Maxent modelling for habitat suitability of vulnerable tree <i>Dalbergia latifolia</i> in Nepal

Abstract

Roxb., commonly known as rosewood, is one of the highly valuable tropical timber species of Nepal. The tree species was widely distributed in the past, however, over-exploitation of natural habitat, deforestation, forest conversion for agriculture, illegal logging and the invasion of alien species resulted in the classification of this species as vulnerable by the IUCN (International Union for Conservation of Nature) category. So, the prediction of habitat suitability and potential distribution of the species is required to develop restoration mechanisms and conservation interventions. In this study, we modelled the suitable habitat of over the entire possible range of Nepal using a Maxent model. We compiled 23 environmental variables (19 bioclimatic, 3 topographic and a vegetative layer), however, only 12 least correlated variables along with 43 spatially representative presence locations were retained for model prediction. We used a receiver operating characteristic (ROC) curve to assess the model’s performance and a Jackknife procedure to evaluate the relative importance of predictor variables. The model was statistically significant with an area under the curve (AUC) value of 0.969. The internal Jackknife test indicated that elevation was the most important variable for the model prediction with 71.3% contribution followed by mean temperature of driest quarter (9.8%). The most (&gt;0.6) suitable habitat for the was 235 484 hectares with large sections of area in two provinces whereas, the western most provinces were not suitable for as per Maxent model. The information presented here can provide a framework for nature conservation planning, monitoring and habitat management of this rare and endangered species.Dalbergia latifoliaD. latifoliaD. latifoliaD. latifolia

Climate‐induced range shifts of invasive species (Diaphorina citri Kuwayama)

“…The final suitability maps have values ranging from 0 to 1, indicating the probability ( p ) of occurrence of the pest. The Maximum Test Sensitivity Plus Specificity threshold (MTSPS), considered simple and effective and at least as good as other more complicated approaches, was chosen to separate the final model's suitable and unsuitable areas for ACP, 55, 68 which was further regrouped into four classes for interpretation: unsuitable p ≤ MTSPS; low suitability MTSPS < p ≤ 0.4; medium suitability 0.4 < p ≤ 0.6; and high suitability p > 0.6 69,70,71 …”

“…The Maximum Test Sensitivity Plus Specificity threshold (MTSPS), considered simple and effective and at least as good as other more complicated approaches, was chosen to separate the final model's suitable and unsuitable areas for ACP, 55,68 which was further regrouped into four classes for interpretation: unsuitable p ≤ MTSPS; low suitability MTSPS < p ≤ 0.4; medium suitability 0.4 < p ≤ 0.6; and high suitability p > 0.6. 69,70,71 Response curves were generated in MaxEnt, and we chose only the environmental variables that represented strong relationships between the probabilities of the presence for the species with each environmental predictor. All the response curves were evaluated in terms of providing powerful biological logic.…”

Predicting the distribution of suitable habitat of the poisonous weed Astragalus variabilis in China under current and future climate conditions

“…It is less affected by the occurrence:background point ratio and species prevalence, which reduces omission errors for low-prevalence species and misclassification errors for high-prevalence species ( Liu et al, 2016 ). The habitat suitability of A. variabilis was reclassified into four categories according to the output logistic value: unsuitable (0–MTSPS), low suitability (MTSPS–0.4), moderate suitability (0.4–0.6), and high suitability (0.6–1.0) ( Zhang et al, 2018 ; Mahatara et al, 2021 ).…”

Prediction of global potential suitable habitats of Nicotiana alata Link et Otto based on MaxEnt model

“…Among these models, MaxEnt is the most widely used ecological niche model because of its easy operation, fast running speed 19 , and accurate prediction of the potential suitable habitats of species based on the longitude and latitude of known presence points and climate factors or even small sample data. MaxEnt has been widely used in invasion management of biological species [20][21][22] , protection of endangered plants and animals [23][24][25][26] , control of pests and diseases 19 , and environmental risk assessment of species habitats 27 .…”