A comparison of absolute performance of different correlative and mechanistic species distribution models in an independent area
To investigate the comparative abilities of six different bioclimatic models in an independent area, utilizing the distribution of eight different species available at a global scale and in Australia. Global scale and Australia. We tested a variety of bioclimatic models for eight different plant species employing five discriminatory correlative species distribution models (SDMs) including Generalized Linear Model (GLM), MaxEnt, Random Forest (RF), Boosted Regression Tree (BRT), Bioclim, together with CLIMEX (CL) as a mechanistic niche model. These models were fitted using a training dataset of available global data, but with the exclusion of Australian locations. The capabilities of these techniques in projecting suitable climate, based on independent records for these species in Australia, were compared. Thus, Australia is not used to calibrate the models and therefore it is as an independent area regarding geographic locations. To assess and compare performance, we utilized the area under the receiver operating characteristic (ROC) curves (AUC), true skill statistic (TSS), and fractional predicted areas for all SDMs. In addition, we assessed satisfactory agreements between the outputs of the six different bioclimatic models, for all eight species in Australia. The modeling method impacted on potential distribution predictions under current climate. However, the utilization of sensitivity and the fractional predicted areas showed that GLM, MaxEnt, Bioclim, and CL had the highest sensitivity for Australian climate conditions. Bioclim calculated the highest fractional predicted area of an independent area, while RF and BRT were poor. For many applications, it is difficult to decide which bioclimatic model to use. This research shows that variable results are obtained using different SDMs in an independent area. This research also shows that the SDMs produce different results for different species; for example, Bioclim may not be good for one species but works better for other species. Also, when projecting a "large" number of species into novel environments or in an independent area, the selection of the "best" model/technique is often less reliable than an ensemble modeling approach. In addition, it is vital to understand the accuracy of SDMs' predictions. Further, while TSS, together with fractional predicted areas, are appropriate tools for the measurement of accuracy between model results, particularly when undertaking projections on an independent area, AUC has been proved not to be. Our study highlights that each one of these models (CL, Bioclim, GLM, MaxEnt, BRT, and RF) provides slightly different results on projections and that it may be safer to use an ensemble of models.
Are research efforts on Animalia in the South Pacific associated with the conservation status or population trends?
Analyses of knowledge gaps can highlight imbalances in research, encouraging greater proportionality in the distribution of research efforts. In this research we used generalized linear mixed models (GLMM) with the aim to determine if research efforts for the period 2005–2015 for terrestrial vertebrates of Amphibia, Aves, Mammalia and Reptilia in the South Pacific region were correlated with conservation status (critically endangered (CR), endangered (EN), vulnerable (VU), least concern (LC) and near threatened (NT)) or population trends (increasing, stable, decreasing and unknown) through the International Union for Conservation of Nature (IUCN) database. Our results showed that research distribution was uneven across different classes. Out of 633623 investigated papers, the average number of publications per species was 43.7, 306.7, 717.6 and 115.3 for Amphibia (284 species), Aves (1306 species), Mammalia (243 species) and Reptilia (400 species), respectively. Consistently, the lower publication effort on Amphibia compared to other taxonomic classes was revealed as significant by GLMM analysis. There was no significant differences in research effort among levels of conservation status. However, we found significantly different publication efforts among population trends of all examined species in that species with "unknown" population trends gained significantly lower researchers' attention compared to species with "decreasing" trend. Results also indicated that, although it was not significant, the highest attention is given to species with "increasing" population trend over all taxonomic classes. Using the Information Theoretic approach we also generated a set of competing models to identify most important factors influencing research efforts, revealing that the highest ranked model included taxonomic class and population.
Invasive weed species' threats to global biodiversity: Future scenarios of changes in the number of invasive species in a changing climate
Invasive weed species (IWS) threaten ecosystems, the distribution of specific plant species, as well as agricultural productivity. Predicting the impact of climate change on the current and future distributions of these unwanted species forms an important category of ecological research. Our study investigated 32 globally important IWS to assess whether climate alteration may lead to spatial changes in the overlapping of specific IWS globally. We utilized the versatile species distribution model MaxEnt, coupled with Geographic Information Systems, to evaluate the potential alterations (gain/loss/static) in the number of potential ecoregion invasions by IWS, under four Representative Concentration Pathways, which differ in terms of predicted year of peak greenhouse gas emission. We based our projection on a forecast of climatic variables (extracted from WorldClim) from two global circulation models (CCSM4 and MIROC-ESM). Initially, we modeled current climatic suitability of habitat, individually for each of the 32 IWS, identifying those with a common spatial range of suitability. Thereafter, we modeled the suitability of all 32 species under the projected climate for 2050, incorporating each of the four Representative Concentration Pathways (2.6, 4.5, 6.0, and 8.5) in separate models, again examining the common spatial overlaps. The discrimination capacity and accuracy of the model were assessed for all 32 IWS individually, using the area under the curve and true skill statistic rate, with results averaging 0.87 and 0.75 respectively, indicating a high level of accuracy. Our final methodological step compared the extent of the overlaps and alterations under the current and future projected climates. Our results mainly predicted decrease on a global scale, in areas of habitat suitable for most IWS, under future climatic conditions, excluding European countries, northern Brazil, eastern US, and south-eastern Australia. The following should be considered when interpreting these results: there are many inherent assumptions and limitations in presence-only data of this type, as well as with the modeling techniques projecting climate conditions, and the envelopes themselves, such as scale and resolution mismatches, dispersal barriers, lack of documentation on potential disturbances, and unknown or unforeseen biotic interactions.
Assessing Accuracy Methods of Species Distribution Models: AUC, Specificity, Sensitivity and the True Skill Statistic
We aimed to assess different methods for evaluating performance accuracy in species distribution models based on the application of five types of bioclimatic models under three threshold selections to predict the distributions of eight different species in Australia, treated as an independent area. Five discriminatory correlative species distribution models (SDMs), were used to predict the species distributions of eight different plants. A global training data set, excluding the Australian locations, was used for model fitting. Four accuracy measurement methods were compared under three threshold selections of i) maximum sensitivity + specificity, ii) sensitivity = specificity and iii) predicted probability of 0.5 (default). Results showed that the choice of modeling methods had an impact on potential distribution predictions for an independent area. Examination of the four accuracy methods underexamined threshold selections demonstrated that TSS is a more realistic and practical method, in comparison with AUC, Sensitivity and Specificity. Accurate projection of the distribution of a species is extremely complex.