Ecological niche modeling the potential geographic distribution of fourCulicoidesspecies of veterinary significance in Florida

Abstract

25Epizootic hemorrhagic disease (EHD) is a viral arthropod-borne disease affecting wild 26 and domestic ruminants. EHD virus (EHDV) is transmitted to vertebrate animal hosts by biting 27 midges in the genus Culicoides. Culicoides sonorensis Latreille is the only confirmed vector of 28 EHDV in the United States but is considered rare in Florida and not sufficiently abundant to 29 support EHDV transmission. This study used ecological niche modeling to map the potential 30 geographical distributions and associated ecological variable space of four Culicoides species 31 suspected of transmitting EHDV in Florida, including Culicoides insignis, Culicoides stellifer, 32 Culicoides debilipalpis and Culicoides venustus. Models were developed with the Genetic 33 Algorithm for Rule Set Production in DesktopGARP v1.1.3 using species occurrence data from 34 field sampling along with environmental variables from WorldClim and Trypanosomiasis and 35 Land use in Africa. For three Culicoides species (C. insignis, C. stellifer and C. debilipalpis) 96-36 98% of the presence points were predicted across the Florida landscape (63.77% -72.53%). 37 For C. venustus, models predicted 98.00% of presence points across 27.42% of Florida. 38 Geographic variations were detected between species. Culicoides insignis was predicted to be 39 restricted to peninsular Florida, and in contrast, C. venustus was predicted to be primarily in 40 north Florida and the panhandle region. Culicoides stellifer and C. debilipalpis were predicted 41 nearly statewide. Environmental conditions also differed by species, with some species' ranges 42 predicted by more narrow ranges of variables than others. The Normalized Difference 43 Vegetation Index (NDVI) was a major predictor of C. venustus and C. insignis presence. For C. 44 stellifer, Land Surface Temperature, Middle Infrared were the most limiting predictors of 45 presence. The limiting variables for C. debilipalpis were NDVI Bi-Annual Amplitude and NDVI 46 Annual Amplitude at 22.45% and 28.09%, respectively. The model outputs, including maps and 49 cannot exist in the environment without the vector, model outputs can be used to estimate the 50 potential risk of disease for animal hosts across Florida. Results also provide distribution and 51 habitat information useful for integrated pest management practices. 52 53

Exploring environmental coverages of species: a new variable contribution estimation methodology for rulesets from the genetic algorithm for rule-set prediction

“…Predictive accuracies for the best subsets from the GARP experiments with the UI-based reduced variable sets was evaluated using a combination of AUC, omission, and commission rates based on the external testing dataset (Lim & Klein, 2006;Peterson, Papeş & Eaton, 2007). The AUC, although not an ideal metric for accuracy estimation (Lobo, Jiménez-Valverde & Real, 2008), is useful to identify models that perform well (Hanley & McNeil, 1982;Mullins et al, 2013;Sloyer et al, 2018). The 10-model best Figure 2 Simulated species distributions, occurrence (presence-absence) maps, and GARP prediction map for the best subset under the two scenarios where the correlation between species occurrence and environment are weak and strong.…”

“…Although several algorithms and advanced methods have been introduced, and many, like MaxEnt, frequently used in the literature, GARP is still widely used to model species distribution and understand their ecological affinities. Some recent applications include predicting distributions of different species, such as the invasive species (e.g., pignut in India Padalia, Srivastava & Kushwaha, 2014 and creeping oxeye in Central America, Qin et al, 2015), modeling bird abundance patterns (Martínez-Meyer et al, 2013), endangered bird species (Montenegro et al, 2017), and ecological niche of tree species (Prakash Singh et al, 2013), and delineating disease risk areas by estimating the geographical distribution of pathogens (Barro et al, 2016;Chikerema et al, 2017) and vector species (Ramsey et al, 2015;Sloyer et al, 2018;Lippi et al, 2019). Other research compares GARP with some other ENMs (especially MaxEnt) to show how species distributions change using different approaches to provide reliable predictions (Padalia, Srivastava & Kushwaha, 2014;Wang et al, 2017;Ray, Behera & Jacob, 2018), to compare the predictive performance of different methods (Khatchikian et al, 2011;Zhu & Peterson, 2017), or to understand why the differences in the performance exist (Elith & Graham, 2009).…”