and resource management), owing to the broad use of geographical information systems in recent years, could be used for identifying suitable habitat and to predict potential distributions of vulture populations ( Robertson et al., 2003, Rushton et al., 2004Vaz et al., 2008). ...
This chapter examines the development of Global Positioning Systems (GPS) for animal tracking and their applications to the key issues of conservation biology. It also presents case studies of wildlife conservation and enumerates parameters that would be the focus of advanced GPS and associated techniques. GPS are the basis for advanced techniques in animal tracking and compare favorably with other tracking methods. Cases are cited from the relevant literature, and hypothetical cases are also examined, where ecological problems may be solved using tracking devices to record measurable parameters. The object would be to record the movement of animals, as individuals or as part of migratory patterns, to determine their ecological relations with their encountered environments and the spatial dynamics that may affect their presence and movements. This involves the examination of the parameters that would be used for the applications of geomatics techniques, especially recent, advanced techniques for tracking and locational analyses of target animals. It is concluded the GPS systems are vital for current studies of conservation biology, but such studies and methods must be supported by related geomatics methods, parameter identification and social studies.
... This would be helpful in expanding the vulture conservation area to be used by future populations after facilitative intervention, to increase the population from the current level. The species distribution modelling (a tool for conservation planning and resource management), owing to the broad use of geographical information systems in recent years, could be used for identifying suitable habitat and to predict potential distributions of vulture populations (Robertson et al., 2003, Rushton et al., 2004Vaz et al., 2008). ...
In seeking the answer to a question “By how much the vulture population has been decimated towards extinction?” the author went through different stages of surveying literature on counting, mapping and predicting the vulture habitats. This article pertains to avian conservation research with emphasis on vulture demography and ecology for a decade. The review investigates status of endangered vultures, their conservation need and use of methodology for collection of data with emphasis in Central India which is a vulture stronghold. Further it explores upon use of technology in analyzing the data and understanding the problems of vulture conservation. Different kind of maps generated and used for various purposes are discussed. Some management recommendations are also explored to revive the species from the brink.
... The limited availability of experimental data also remains a major constraint, and it is thus uncertain if mechanistic models can live up to their promise of providing more accurate forecasts of species' range shifts under climate change [62,89]. Indeed, a few studies have found mechanistic and correlative models to perform equally well [63,90]. Mechanistic model implementation also comes at the cost of increased data and computational requirements, limiting their wider use. ...
Predictive models are central to many scientific disciplines and vital for informing management in a rapidly changing world. However, limited understanding of the accuracy and precision of models transferred to novel conditions (their ‘transferability’) undermines confidence in their predictions. Here, 50 experts identified priority knowledge gaps which, if filled, will most improve model transfers. These are summarized into six technical and six fundamental challenges, which underlie the combined need to intensify research on the determinants of ecological predictability, including species traits and data quality, and develop best practices for transferring models. Of high importance is the identification of a widely applicable set of transferability metrics, with appropriate tools to quantify the sources and impacts of prediction uncertainty under novel conditions.
... Species-distribution models (SDMs) are numerical tools that combine observations of species occurrence or abundance with measurements of environmental parameters (Elith & Leathwick 2009). These models are used to interpret the role of environmental conditions in driving population distributions (Hacohen-Domené et al. 2015), and may be useful in predicting the likelihood of species occurrence in areas where biological knowledge is limited (Robertson et al. 2003). SDMs can be constructed using a variety of methods, ranging from relatively simple regression models to complex non-linear models (e.g. ...
Despite being a large, relatively abundant predator, the distribution and seasonal occurrence of the broadnose sevengill shark, Notorynchus cepedianus, in New Zealand is poorly understood. During 71 sampling trips conducted from July 2013 to May 2015, sharks were attracted to coastal sampling sites in southern New Zealand at Ōtākou/Otago Harbour and Te Whaka ā Te Wera/Paterson Inlet, Stewart Island, using chum. Using a logistic regression model, water temperature was identified as a key predictor of encountering sevengill sharks. In addition, location, cloud cover and sea state were also identified as influential predictors. At Ōtākou, a clear seasonal pattern of sevengill shark sightings emerged. Sharks were detected on 86% of survey trips in summer, whilst no sharks were detected in winter or spring. At Te Whaka ā Te Wera, sharks were sighted throughout all seasons, but a decline in shark encounters occurred during winter. This study represents the first systematic data on seasonal habitat use by sevengill sharks in New Zealand.
... Researchers and decision makers have used various types of SDMs along with systematic conservation planning (SCP) (Buisson, Thuiller, Casajus, Lek, & Grenouillet, 2010) software packages to assess or delimit conservation areas (Guisan et al., 2013). SDM outputs generally represent the probability of species presence (Liu, Berry, Dawson, & Pearson, 2005), and/or environmental suitability (Guisan et al., 2013;Robertson, Peter, Villet, & Ripley, 2003). The primary objective of SCP is to prioritize areas that maximize biodiversity. ...
Systematic conservation initiatives attempt to cater to the needs of many species via the integration of multiple species distribution models (SDMs), or via the integration of Systematic Conservation Planning (SCP) software, such as Zonation. Unfortunately, due to limited data and knowledge, it is often difficult to select the most suitable model for specific species, let alone an appropriate ensemble modeling method for multiple species. In general, model selection criteria are based on either model performance or consensus. The former integrates the highest-performing SDM for all focal species, whereas the latter integrates multiple SDM outputs based on consensus. While higher-performing ensemble models presumably identify high-quality habitats better, many have argued that high consensus ensemble models have less uncertainty originating from sporadic model variability. This study develops and validates seven ensemble-modeling strategies for integrating outputs of the systematic conservation tool Zonation. First, we considered the distributions of 11 bird species via 100 runs of five SDMs across Taiwan. Second, we evaluated the local and global uncertainty of all five models. Third, we used Zonation to obtain conservation priorities. We then used Principal Component Analysis (PCA) to quantify different sources of uncertainty. Finally, we used independent third-party habitat data to validate each strategy. On average, the ‘best model’ strategy (based on the highest AUC value) performed best. Based on our modeling exercise we present a comprehensive framework for conservation prioritization, validation and the quantification of uncertainty intrinsic to SDMs according to different conservation scenarios and goals.
... Besides Kappa, confusion matrices allowed calculation of accuracy and error measures such as global predictive success, sensitivity, specificity, and commission and omission errors (e.g., Forbes, 1995;Manel, Williams, & Ormerod, 2001;Robertson, Peters, Villet, & Ripley, 2003). The application of ROC complements potential weakness of the measures obtained from the error matrices (Fielding & Bell, 1997). ...
Considering the high biodiversity and conservation concerns of the tropical dry forest, this study aim is to predict and evaluate the potential and current distributions of twelve species of endemic birds which distribute along the western slope of Mexico. The main goal is to evaluate altogether different methods for predicting actual species distribution models (ADMs) of the twelve species including the identification of key environmental potential limiting factors. ADMs for twelve endemic Mexican birds were generated and validated by means of applying: (1) three widely used species niche modeling approaches (ENFA, Garp, and Maxent); (2) two thresholding methods, based on ROC curves and Kappa Index, for transforming continuous models to presence/absence (binary) models; (3) documented habitat–species associations for reducing species potential distribution models (PDMs); and (4) field occurrence data for validating final ADMs. Binary PDMs' predicted areas seemed overestimated, while ADMs looked drastically reduced and fragmented because of the approach taken for eliminating those predicted areas which were documented as unsuitable habitat types for individual species. Results indicated that both thresholding methods generated similar threshold values for species modeled by each of the three species distribution modeling algorithms (SDMAs). A Wilcoxon signed-rank test, however, showed that Kappa values were generally higher than ROC curve for species modeled by ENFA and Maxent, while for Garp models there were no significant differences. Prediction success (e.g., true presences percentage) obtained from field occurrence data revealed a range of 50%–82% among the 12 species. The three modeling approaches applied enabled to test the application of two thresholding methods for transforming continuous to binary (presence/absence) models. The use of documented habitat preferences resulted in drastic reductions and fragmentation of PDMs. However, ADMs predictive success rate, tested using field species occurrence data, varied between 50 and 82%.
... Most approaches have their roots in quantifying species-environment (or community-environment) relationships and extrapolating those relationships in space (e.g., through the use of GIS). Such models range from correlative or physiologically based models of species distributions (Anderson et al., 2002;Robertson et al., 2003) to simulations of species pattern and dispersal on hypothetical landscapes (Malanson and Cramer 1999;Nesslage et al., 2007) (see Chapters 22 and 23). Common topics that link the interests of biogeographers and landscape ecologists include the identification and protection of habitat for threatened and endangered species (Rayner et al., 2007), the spread of nonnative species (Schussman et al., 2006), and the effects of climate change on species ranges (Thuiller et al., 2008). ...
... Imbernon and Branthomme (2001) proposed a two-stage methodology using unsupervised classification followed by analysis of aerial photographs. Robertson, Peter, Villet, and Ripley (2003), Pedroni (2003), and Cayuela, Benayas, and Echeverría (2006) used algorithms, which allow the integration of spectral and statistical information in the classification process. Birger (2002) and Siegmann, Glässer, Itzerott, and Neumann (2014) developed a multi-sensor and multi-temporal approach for classification and monitoring of pioneer communities, while Steele, Bestelmeyer, Burkett, Smith, and Yannoff (2012) developed a multifactor classification system for the mapping of heterogeneous landscapes. ...
This paper aims to develop a flexible decision-tree framework for the classification of nine spectrally highly similar pioneer species for monitoring heterogeneous grassland habitats, which is based on different plant phenological indicators. For typical pioneer species, laboratory spectroscopic measurements were taken. Reflectance spectra were collected to cover a complete phenological cycle. First, a combined spectral similarity measure that consists of two independent methods was applied. Second, phenological metrics were derived from time series of NDVI for each species to describe differences in annual plant development. Further, spectral vegetation indices were applied that are related to plant physiological properties. The investigated species could be grouped into three spectral separability clusters. With increasing spectral similarities, species of the particular groups can be separated into subgroups and individual species that show a similar phenological development. The remaining subgroups with the highest spectral and phenological similarities could be divided to individual species by consideration of plant physiological parameters. The results show that the combination of different spectral methods with phenological metrics enhanced the classification of pioneer vegetation and other spectrally similar species. The approach can be used as basis for a continuous monitoring of fast changing habitats.
... Chart: Some groups of models associated with ruminant production systems, showing their scale of focus and modeling approach. Model groups are those discussed in this paper, addressing aspects felt to be most relevant in the context of climate change are based on detailed knowledge of host and pathogen physiology and attempt to replicate underlying mechanisms that drive species' responses to environmental variables ( Robertson et al., 2003). As such models do not rely on empirical relationships between climate variables that may alter with climate change, they are comparatively robust under spatio-temporal extrapolation ( Dormann, 2007;Hijmans and Graham, 2006) and can predict consequences of subtle interactions between system components under climate influence. ...
The livestock and grassland theme (LiveM) of the MACSUR (Modelling Agriculture with Climate Change for Food Security) (www.macsur.eu) knowledge hub brings together partners from across Europe to develop a pan-European modelling capability in the area of livestock systems modelling of climate change adaptation and mitigation. Through the project, inventories of grassland, animal and farm-scale models, as well as datasets related to grasslands and livestock have been compiled. Model inter-comparisons have taken place for grassland models, and a model evaluation protocol is being developed. Farm-scale modellers are undertaking a model inter-comparison exercise, and the theme has formed links to related projects in order to bring together a more coherent livestock systems modelling community. The need for better knowledge exchange within the livestock research community has been highlighted within the project, and is a focus for further action. The knowledge hub concept creates an arena for collaboration between research groups, disciplines and projects essential for tackling complex global issues such as the impact of climate change on agriculture.
... Direct comparisons of results generated from correlative and mechanistic approaches have produced divergent results. In some instances, little difference has been observed in their predictive power (Robertson et al., 2003), especially when species functional traits are included in correlative models (Kearney et al., 2010b). Other modelling efforts have shown large differences in predictions made by the two approaches , whereas different studies have suggested that the inclusion of ecological traits such as migratory ability and trophic level can increase the accuracy of species distribution models (SDMs) (McPherson and Jetz, 2007). ...
The emphasis in recent scientific studies has gradually shifted from merely documenting the numerous biological impacts of global climate change to developing predictive tools that help forecast which organisms, ecosystems and locations are most (and least) likely to be affected. This work often focuses on two very different scales of approach: the impacts of environmental change on individual organisms and their physiological vulnerability; and large-scale, biogeographic shifts in patterns of species distributions. While both scales of research are important and potentially informative of one another, with a few key exceptions, biogeographic (large scale) and physiological (small scale) approaches often operate in isolation from one another. There is a general consensus that a better understanding of mechanistic drivers will likely improve our ability to develop a more predictive framework. To this end, experimental and theoretical research has begun to tease apart the complexities of how changes in climate (as reflected in weather) ultimately translate into changes in growth, survival, productivity, species distribution and abundances, and the provision of ecosystem services. Yet, considerable debate still remains over how much detail is required to make effective predictions. Many authors have raised concerns about the implicit assumptions inherent in biological forecasts and while we cannot wait for perfect models before acting, oversimplifications can lead to unintended consequences. Additionally, what makes one forecasting method “good enough” or one approach “better” than another may depend largely on the application to which the predictions are being made. This review explores and summarizes the concerns raised by researchers working on problems at diverse scales and offers suggestions of how cross-scale research can best avoid potential pitfalls while preparing society for the ongoing impacts of climate change.
... Correlative models use distribution records of a species along with values of a set of predictor variables -such as resource gradients, or -to predict the likelihood of its occurrence, based on suitability of niche space. When these models use presence as well as absence records, they are considered a type of group discrimination techniques (Guisan et al. 2002;Robertson et al. 2003). We used generalized linear models with binomial error distribution to relate the Presence/Absence of L. camara to the independent variables. ...
Invasive plant species spread presents a challenge, requiring better mapping and monitoring for control. Remote sensing (RS) provides an efficient tool to map invasive plants in diverse ecosystems. Yet applications of RS for invasive plant mapping largely rely on spatial and spectral patterns. The use of invasive plant functional traits can improve RS mapping, using ecological insights on processes and functions associated with invasion. We summarize research utilizing plant functional traits in RS mapping of invasive species from the years 2000 to 2014. Based on this review, we summarize plant traits that can be related to spatial and spectral properties, and used to discriminate invasive alien plants from native vegetation. Phenological and structural plant traits have been relatively well exploited via RS for invasion studies. In comparison, there has been limited utilization of physiological traits (with the exception of properties such as nitrogen content). This is an area that merits further research attention, via the linkage of ecophysiological field research with RS.
... Chart: Some groups of models associated with ruminant production systems, showing their scale of focus and modeling approach. Model groups are those discussed in this paper, addressing aspects felt to be most relevant in the context of climate change are based on detailed knowledge of host and pathogen physiology and attempt to replicate underlying mechanisms that drive species' responses to environmental variables (Robertson et al., 2003). As such models do not rely on empirical relationships between climate variables that may alter with climate change, they are comparatively robust under spatio-temporal extrapolation (Dormann, 2007;Hijmans and Graham, 2006) and can predict consequences of subtle interactions between system components under climate influence. ...
Ruminant production systems are important producers of food, support rural communities and culture, and help to maintain a range of ecosystem services including the sequestering of carbon in grassland soils. However, these systems also contribute significantly to climate change through greenhouse gas (GHG) emissions, while intensification of production has driven biodiversity and nutrient loss, and soil degradation. Modeling can offer insights into the complexity underlying the relationships between climate change, management and policy choices, food production, and the maintenance of ecosystem services. This paper 1) provides an overview of how ruminant systems modeling supports the efforts of stakeholders and policymakers to predict, mitigate and adapt to climate change and 2) provides ideas for enhancing modeling to fulfil this role. Many grassland models can predict plant growth, yield and GHG emissions from mono-specific swards, but modeling multi-species swards, grassland quality and the impact of management changes requires further development. Current livestock models provide a good basis for predicting animal production; linking these with models of animal health and disease is a priority. Farm-scale modeling provides tools for policymakers to predict the emissions of GHG and other pollutants from livestock farms, and to support the management decisions of farmers from environmental and economic standpoints. Other models focus on how policy and associated management changes affect a range of economic and environmental variables at regional, national and European scales. Models at larger scales generally utilise more empirical approaches than those applied at animal, field and farm-scales and include assumptions which may not be valid under climate change conditions. It is therefore important to continue to develop more realistic representations of processes in regional and global models, using the understanding gained from finer-scale modeling. An iterative process of model development, in which lessons learnt from mechanistic models are applied to develop ‘smart’ empirical modeling, may overcome the trade-off between complexity and usability. Developing the modeling capacity to tackle the complex challenges related to climate change, is reliant on closer links between modelers and experimental researchers, and also requires knowledge-sharing and increasing technical compatibility across modeling disciplines. Stakeholder engagement throughout the process of model development and application is vital for the creation of relevant models, and important in reducing problems related to the interpretation of modeling outcomes. Enabling modeling to meet the demands of policymakers and other stakeholders under climate change will require collaboration within adequately-resourced, long-term inter-disciplinary research networks.
... The various steps of the development process of predictive modelling were described by several authors (Jørgensen, 1999;Manel et al., 1999;Guisan and Zimmermann, 2000;Jørgensen, 2000;Olden and Jackson, 2002;Robertson et al., 2003;Jørgensen, 2005). The features that are important for the models applied in this dissertation, will be discussed in the following paragraphs. ...
... assumption of spatial stationarity) has also been tested (Mellin et al. 2010). A major strength of these models is their ability to predict species occurrence patterns in areas where no data are available (Robertson et al. 2003), whether by (i) interpolating results and providing predictions for unsampled sites within the sampled area (Mellin et al. 2010), or by (ii) forecasting results under expected future environments (Ara ujo & New 2006;Sequeira et al. 2014a) or to new geographical areas (Sequeira et al. 2014a,b), as defined by Elith & Leathwick (2009). The application of such models to unsampled areas is termed transferability (Phillips 2008). ...
Understanding biodiversity patterns depends on data collection, which in marine environments can be prohibitively expensive. Transferable predictive models could therefore provide time‐ and cost‐effective tools for understanding biodiversity–environment relationships.
We used fish species counts and spatial and environmental predictors to develop predictive models of fish species richness ( S ) for two major coral reefs located in separate ocean basins: A ustralia's G reat B arrier R eef ( GBR ; Q ueensland) and N ingaloo R eef ( NR ; W estern A ustralia). We tested the ability of the GBR model to predict S at NR (its transferability) under various scenarios using different sampling durations, years sampled and transect sizes.
Based on R ² , the GBR model poorly predicted S at NR ( R ² < 16%) with few predicted values strongly correlated with observations. However, comparable spatial patterns in S across NR were predicted by both the NR and the GBR models when calibrated at similar spatio‐temporal scales.
This result suggests that poor validation of the transferred models may indicate low deviance explained by the predictors in the new system (where other predictors not included might have a more direct effect on the response) and that in some situations, model transferability may be considerably improved by using data sets of similar spatio‐temporal scales. Therefore, data filtering by time and space may be required prior to transferring models.
Policy implications . Transferable models can provide initial estimates of fish species richness patterns in poorly sampled systems, and thereby guide the design of better and more efficient sampling programs. Further improvements in model transferability will increase their predictive power and utility in conservation planning and management.
Spatiotemporal modelling has received increased attention from entomologists as it meets various demands, from understanding ecological patterns to predicting pest outbreaks. Recent challenges from modern agriculture require efficient analytical methods to analyze population dynamics and insect movement patterns. Also, the efficacy of control actions and outbreak predictions depends on well-trained professionals properly employing models to guarantee successful management programs. This chapter aims to provide the reader with an overview of the diversity of analytical tools available for application in various scenarios, accompanied by examples of applications in entomology taxonomically classified by orders of insects.
Brassica napus, a versatile crop with significant socioeconomic importance, serves as a valuable source of nutrition for humans and animals while also being utilized in biodiesel production. The expansion potential of B. napus is profoundly influenced by climatic variations, yet there remains a scarcity of studies investigating the correlation between climatic factors and its distribution. This research employs CLIMEX to identify the current and future ecological niches of B. napus under the RCP 8.5 emission scenario, utilizing the Access 1.0 and CNRM-CM5 models for the time frame of 2040–2059. Additionally, a sensitivity analysis of parameters was conducted to determine the primary climatic factors affecting B. napus distribution and model responsiveness. The simulated outcomes demonstrate a satisfactory alignment with the known current distribution of B. napus, with 98% of occurrence records classified as having medium to high climatic suitability. However, the species displays high sensitivity to thermal parameters, thereby suggesting that temperature increases could trigger shifts in suitable and unsuitable areas for B. napus, impacting regions such as Canada, China, Brazil, and the United States.
Biogeoscience is a rapidly growing interdisciplinary field that aims to bring together biological and geophysical processes. This book builds an enhanced understanding of ecosystems by focusing on the integrative connections between ecological processes and the geosphere, hydrosphere and atmosphere. Each chapter provides studies by researchers who have contributed to the biogeoscience synthesis, presenting the latest research on the relationships between ecological processes, such as conservation laws and heat and transport processes, and geophysical processes, such as hillslope, fluvial and aeolian geomorphology, and hydrology. Highlighting the value of biogeoscience as an approach to understand ecosystems, this is an ideal resource for researchers and students in both ecology and the physical sciences.
Natural control of invertebrate crop pests has the potential to complement or replace conventional insecticide-based practices, but its mainstream application is hampered by predictive unreliability across agroecosystems. Inconsistent responses of natural pest control to changes in landscape characteristics have been attributed to ecological complexity and system-specific conditions. Here, we review agroecological models and their potential to provide predictions of natural pest control across agricultural landscapes. Existing models have used a multitude of techniques to represent specific crop-pest-enemy systems at various spatiotemporal scales, but less wealthy regions of the world are underrepresented. A realistic representation of natural pest control across systems appears to be hindered by a practical trade-off between generality and realism. Nonetheless, observations of context-sensitive, trait-mediated responses of natural pest control to land-use gradients indicate the potential of ecological models that explicitly represent the underlying mechanisms. We conclude that modelling natural pest control across agroecosystems should exploit existing mechanistic techniques towards a framework of contextually bound generalizations. Observed similarities in causal relationships can inform the functional grouping of diverse agroecosystems worldwide and the development of the respective models based on general, but context-sensitive, ecological mechanisms. The combined use of qualitative and quantitative techniques should allow the flexible integration of empirical evidence and ecological theory for robust predictions of natural pest control across a wide range of agroecological contexts and levels of knowledge availability. We highlight challenges and promising directions towards developing such a general modelling framework.
With the advent of next-generation sequencing technologies, there is a reduction in sequencing costs, which on the other hand have increased the importance of cancer survival studies due to availability of high-throughput data and advancement in high-resolution cancer genome analysis with increased sensitivity. In this chapter, we have discussed selected statistical methods for survival analysis along with their strengths and limitations. We have considered global data corresponding to cancer statistics and patient survival rates. Further, we have emphasized on the biological and technical concepts of survival analysis. Here we have mainly focused on two popular female cancers, i.e., breast and ovarian carcinoma. We have employed Kaplan-Meier estimator and Cox proportional hazard regression to illustrate the case studies based on the featured design principles using cancer dataset. The performance of the Cox proportional hazard model surpasses that of the 2D statistical comparison model of Kaplan-Meier estimator as it can incorporate multiple layers of clinical information which enriches the prediction outcome.
This study assesses the potential impact of drought on arsenic exposure from private domestic wells by using a previously developed statistical model that predicts the probability of elevated arsenic concentrations (>10 μg per liter) in water from domestic wells located in the conterminous United States (CONUS). The application of the model to simulate drought conditions used systematically reduced precipitation and recharge values. The drought conditions resulted in higher probabilities of elevated arsenic throughout most of the CONUS. While the increase in the probability of elevated arsenic was generally less than 10% at any one location, when considered over the entire CONUS, the increase has considerable public health implications. The population exposed to elevated arsenic from domestic wells was estimated to increase from approximately 2.7 million to 4.1 million people during drought. The model was also run using total annual precipitation and groundwater recharge values from the year 2012 when drought existed over a large extent of the CONUS. This simulation provided a method for comparing the duration of drought to changes in the predicted probability of high arsenic in domestic wells. These results suggest that the probability of exposure to arsenic concentrations greater than 10 μg per liter increases with increasing duration of drought. These findings indicate that drought has a potentially adverse impact on the arsenic hazard from domestic wells throughout the CONUS.
For more than 70 years, Hutchinson's concept of the fundamental niche has guided ecological research. Hutchinson envisioned the niche as a multidimensional hypervolume relating the fitness of an organism to relevant environmental factors. Here, we challenge the utility of the concept to modern ecologists, based on its inability to account for environmental variation and phenotypic plasticity. We have ample evidence that the frequency, duration, and sequence of abiotic stress influence the survivorship and performance of organisms. Recent work shows that organisms also respond to the spatial configuration of abiotic conditions. Spatiotemporal variation of the environment interacts with the genotype to generate a unique phenotype at each life stage. These dynamics cannot be captured adequately by a multidimensional hypervolume. Therefore, we recommend that ecologists abandon the niche as a tool for predicting the persistence of species and embrace mechanistic models of population growth that incorporate spatiotemporal dynamics.
It has long been a concern that performance measures of species distribution models react to attributes of the modeled entity arising from the input data structure rather than to model performance. Thus, the study of Allouche et al. (Journal of Applied Ecology, 43, 1223, 2006) identifying the true skill statistics (TSS) as being independent of prevalence had a great impact. However, empirical experience questioned the validity of the statement. We searched for technical reasons behind these observations. We explored possible sources of prevalence dependence in TSS including sampling constraints and species characteristics, which influence the calculation of TSS. We also examined whether the widespread solution of using the maximum of TSS for comparison among species introduces a prevalence effect. We found that the design of Allouche et al. (Journal of Applied Ecology, 43, 1223, 2006) was flawed, but TSS is indeed independent of prevalence if model predictions are binary and under the strict set of assumptions methodological studies usually apply. However, if we take realistic sources of prevalence dependence, effects appear even in binary calculations. Furthermore, in the widespread approach of using maximum TSS for continuous predictions, the use of the maximum alone induces prevalence dependence for small, but realistic samples. Thus, prevalence differences need to be taken into account when model comparisons are carried out based on discrimination capacity. The sources we identified can serve as a checklist to safely control comparisons, so that true discrimination capacity is compared as opposed to artefacts arising from data structure, species characteristics, or the calculation of the comparison measure (here TSS).
Mountain Anoa (Bovidae: Bubalus quarlesi. Ouwens,1910) endemic to Indonesian island of Sulawesi and Buton. This animal has been classified into endangered species by IUCN, CITES and fully protected under Indonesian Law. This research aim is to produce a model of Mountain Anoa habitat relationships and to identify habitat variables affecting Mountain Anoa presence at the certain level. Approximately 2300 hectare area (2300 observation grid of one hectare each), has been surveyed to determine Anoa habitat. Logistic regression GIS coupled model is use to predict the habitat suitability of Mountain Anoa through Lore Lindu National Park. Environmental variable used in model like distance slope, elevation, distance from river, distance from settlement and distance from forest edge are taken from GIS data layer and remote sensing data. As result, altitude, distance from settlement, distance from forest edge, and distance from river have a significant factor while slope have insignificant factor for suitable habitat of Mountain Anoa. The overall classification accuracy was resulted on 96.0 % while the habitat suitability model was tested using Hosmer and Lemeshow goodness of fit resulted 0.248 and Nagelkerke's R2 score 73.4 % as result of variable representation in this model. The simulation result shows that suitable habitat for Mountain Anoa located in core zone area, deep in undisturbed forest. This result is consistent with Mountain Anoa behaviors that tend to avoid human activity area. This result also indicating there are human activities inside national park that force Mountain Anoa to move deeply to undisturbed forest area.
Animal and plant populations often occupy a variety of local areas and may experience different local birth and death rates in different areas. When this occurs, reproductive surpluses from productive source habitats may maintain populations in sink habitats, where local reproductive succes fails to keep pace with local mortality. For animals with active habitat selection, an equilibrium with both source and sink habitats occupied can be both ecologically and evolutionarily stable. If the surplus population of the source is large and the per capit deficit in the sink is small, only a small fraction of the total population will occur in areas where local reproduction is sufficient to compensate for local mortality. In this sense, the realized niche may be larger than the fundamental niche. Consequently, the particular species assemblage occupying any local study site may consist of a mixture of source and sink populations and may be as much or more influenced by the type and proximity of other habitats as by the resources and other conditions at the site. -Author
We propose a multivariate approach to the study of geographic species dis- tribution which does not require absence data. Building on Hutchinson's concept of the ecological niche, this factor analysis compares, in the multidimensional space of ecological variables, the distribution of the localities where the focal species was observed to a reference set describing the whole study area. The first factor extracted maximizes the marginality of the focal species, defined as the ecological distance between the species optimum and the mean habitat within the reference area. The other factors maximize the specialization of this focal species, defined as the ratio of the ecological variance in mean habitat to that observed for the focal species. Eigenvectors and eigenvalues are readily interpreted and can be used to build habitat-suitability maps. This approach is recommended in situations where absence data are not available (many data banks), unreliable (most cryptic or rare species), or meaningless (invaders). We provide an illustration and validation of the method for the alpine ibex, a species reintroduced in Switzerland which presumably has not yet recolonized its entire range.
The importance of the shape of a species response curve to an environmental gradient is reviewed. The implications for vegetation theory, ordination methods, species as indicators of environmental conditions, predicting species distribution from surveys and simulation models of climatic impact on vegetation are examined. A beta-function V = k (x - a)alpha . (b - x)gamma is used to model species response curves using generalized linear modelling (GLM). Two hypotheses are tested; (1) that response curves differ significantly from a unimodal symmetric (Gaussian) shape and (2) that the direction of skew is a function of species position along the gradient. Nine eucalypt species are modelled using GLM with a beta-function fitted for mean annual temperature. Six other environmental variables and factors are considered in fitting the statistical models; mean annual rainfall, mean monthly solar radiation, topographic position, lithological type, nutrient index and rainfall seasonality. All nine species are significantly skewed in response to temperature. The direction of skew is positive when a species optimum temperature falls below 11.5-degrees-C and negative for species with optimum above. These results indicate that current vegetation analyses which require the modelling of species response curves, or assumptions about the shape of the response, require reassessment.
Aims To show how logistic regression models for individual species can be used to produce improved estimates of species richness at a continental scale; to present these data for African ticks (Acari: Ixodida); and to address the question of whether there is a latitudinal gradient in tick species richness.
Location Africa.
Methods A database of 34,060 collection records for African ticks is used to produce a pan-African map of known tick species richness at 0.25 × 0.25-degree resolution. The likely distributions of seventy-three species are then estimated from environmental factors using logistic regression, and localities where there is a suitably high probability of occurrence for a given species are added to the original data for that species. These augmented data are combined to produce a map of the predicted pan-African distribution of tick species richness. The relationship of species richness to latitude is considered along a transect placed across some of the more extensively collected areas.
Results Maps of known and predicted pan-African tick species richness are presented, and deficiencies in the available data are highlighted. Correlations using both known and predicted estimates of tick species richness suggest that ticks follow similar species richness patterns to those described for African mammals and birds, with a latitudinal gradient and highest species richness in east equatorial Africa. Tick species ranges are log-normally distributed.
Main conclusions Carefully constructed probability surfaces offer a more powerful approach to mapping species ranges than simple presence-absence maps. Such models are a useful extension to current biogeographical methods and have a wide range of potential applications in ecology, epidemiology and conservation. Tick species richness at a continental scale follows similar trends to those reported for mammals and birds.
A method is described, using Generalized Linear Modelling, for determining the qualitative environmental realized niche (QERN) of plant species. Five species of Eucalyptus growing in south-eastern Australia are used as examples. The frequency of each species occurrence in a data set comprising 6080 observations over an area of ≈40000 km2 is modelled as a function of four environmental variables: mean annual rainfall, mean annual temperature, a solar radiation index, and rock type. The position and shape of the response of a species to these four variables is the species' QERN. The effect of rock type on the distribution of each species was found to be highly significant. To reduce model complexity, the data were stratified on rock type, and models for each species were generated for the two common rock types, granites and soft sediments. Whilst the model-fitting procedure generally is the same, each species' response demanded some modifications of the procedure. The continuous variates, mean annual temperature, mean annual rainfall, and solar radiation index, were categorized so that the magnitude of the model coefficients from each category would indicate the shape of each species' response. Continuous functions were then fitted to the categorical response shapes. Of the 10 models (5 species and 2 rock types) 5 revealed a skewed response to one or more of the environmental variables, one was linear, and two were bell shaped. Two more models revealed complex response shapes, suggesting that the environmental variables were inappropriate or insufficient to model those species' realized niches. The results provide evidence for asymmetric responses of species to environmental variables contrary to the bell-shaped symmetrical responses commonly assumed in ecological theory. Models of species' realized niches are necessary in community ecology to describe the role of environment, prior to studies on the role of competition in determining community composition. Limitations of the models, both statistical and ecological, are considered, with suggestions for improvements.
Predictive vegetation mapping can be defined as pedicting the geographic distribution of the vegetation composition across a landscape from mapped environmental variables. Computerized predictive vegetation mapping is made possibly by the availability of digital maps of topography and other environmental variables such as soils, geology and climate variables, and geographic information system software for manipulating these data. Especially important to predictive vegetation mapping are interpolated climatic variables related to physiological tolerances, and topographic variables, derived from digital elevation grids, related to site energy and moisture balance. Predictive vegetation mapping is founded in ecological niches theory and gradient analysis, and driven by the need to map vegetation patterns over large areas for resource conservation planning, and to predict the effects of environmental change on vegetation distributions. -from Author
SUMMARY After an introduction outlining the differences between the biological and statistical approaches to understanding the distribution and abundance of organisms, this chapter gives two examples of dimension-reducing statistical techniques whereby large amounts of environmental data can be processed and sifted to extract useful correlates of the distributional ranges of animal species. These techniques are illustrated using the tsetse fly, Glossina morsitans, as an example. The first technique is that of linear discriminant analysis which predicts the past and present distribution of G. morsitans in Zimbabwe, Kenya and Tanzania with an accuracy of >80%. Conclusions from the statistical analysis coincide with previous biological interpretations of the dis- tribution of this species in Africa. The message to emerge from the analysis is that Global Circulation Models (GCMs) will need to achieve a greater degree of accuracy than at present if they are to be useful in making predictions about changing vector distributions with global climate change. The second technique is temporal Fourier analysis of a series of Normalized Difference Vegetation Indices (NDVIs) of Africa derived frOItl the Advanced Very High Resolution Radiometers (A VHRR) of Earth-orbiting meteorological satellites of the National Oceanic and Atmospheric Administration (NOAA) series. The analysis captures the important characteristics (i.e. the average, amplitude and phase) of the major annual and biannual cycles of vegetation growth. Examples are given of features (the Gezira irrigation project in Sudan) and processes (the timing of the peak vegetation growth along the Nile from Uganda to the Mediterranean) which are revealed clearly by Fourier analysis. A strong association is demonstrated between the amplitude of the first term of the Fourier expansion (= the amplitude of the annual cycle of vegetation growth) and savannah woodland areas of Africa, and a similar
A model to predict global patterns in vegetation physiognomy was developed from physiological considerations influencing the distributions of different functional types of plant. Primary driving variables are mean coldest-month temperature, annual accumulated temperature over 5-degrees-C, and a drought index incorporating the seasonality of precipitation and the available water capacity of the soil. The model predicts which plant types can occur in a given environment, and selects the potentially dominant types from among them. Biomes arise as combinations of dominant types. Global environmental data were supplied as monthly means of temperature, precipitation and sunshine (interpolated to a global 0.5-degrees grid, with a lapse-rate correction) and soil texture class. The resulting predictions of global vegetation patterns were in good agreement with the mapped distribution of actual ecosystem complexes (Olson, J.S., Watts, J.A. & Allison, L.J. (1983) ORNL-5862, Oak Ridge Nat. Lab., 164 pp.), except where intensive agriculture has obliterated the natural patterns. The model will help in assessing impacts of future climate changes on potential natural vegetation patterns, land-surface characteristics and terrestrial carbon storage, and in analysis of the effects of past climate change on these variables.
New Zealand's four broad-leaved evergreen tree species from the genus Nothofagus all show pronounced distributional disjunctions, independent of environmental factors known to influence tree distributions. Here, we use these disjunctions as the basis for a natural removal experiment to investigate competitive interactions between Nothofagus and a range of other widespread conifer and broad-leaved tree species. We first model the abundance of non-Nothofagus species as a function of environment, using Generalized Additive Models (GAMs) and an extensive data set sampling much of New Zealand's remaining old-growth forests. We then assess the effects of competitive interaction with Nothofagus by adding statistical terms describing (1) Nothofagus abundance, and (2) interactions between Nothofagus abundance and annual temperature, the dominant environmental gradient. Results indicate substantial reductions in the abundance of many species as Nothofagus abundance increases. The magnitude of this reduction varies with position along the dominant environmental gradient; species overlapping most strongly with Nothofagus are generally most sensitive to increases in Nothofagus abundance. In addition, both the shapes of species responses to mean annual temperature and the positions of their optima change as Nothofagus abundance increases. This demonstration of competition using community compositional data has implications both for vegetation theory and for prediction of the likely impacts of global warming on New Zealand's forest pattern.
Animal and plant populations often occupy a variety of local areas and may experience different local birth and death rates in different areas. When this occurs, reproductive surpluses from productive source habitats may maintain populations in sink habitats, where local reproductive success fails to keep pace with local mortality. For animals with active habitat selection, an equilibrium with both source and sink habitats occupied can be both ecologically and evolutionarily stable. If the surplus population of the source is large and the per capita deficit in the sink is small, only a small fraction of the total population will occur in areas where local reproduction is sufficient to compensate for local mortality. In this sense, the realized niche may be larger than the fundamental niche. Consequently, the particular species assemblage occupying any local study site may consist of a mixture of source and sink populations and may be as much or more influenced by the type and proximity of other habitats a...
Aim To offer an objective approach to some of the problems associated with the development of logistic regression models: how to compare different models, determination of sample size adequacy, the influence of the ratio of positive to negative cells on model accuracy, and the appropriate scale at which the hypothesis of a non-random distribution should be tested.
Location Test data were taken from Southern Africa.
Methods The approach relies mainly on the use of the AUC (Area under the Curve) statistic, based on ROC (threshold Receiver Operating Characteristic) plots, for between-model comparisons. Data for the distribution of the bont tick Amblyomma hebraeum Koch (Acari: Ixodidae) are used to illustrate the methods.
Results Methods for the estimation of minimum sample sizes and more accurate hypothesis-testing are outlined. Logistic regression is robust to the assumption that uncollected cells can be scored as negative, provided that the sample size of cells scored as positive is adequate. The variation in temperature and rainfall at localities where A. hebraeum has been collected is significantly lower than expected from a random sample of points across the data set, suggesting that within-site variation may be an important determinant of its distribution.
Main conclusions Between-model comparisons relying on AUCs can be used to enhance objectivity in the development and refinement of logistic regression models. Both between-site and within-site variability should be considered as potentially important factors determining species distributions.
Neglect of ecological knowledge is a limiting factor in the use of statistical modelling to predict species distribution. Three components are needed for statistical modelling, an ecological model concerning the ecological theory to be used or tested, a data model concerning the collection and measurement of the data, and a statistical model concerning the statistical theory and methods used. This component framework is reviewed with emphasis on ecological theory. The expected shape of a species response curve to an environmental gradient is a central assumption on which agreement has yet to be reached. The nature of the environmental predictors whether indirect variables, e.g. latitude that have no physiological impact on plants, or direct variables, e.g. temperature also influence the type of response expected. Straight-line relationships between organisms and environment are often used uncritically. Many users of canonical correlation analysis use linear (straight-line) functions to relate ordination axes to variables such as slope and aspect though this is not a necessary part of the method. Some statisticians have used straight lines for species/environment relationships without testing, when evaluating new statistical procedures. Assumptions used in one component often conflict with those in another component. Statistical models can be used to explore ecological theory. Skewed species response curves predominate contrary to the symmetric unimodal curves assumed by some statistical methods. Improvements in statistical modelling can be achieved based on ecological concepts. Examples include incorporating interspecific competition from dominant species; more proximal predictors based on water balance models and spatial autocorrelation procedures to accommodate non-equilibrium vegetation. # 2002 Elsevier Science B.V. All rights reserved.
1. An extensive data set describing the composition of New Zealand’s remaining indigenous forests was used to estimate the degree of correlation between measures of both soil and atmospheric water deficit and the distribution of common tree species.
2. For most species, regression models incorporating measures of air saturation deficit in early autumn, as well as an annual integral of root zone water deficit, provided the best explanation of spatial distribution. This accords strongly with the mechanistic effects of air saturation deficits on transpiration from trees, and the hydraulic risks experienced by trees under high evaporative demand.
3. Adjustment of root zone water deficits to account for reductions in rainfall in dry years substantially improved model predictions. This suggests that extreme climatic events, such as the El Niño phase of the Southern Oscillation, are likely to have strongly influenced the historic composition of forests in New Zealand’s drier eastern lowlands.
A reformulation of the continuum concept is presented after considering the implications of the community/continuum controversy and current niche theory. Community is a spatial concept dependent on landscape pattern while the continuum is an environmental concept referring to an abstract space. When applying niche theory to plants, the mechanisms of competition are ill-defined and the assumption of bell-shaped response curves for species unrealistic.
Eight testable propositions on the pattern of response of vegetation to environmental gradients are presented 1. Environmental gradients are of two types. a) resource gradients or b) direct physiological gradients. 2. The fundamental niche response of species to resource gradients is a series of similar nested response curves. 3. The fundamental niche response of species to direct gradients is a series of separate, independent, overlapping response curves. 4. Species fundamental response curves are such that they have a relative performance advantage in some part of the environmental space. 5. The shape of the realized niche is variable even bimodal but predictable from the fundamental response given the other species present. Propositions 6–8 describe the response shapes of emergent community properties to environmental gradient; species richness is bimodal, dominance trimodal and standing crop unimodal. Detailed comparisons of these propositions are made with the alternative theories of Ellenberg, Gauch and Whittaker, Grime, and Tilman. These theories are incomplete lacking several generally accepted properties of plants and vegetation.
1. Empirical models for spatial distribution of wildlife, given data from a complete census or a random sample of sites, are reviewed briefly. 2. The use of covariates, recorded at different resolutions, for modelling spatial distribution is explored. 3. Presentation of model predictions in map form is discussed. 4. A framework of models for change in spatial distribution, given data from successive surveys, is developed. 5. Methods for quantifying and presenting precision and bias are described. 6. The methods are illustrated for two bird species (green woodpecker Picus viridis and redstart Phoenicurus phoenicurus) and for red deer Cervus elaphus, using data from north-east Scotland.
The paper describes how the areal distributions of three kangaroo species (Macropus spp.) in Australia can be re-expressed in terms of their climatic characteristics using the modelling capabilities of SIMPLE. SIMPLE is a computer system which integrates modelling techniques with the data base and mapping functions of a geographic information system. Two modelling techniques suitable for modelling the presence and absence of entities, namely classification trees and logistic regression are explored.
A Mapped Atmosphere-Plant-Soil System (MAPSS) has been constructed for simulating the potential biosphere impacts and biosphere-atmosphere feedbacks from climatic change. The system calculates the potential vegetation type and leaf area that could be supported at a site, within the constraints of the abiotic climate. Both woody vegetation and grass are supported and compete for light and water. The woody vegetation can be either trees or shrubs, evergreen or deciduous, and needleleaved or broadleaved. A complete site water balance is calculated and integrates the vegetation leaf area and stomatal conductance in canopy transpiration and soil hydrology. The MAPSS model accurately simulates the distributions of forests, grasslands, and deserts and reproduces observed monthly runoff. The model can be used for predictions of new vegetation distribution patterns, soil moisture, and runoff patterns in alternative climates.
Partial table of contents: Discriminant Analysis in Research. PREDICTION. Basic Ideas of Classification. Multivariate Normal Rules. Classification Results. Hit Rate Estimation. Nonnormal Rules. Reporting Results of a PDA. Applications of PDA. DESCRIPTION. Group Separation. Assessing Effects. Describing Effects. Selecting and Ordering Variables. Reporting Results of a DDA. Applications of DDA. ISSUES AND PROBLEMS. Special Problems. Appendices. Answers to Exercises. References. Index.
. Past explanations of the large disjunctions in the distribution of New Zealand's four Nothofagus species have emphasized displacement during glacial cycles followed by slow re-occupation of suitable sites, or the effects of plate tectonics coupled with ecological and/or environmental limitations to further spread. In this study the degree of equilibrium between Nothofagus distribution and environment was compared with that of other widespread tree species by statistical analysis. Generalized additive regression models were used to relate species distribution data to estimates of temperature, solar radiation, soil water deficit, atmospheric humidity, lithology and drainage. For each species, the amount of spatial patterning remaining unexplained by environment was assessed by adding a variable describing species presence/absence on adjacent plots. Results indicate that Nothofagus species occur more frequently in environments suboptimal for tree growth, i.e. having various combinations of cool temperatures, low winter solar radiation, high root-zone water deficit, low humidity, and infertile granitic substrates. Despite these demonstrated preferences, they exhibit substantially more spatial clustering which is unexplained by environment, than most other widespread tree species. Predictions formed from regressions using environment alone confirm that several major Nothofagus disjunctions are not explicable in terms of the environmental factors used in this analysis, but more likely reflect the effects of historic displacement coupled with slowness to invade forest dominated by more rapidly dispersing endomycorrhizal species. The technique used in this study for detecting residual spatial autocorrelation after fitting explanatory variables has potentially wide application in other studies where either regression or ordination techniques are used for analysis of compositional data.
With the rise of new powerful statistical techniques and GIS tools, the development of predictive habitat distribution models has rapidly increased in ecology. Such models are static and probabilistic in nature, since they statistically relate the geographical distribution of species or communities to their present environment. A wide array of models has been developed to cover aspects as diverse as biogeography, conservation biology, climate change research, and habitat or species management. In this paper, we present a review of predictive habitat distribution modeling. The variety of statistical techniques used is growing. Ordinary multiple regression and its generalized form (GLM) are very popular and are often used for modeling species distributions. Other methods include neural networks, ordination and classification methods, Bayesian models, locally weighted approaches (e.g. GAM), environmental envelopes or even combinations of these models. The selection of an appropriate method should not depend solely on statistical considerations. Some models are better suited to reflect theoretical findings on the shape and nature of the species’ response (or realized niche). Conceptual considerations include e.g. the trade-off between optimizing accuracy versus optimizing generality. In the field of static distribution modeling, the latter is mostly related to selecting appropriate predictor variables and to designing an appropriate procedure for model selection. New methods, including threshold-independent measures (e.g. receiver operating characteristic (ROC)-plots) and resampling techniques (e.g. bootstrap, cross-validation) have been introduced in ecology for testing the accuracy of predictive models. The choice of an evaluation measure should be driven primarily by the goals of the study. This may possibly lead to the attribution of different weights to the various types of prediction errors (e.g. omission, commission or confusion). Testing the model in a wider range of situations (in space and time) will permit one to define the range of applications for which the model predictions are suitable. In turn, the qualification of the model depends primarily on the goals of the study that define the qualification criteria and on the usability of the model, rather than on statistics alone.
The application of a model modified from Jeffree and Jeffree (1994) for investigating the distribution responses of selected antlion species to a climate change scenario was explored in this study. Modifications include a multivariate capability that facilitates the incorporation of precipitation seasonality, and provides useful output in the form of probability of occurrence values for each species. The model can be used to interpolate the distributions of poorly sampled taxa as well as predict responses to a changing climate. It is predicted that species from the more arid western parts of South Africa will be subject to severe range contraction and range shifts whereas the species from the more mesic eastern parts will experience range contraction with limited range shift. The likelihood of successful range shifts will be affected by the nature of novel communities, habitat suitability and the degree of land transformation. Given the extent of the predicted spatial responses, conservation planners can no longer afford to ignore future climate impacts on species distribution patterns.
A model to predict the effects of global warming on the composition of New Zealand's native forests is described. Relationships between the current distribution of 41 tree species and site temperature, solar radiation, water balance, lithology, and drainage for 14 500 plots have been analysed using non-parametric regression. Distributions of species were then predicted for points on a 5 km grid across New Zealand. A test for residual spatial autocorrelation using a ‘proximity’ variable, indicated that New Zealand's four Nothofagus species have distributions less well correlated with environmental variables than most other species. Inclusion of the ‘proximity’ variable in the regressions also substantially improved predictions of Nothofagus distribution. Predictions for other species were improved by incorporating a term representing interaction with the patchily distributed but strongly dominant Nothofagus species. Preliminary results from a cluster analysis of the combined predictions for all species indicate that the model successfully reconstructs the existing pattern of New Zealand's indigenous forests. Estimation of the effects of global warming on species distribution was done by introducing a perturbation to represent an overall increase in temperature of 2°C. The results indicate that a substantial disequilibrium is likely to occur between the current forest pattern and expected warmer temperatures.
. The relationship between present climate and the distribution in Europe of the aggressively invasive exotic Fallopia japonica is described by fitting a response surface based on three bioclimatic variables: mean temperature of the coldest month, the annual temperature sum > 5 °C, and the ratio of actual to potential evapotranspiration. The close fit between the observed and simulated distributions suggests that the species' European distribution is climatically determined. The response surface also provides a simulation of the extent of the area of native distribution of F. japonica in Southeast Asia that is generally accurate, confirming the robustness of the static correlative model upon which it is based. Simulations of the potential distribution of F. japonica under two alternative 2 x CO2 climate change scenarios indicate the likelihood of considerable spread into higher latitudes and possible eventual exclusion of the species from central Europe. However, despite the robustness of the response surface with present-day climate, the reliability of these simulations as forecasts is likely to be limited because no account is taken of the direct effects of CO2 and their interaction with the species' physiological responses to climate. Similarly, no account is taken of the potential impact of interactions with ‘new’ species as ecosystems change in composition in response to climate change. Nevertheless, the simulations indicate both the possible magnitude of the impacts of forecast climate changes and the regions that may be susceptible to invasion by F. japonica.
Correlative approaches to understanding the climatic controls of vegetation distribution have exhibited at least two important weaknesses: they have been conceptually divorced across spatial scales, and their climatic parameters have not necessarily represented aspects of climate of broad physiological importance to plants. Using examples from the literature and from the Sierra Nevada of California, I argue that two water balance parameters—actual evapotranspiration (AET) and deficit (D)—are biologically meaningful, are well correlated with the distribution of vegetation types, and exhibit these qualities over several orders of magnitude of spatial scale (continental to local). I reach four additional conclusions. (1) Some pairs of climatic parameters presently in use are functionally similar to AET and D; however, AET and D may be easier to interpret biologically. (2) Several well-known climatic parameters are biologically less meaningful or less important than AET and D, and consequently are poorer correlates of the distribution of vegetation types. Of particular interest, AET is a much better correlate of the distributions of coniferous and deciduous forests than minimum temperature. (3) The effects of evaporative demand and water availability on a site's water balance are intrinsically different. For example, the ‘dry’ experienced by plants on sunward slopes (high evaporative demand) is not comparable to the ‘dry’ experienced by plants on soils with low water-holding capacities (low water availability), and these differences are reflected in vegetation patterns. (4) Many traditional topographic moisture scalars—those that additively combine measures related to evaporative demand and water availability—are not necessarily meaningful for describing site conditions as sensed by plants; the same holds for measured soil moisture. However, using AET and D in place of moisture scalars and measured soil moisture can solve these problems.
We present a correlative modelling technique that uses locality records (associated with species presence) and a set of predictor variables to produce a statistically justifiable probability response surface for a target species. The probability response surface indicates the suitability of each grid cell in a map for the target species in terms of the suite of predictor variables. The technique constructs a hyperspace for the target species using principal component axes derived from a principal components analysis performed on a training dataset. The training dataset comprises the values of the predictor variables associated with the localities where the species has been recorded as present. The origin of this hyperspace is taken to characterize the centre of the niche of the organism. All the localities (grid-cells) in the map region are then fitted into this hyperspace using the values of the predictor variables at these localities (the prediction dataset). The Euclidean distance from any locality to the origin of the hyperspace gives a measure of the ‘centrality’ of that locality in the hyperspace. These distances are used to derive probability values for each grid cell in the map region. The modelling technique was applied to bioclimatic data to predict bioclimatic suitability for three alien invasive plant species (Lantana camara L., Ricinus communis L. and Solanum mauritianum Scop.) in South Africa, Lesotho and Swaziland. The models were tested against independent test records by calculating area under the curve (AUC) values of receiver operator characteristic (ROC) curves and kappa statistics. There was good agreement between the models and the independent test records. The pre-processing of climatic variable data to reduce the deleterious effects of multicollinearity, and the use of stopping rules to prevent overfitting of the models are important aspects of the modelling process.
BIOCLIM is a bioclimate analysis and prediction system which can be used to stratify an area on a climatic basis prior to survey and also to predict distributions of individual entities such as species or vegetation types. BIOCLIM is based on continuous mathematical surfaces fitted to measured meteorological data, and can be used to generate estimates of monthly mean minimum and maximum temperatures and precipitation for any point on or near mainland Australia and Tasmania, from inputs of latitude, longitude and elevation. -from Author
Scaevola plumieri is an important pioneer on many tropical and subtropical sand dunes, forming a large perennial subterranean plant with only the tips of the branches emerging above accreting sand. In South Africa it is the dominant pioneer on sandy beaches along the east coast, less abundant on the south coast and absent from the southwest and west coasts. Transpiration rates (E) of S. plumieri are predictably related to atmospheric vapour pressure deficit under a wide range of conditions and can therefore be predicted from measurement of ambient temperature and relative humidity. Scaling measurements of E at the leaf level to the canopy level has been demonstrated previously. Using a geographic information system, digital maps of regional climatic variables were used to calculate digital maps of potential transpiration from mean monthly temperature and relative humidity values, effectively scaling canopy level transpiration rates to a regional level. Monthly potential transpiration was subtracted from the monthly median rainfall to produce a map of mean monthly water balance. Seasonal growth was correlated with seasonal water balance. Localities along the coast with water deficits in summer corresponded with the recorded absence of S. plumieri, which grows and reproduces most actively in the summer months. This suggests that reduced water availability during the summer growth period limits the distribution of S. plumieri along the southwest coast, where water deficits develop in summer. Temperature is also important in limiting the distribution of S. plumieri on the southwest coast of South Africa through its effects on the growth and phenology of the plant. Nomenclature: Dyer (1967). Abbreviations: E = Transpiration; ISSR = Inter Simple Sequence Repeat; LAI = Leaf area index; RH = Relative humidity; SVP = Saturation vapour pressure; VPD = Vapour pressure deficit.
A quantitative study of relationships between forest pattern and environment in the central North Island, New Zealand, is based on forest composition data from ca. 2000 existing plots distributed throughout the forests of the region. Estimates of mean annual temperature, rainfall, and solar radiation are derived for each plot from mathematical surfaces fitted to climate station data. Estimates of the depth of the last major rhyolitic eruption, (Taupo Pumice, ca. 130 AD) are derived from isopach maps. A classification procedure is used to identify broad compositional groups. Generalised linear models are used to examine relationships between major species and climatic and other physical factors. Significant relationships are identified between the distributions of both plot groups and species, and climate, vulcanism, topography and drainage. Among these factors, temperature and/or solar radiation are indicated as major determinants of the regional forest pattern, with rainfall, topography, and drainage acting at a secondary level. The role of the Taupo Pumice eruption is more difficult to interpret, and its effects seem to have been greatly influenced by topography. Deep extensive deposits of tephra on flat-to-rolling sites close to the eruption centre have probably favoured the current dominance of these sites by more rapidly dispersing conifers. In contrast, on adjacent steep sites where forest destruction was likely to be less severe, slow-dispersing Nothofagus species are largely dominant. Further work is needed to understand the factors favouring conifer dominance of the central basins and the degree to which Nothofagus species might expand their range in the future.
1. Presence-absence maps of the occurrence of species may be difficult to interpret because of uneven coverage in field surveys. 2. Using data collected in the mountain Kingdom of Lesotho, southern Africa, where even coverage was virtually impossible to achieve, a method is presented to improve bird atlas maps by plotting predicted probabilities of occurrence calculated from logistic models. 3. A set of habitat variables derived from maps was reduced to principal components and then used in logistic models to predict the occurrence of bird species across the country. The models generate a probability of occurrence for all survey squares irrespective of coverage. Maps for the Cape vulture Gyps coprotheres (Forster), ground woodpecker Geocolaptes olivaceus (Gmelin) and pied crow Corvus albus Muller are presented as examples. 4. The models were assessed by jack-knife analysis and correctly predicted presence and absence in the field data. 5. Relationships between the predicted probabilities, recording frequency (reporting rate) and density are discussed.
Approaches to determining the number of components to interpret from principal components analysis were compared. Heuristic procedures included: retaining components with eigenvalues (Xs) > 1 (i.e., Kaiser-Guttman criterion); components with bootstrapped Xs > 1 (bootstrapped Kaiser-Guttman); the scree plot; the broken-stick model; and components with Xs totalling to a fixed amount of the total variance. Statistical ap- proaches included: Bartlett's test of sphericity; Bartlett's test of homogeneity of the cor- relation matrix, Lawley's test of the second X; bootstrapped confidence limits on successive Xs (i.e., significant differences between Xs); and bootstrapped confidence limits on eigen- vector coefficients (i.e., coefficients that differ significantly from zero). All methods were compared using simulated data matrices of uniform correlation structure, patterned ma- trices of varying correlation structure and data sets of lake morphometry, water chemistry, and benthic invertebrate abundance. The most consistent results were obtained from the broken-stick model and a combined measure using bootstrapped Xs and associated eigen- vector coefficients. The traditional and bootstrapped Kaiser-Guttman approaches over- estimated the number of nontrivial dimensions as did the fixed-amount-of-variance model. The scree plot consistently estimated one dimension more than the number of simulated dimensions. Bartlett's test of sphericity showed inconsistent results. Both Bartlett's test of homogeneity of the correlation matrix and Lawley's test are limited to testing for only one and two dimensions, respectively.
. Using the results of a total floristic survey of two veld types (Arid Sweet Bushveld and Mixed Bushveld) in the northeastern Transvaal, South Africa, we linked median annual rainfall from a surface response model to each of 139 samples. The samples had been classified floristically into 15 plant communities. These communities represent two broad divisions, corresponding with the concepts embodied in the two veld types. Using contingency tables, we defined the conditions of median annual rainfall and elevation for each of the veld types. Using a geographic analysis system we predicted the distribution of the veld types in an area of 120 000 km2 outside the study area. The predicted distribution was validated by comparison with a digitized version of the Acocks map. We conclude that the defined conditions of median annual rainfall and elevation provide confident criteria for the definition of these veld types.
A dataset of some 10 000 plots was used to describe the climatic relationships of 33 widespread New Zealand tree species. Estimates of mean annual temperature, temperature seasonality, mean annual solar radiation, and moisture balance were derived from mathematical surfaces fitted to climate station data. Plots were also categorized into five lithological classes and three drainage classes. Generalized additive models were used to examine species/environment relationships. Mean annual temperature and mean annual solar radiation are most strongly correlated with current tree distributions, followed by moisture balance, temperature seasonality, lithology, and drainage. Most broad-leaved tree species other than Nothofagus spp. reach their greatest levels of occurrence in warm, moist environments with high solar radiation. In contrast, Nothofagus spp. generally reach their greatest levels of occurrence in cooler and/or lower insolation environments, and all have lower levels of occurrence on rhyolitic substrates which have resulted from large-scale geomorphic disturbance, mostly over the past few thousand years. Although coniferous species have widely differing climatic optima, many are biased towards lithological classes characterized either by large-scale geomorphic disturbance or harsh edaphic conditions. The relevance of these results to particular synecological questions is briefly discussed. Continuing adjustments in the range of slow-dispersing Nothofagus spp. are strongly suggested, and the climatic suitability of extensive rhyolitic basins in the central North Island, from which these species are largely absent, is confirmed.
The relationships between the distribution of alpine species and selected environmental variables are investigated by using two types of generalized linear models (GLMs) in a limited study area in the Valais region (Switzerland). The empirical relationships are used in a predictive sense to mimic the potential abundances of alpine species over a regular grid. Here, we present the results for the alpine sedge Carex curvula ssp. curvula . The modelling approach consists of (1) a binomial GLM, including only the mean annual temperature as explanatory variable, which is adjusted to species presence/absence data in the entire study area; (2) a logistic model restricted to stands occurring within the a priori defined temperature range for the species ‐ which allows ordinal abundance data to be adjusted; (3) the two species‐response functions combined in a GIS to generate a map of the species' potential abundance in the study area; (4) model predictions filtered by the classes of the qualitative variables under which the species never occur.
Such a stratified approach used to better fit the variability within the optimal altitudinal zone for the species. Removing stand descriptions from altitudes too high or too low, where the species is unlikely to occur, enhances the global modelling performance by allowing the identification of important environmental variables only retained in the second model.
The model evaluation is finally carried out with the γ‐measure of association in an ordinal contingency table. It shows that abundance is satisfactorily predicted for C. curvula .
Monitoring of regional vegetation and surface biophysical properties is tightly constrained by both the quantity and quality of ground data. Stratified sampling is often used to increase sampling efficiency, but its effectiveness hinges on appropriate classification of the land surface. A good classification must be sufficiently detailed to include the important sources of spatial variability, but at the same time it should be as parsimonious as possible to conserve scarce and expensive degrees of freedom in ground data. As part of the First ISLSCP (International Satellite Land Surface Climatology Program) Field Experiment (FIFE), we used Regression Tree Analysis to derive an ecological classification of a tall grass prairie landscape. The classification is derived from digital terrain, land use, and land cover data and is based on their association with spectral vegetation indices calculated from single-date and multi-temporal satellite imagery. The regression tree analysis produced a site stratification that is similar to the a priori scheme actually used in FIFE, but is simpler and considerably more effective in reducing sample variance in surface measurements of variables such as biomass, soil moisture and Bowen Ratio. More generally, regression tree analysis is a useful technique for identifying and estimating complex hierarchical relationships in multivariate data sets.
Invasive alien organisms pose a major threat to global biodiversity. The Cape Peninsula, South Africa, provides a case study of the threat of alien plants to native plant diversity. We sought to identify where alien plants would invade the landscape and what their threat to plant diversity could be. This information is needed to develop a strategy for managing these invasions at the landscape scale. We used logistic regression models to predict the potential distribution of six important invasive alien plants in relation to several environmental variables. The logistic regression models showed that alien plants could cover over 89% of the Cape Peninsula. Acacia cyclops and Pinus pinaster were predicted to cover the greatest area. These predictions were overlaid on the current distribution of native plant diversity for the Cape Peninsula in order to quantify the threat of alien plants to native plant diversity. We defined the threat to native plant diversity as the number of native plant species (divided into all species, rare and threatened species, and endemic species) whose entire range is covered by the predicted distribution of alien plant species. We used a null model, which assumed a random distribution of invaded sites, to assess whether area invaded is confounded with threat to native plant diversity. The null model showed that most alien species threaten more plant species than might be suggested by the area they are predicted to invade. For instance, the logistic regression model predicted that P. pinaster threatens 350 more native species, 29 more rare and threatened species, and 21 more endemic species than the null model would predict. Comparisons between the null and logistic regression models suggest that species richness and invasibility are positively correlated and that species richness is a poor indicator of invasive resistance in the study site. Our results emphasize the importance of adopting a spatially explicit approach to quantifying threats to biodiversity, and they provide the information needed to prioritize threats from alien species and the sites that need urgent management intervention.
Resumen: Organismos invasores poseen una amenaza grave para la biodiversidad global. La Península Cape, en Sudáfrica provee un caso de estudio de la amenaza de plantas invasoras sobre la diversidad nativa de plantas. Intentamos identificar si las plantas pueden invadir el paisaje y cual sería su amenaza sobre la diversidad de plantas. Esta información es necesaria para desarrollar una estrategia para el manejo de estas invasiones a escala de paisaje. Mediante el uso de modelos de regresion logística predecimos la distribución potencial de seis especies de plantas invasoras importantes en relación con diversas variables ambientales. Los modelos de regresión logística mostraron que las plantas invasoras podrían cubrir mas de un 89% de la Península Cape. Acacia cyclops y Pinus pinaster fueron predecidas como las especies que cubrirían mas área. Estas predicciones fueron sobrepuestas en un mapa de distribución actual de la diversidad de plantas nativas de la Península Cape con la intención de cuantificar la amenaza de las especies invasoras sobre la diversidad de especies nativas. Definimos la amenaza a la diversidad de plantas nativas como el número de especies de plantas nativas (divididas como: todas las especies, especies raras y amenazadas y especies endémicas) que tienen su rango total cubierto por la distribución predecida de plantas invasoras. Un modelo nulo que asume una distribucióñ aleatoria de sitios invadidos se empleó para evaluar si el área invadida estaba siendo confundida con la amenaza de la diversidad sobre plantas nativas. De hecho, el modelo de regresión logística predijo que Pinus pinaster amenaza a 350 especies nativas, 29 consideradas raras y amenazadas y 21 especies endémicas mas de lo que el modelo nulo predijo. Las comparaciones entre los modelos nulos y logísticos sugieren que la riqueza de especies y la invisabilidad estan positivamente correlacionadas y que la riqueza de especies es un indicador pobre de la resistencia contra invasiones en el área de estudio. Nuestros resultados hacen énfasis en la importancia de adoptar aproximanciones espacialmente explícitas para cuantificar las amenazas contra la biodiversidad y proveer la información necesaria para priorizar amenazas por especies invasoras y los sitios que necesitan una intervención de manejo urgente.