scholarly journals Roughing it: terrain is crucial in identifying novel translocation sites for the vulnerable brush-tailed rock-wallaby ( Petrogale pencillata )

2020 ◽  
Vol 7 (12) ◽  
pp. 201603
Author(s):  
Shane D. Morris ◽  
Christopher N. Johnson ◽  
Barry W. Brook
Keyword(s):  
Threatened Species ◽  
Species Distribution Models ◽  
Predictive Accuracy ◽  
Important Variable ◽  
Eastern Coast ◽  
Isolated Populations ◽  
Distribution Models ◽  
Promising Candidate ◽  
In The Wild ◽  
Chronic Threat

Translocations—the movement of species from one place to another—are likely to become more common as conservation attempts to protect small isolated populations from threats posed by extreme events such as bushfires. The recent Australian mega-fires burnt almost 40% of the habitat of the brush-tailed rock-wallaby ( Petrogale pencillata ), a threatened species whose distribution is already restricted, primarily due to predation by invasive species. This chronic threat of over-predation, coupled with the possible extinction of the genetically distinct southern population (approx. 40 individuals in the wild), makes this species a candidate for a conservation translocation. Here, we use species distribution models to identify translocation sites for the brush-tailed rock-wallaby. Our models exhibited high predictive accuracy, and show that terrain roughness, a surrogate for predator refugia, is the most important variable. Tasmania, which currently has no rock-wallabies, showed high suitability and is fox-free, making it a promising candidate site. We outline our argument for the trial translocation of rock-wallaby to Maria Island, located off Tasmania's eastern coast. This research offers a transparent assessment of the translocation potential of a threatened species, which can be adapted to other taxa and systems.

scholarly journals Transferability and scalability of species distribution models: a test with sedentary marine invertebrates

2017 ◽  
Vol 74 (5) ◽  
pp. 766-778 ◽  
Author(s):  
Aaron M. Eger ◽  
Janelle M.R. Curtis ◽  
Marie-Josée Fortin ◽  
Isabelle M. Côté ◽  
Frédéric Guichard
Keyword(s):  
Species Distribution ◽  
Species Distribution Models ◽  
Marine Invertebrates ◽  
Predictive Accuracy ◽  
Distribution Models ◽  
Initial Information ◽  
Reference Models ◽  
Different Levels ◽  
Better Than

We found the predictive accuracy of species distribution models (SDMs) for sedentary marine invertebrates to be dependent on the methodology of their application. We explored three applications of SDMs: first a model tested at a scale smaller than at which it was trained (downscaled), second a model tested at scale larger than its training scale (upscaled), and third a model tested at the same scale but outside the extent for which it was trained (transferred). The accuracies of these models were compared with the “reference” models that were trained and tested at the same scale and extent. We found that downscaled SDMs had higher predictive accuracy than reference SDMs. Transferred and upscaled models had lower predictive accuracy than their reference counterparts but still performed better than random, making them potentially acceptable alternatives where information is lacking for imminent decisions or in cost-restricted scenarios. Our results provide insights into the techniques available for researchers and managers developing SDMs at varying scales, with different species, and with different levels of initial information.

scholarly journals Embracing ensemble species distribution models to inform at-risk species status assessments

2021 ◽  
Author(s):  
Carlos Ramirez-Reyes ◽  
Mona Nazeri ◽  
Garrett Street ◽  
D. Todd Jones-Farrand ◽  
Francisco Vilella ◽  
...  
Keyword(s):  
At Risk ◽  
Species Distribution ◽  
Species Distribution Models ◽  
Predictive Accuracy ◽  
Calibration Data ◽  
Species Status ◽  
Distribution Models ◽  
Modeling Approach ◽  
Modeling Techniques ◽  
Consequent Improvement

Conservation planning depends on reliable information regarding the geographic distribution of species. However, our knowledge of species' distributions is often incomplete, especially when species are cryptic, difficult to survey, or rare. The use of species distribution models has increased in recent years and proven a valuable tool to evaluate habitat suitability for species. However, practitioners have yet to fully adopt the potential of species distribution models to inform conservation efforts for information-limited species. Here, we describe a species distribution modeling approach for at-risk species that could better inform U.S. Fish and Wildlife Service’s species status assessments and help facilitate conservation decisions. We applied four modeling techniques (generalized additive, maximum entropy, generalized boosted, and weighted ensemble) to occurrence data for four at-risk species proposed for listing under the U.S. Endangered Species Act (Papaipema eryngii, Macbridea caroliniana, Scutellaria ocmulgee and Balduina atropurpurea) in the Southeastern U.S. The use of ensemble models reduced uncertainty caused by differences among modeling techniques, with a consequent improvement of predictive accuracy of fitted models. Incorporating an ensemble modeling approach into species status assessments and similar frameworks is likely to benefit survey efforts, inform recovery activities, and provide more robust status assessments for at-risk species. We emphasize that co-producing species distribution models in close collaboration with species experts has the potential to provide better calibration data and model refinements, which could ultimately improve reliance and use of model outputs.

scholarly journals Species Distribution Models and Ecological Suitability Analysis for Potential Tick Vectors of Lyme Disease in Mexico

2012 ◽  
Vol 2012 ◽  
pp. 1-10 ◽  
Author(s):  
Patricia Illoldi-Rangel ◽  
Chissa-Louise Rivaldi ◽  
Blake Sissel ◽  
Rebecca Trout Fryxell ◽  
Guadalupe Gordillo-Pérez ◽  
...  
Keyword(s):  
Lyme Disease ◽  
Species Distribution ◽  
Species Distribution Models ◽  
Eastern Coast ◽  
Suitability Analysis ◽  
Amblyomma Cajennense ◽  
Distribution Models ◽  
Geographic Ranges ◽  
Lower Temperature ◽  
Tick Vectors

Species distribution models were constructed for tenIxodesspecies andAmblyomma cajennensefor a region including Mexico and Texas. The model was based on a maximum entropy algorithm that used environmental layers to predict the relative probability of presence for each taxon. For Mexico, species geographic ranges were predicted by restricting the models to cells which have a higher probability than the lowest probability of the cells in which a presence record was located. There was spatial nonconcordance between the distributions ofAmblyomma cajennenseand theIxodesgroup with the former restricted to lowlands and mainly the eastern coast of Mexico and the latter to montane regions with lower temperature. The risk of Lyme disease is, therefore, mainly present in the highlands where someIxodesspecies are known vectors; ifAmblyomma cajennenseturns out to be a competent vector, the area of risk also extends to the lowlands and the east coast.

scholarly journals A Robust Prediction Model for Species Distribution Using Bagging Ensembles with Deep Neural Networks

2021 ◽  
Vol 13 (8) ◽  
pp. 1495
Author(s):  
Jehyeok Rew ◽  
Yongjang Cho ◽  
Eenjun Hwang
Keyword(s):  
Neural Networks ◽  
Species Distribution ◽  
Best Practice ◽  
Deep Neural Networks ◽  
Species Distribution Models ◽  
Species Distribution Model ◽  
Environmental Data ◽  
Distribution Model ◽  
Distribution Models ◽  
Robust Prediction

Species distribution models have been used for various purposes, such as conserving species, discovering potential habitats, and obtaining evolutionary insights by predicting species occurrence. Many statistical and machine-learning-based approaches have been proposed to construct effective species distribution models, but with limited success due to spatial biases in presences and imbalanced presence-absences. We propose a novel species distribution model to address these problems based on bootstrap aggregating (bagging) ensembles of deep neural networks (DNNs). We first generate bootstraps considering presence-absence data on spatial balance to alleviate the bias problem. Then we construct DNNs using environmental data from presence and absence locations, and finally combine these into an ensemble model using three voting methods to improve prediction accuracy. Extensive experiments verified the proposed model’s effectiveness for species in South Korea using crowdsourced observations that have spatial biases. The proposed model achieved more accurate and robust prediction results than the current best practice models.

Sign in / Sign up

Export Citation Format

Share Document