Forecasting species distributions: Correlation does not equal causation

Accounting for missing ticks: Use (or lack thereof) of hierarchical models in tick ecology studies

Ixodid (hard) ticks play important ecosystem roles and have significant impacts on animal and human health via tick-borne diseases and physiological stress from parasitism. Tick occurrence, abundance, activity, and key life-history traits are highly influenced by host availability, weather, microclimate, and landscape features. As such, changes in the environment can have profound impacts on ticks, their hosts, and the spread of diseases. Researchers recognize that spatial and temporal factors influence activity and abundance and attempt to account for both by conducting replicate sampling bouts spread over the tick questing period. However, common field methods notoriously underestimate abundance, and it is unclear how (or if) tick studies model the confounding effects of factors influencing activity and abundance. This step is critical as unaccounted variance in detection can lead to biased estimates of occurrence and abundance. We performed a descriptive review to evaluate the extent to which studies account for the detection process while modeling tick data. We also categorized the types of analyses that are commonly used to model tick data. We used hierarchical models (HMs) that account for imperfect detection to analyze simulated and empirical tick data, demonstrating that inference is muddled when detection probability is not accounted for in the modeling process. Our review indicates that only 5 of 412 (1 %) papers explicitly accounted for imperfect detection while modeling ticks. By comparing HMs with the most common approaches used for modeling tick data (e.g., ANOVA), we show that population estimates are biased low for simulated and empirical data when using non-HMs, and that confounding occurs due to not explicitly modeling factors that influenced both detection and abundance. Our review and analysis of simulated and empirical data shows that it is important to account for our ability to detect ticks using field methods with imperfect detection. Not doing so leads to biased estimates of occurrence and abundance which could complicate our understanding of parasite-host relationships and the spread of tick-borne diseases. We highlight the resources available for learning HM approaches and applying them to analyzing tick data.

Towards exhaustive community ecology via DNA metabarcoding

Exhaustive biodiversity data, covering all the taxa in an environment, would be fundamental to understand how global changes influence organisms living at different trophic levels, and to evaluate impacts on interspecific interactions. Molecular approaches such as DNA metabarcoding are boosting our ability to perform biodiversity inventories. Nevertheless, even though a few studies have recently attempted exhaustive reconstructions of communities, holistic assessments remain rare. The majority of metabarcoding studies published in the last years used just one or two markers and analysed a limited number of taxonomic groups. Here, we provide an overview of emerging approaches that can allow all-taxa biological inventories. Exhaustive biodiversity assessments can be attempted by combining a large number of specific primers, by exploiting the power of universal primers, or by combining specific and universal primers to obtain good information on key taxa while limiting the overlooked biodiversity. Multiplexes of primers, shotgun sequencing and capture enrichment may provide a better coverage of biodiversity compared to standard metabarcoding, but still require major methodological advances. Here, we identify the strengths and limitations of different approaches, and suggest new development lines that might improve broad scale biodiversity analyses in the near future. More holistic reconstructions of ecological communities can greatly increase the value of metabarcoding studies, improving understanding of the consequences of ongoing environmental changes on the multiple components of biodiversity.

Into the Trees: Random Forests for Predicting Fusarium Head Blight Epidemics of Wheat in the United States

Constructing models that accurately predict Fusarium head blight (FHB) epidemics and are also amenable to large-scale deployment is a challenging task. In the United States, the emphasis has been on simple logistic regression (LR) models, which are easy to implement but may suffer from lower accuracies when compared with more complicated, harder-to-deploy (over large geographies) model frameworks such as functional or boosted regressions. This article examined the plausibility of random forests (RFs) for the binary prediction of FHB epidemics as a possible mediation between model simplicity and complexity without sacrificing accuracy. A minimalist set of predictors was also desirable rather than having the RF model use all 90 candidate variables as predictors. The input predictor set was filtered with the aid of three RF variable selection algorithms (Boruta, varSelRF, and VSURF), using resampling techniques to quantify the variability and stability of selected variable sets. Post-selection filtering produced 58 competitive RF models with no more than 14 predictors each. One variable representing temperature stability in the 20 days before anthesis was the most frequently selected predictor. This was a departure from the prominence of relative humidity-based variables previously reported in LR models for FHB. The RF models had overall superior predictive performance over the LR models and may be suitable candidates for use by the Fusarium Head Blight Prediction Center.

Contrasting seasonal effects of climate change influence density in a cold-adapted species

Many ecological processes are profoundly influenced by abiotic factors, such as temperature and snow. However, despite strong evidence linking shifts in these ecological processes to corresponding shifts in abiotic factors driven by climate change, the mechanisms connecting population size to season-specific climate drivers are little understood. Using a 21-year dataset and a Bayesian state space model, we identified biologically informed seasonal climate covariates that influenced densities of snowshoe hares (Lepus americanus), a cold-adapted boreal herbivore. We found that snow and temperature had strong but conflicting season-dependent effects. Reduced snow duration in spring and fall and warmer summers were associated with lowered hare density, whereas warmer winters were associated with increased density. When modeled simultaneously and under two climate change scenarios, the negative effects of reduced fall and spring snow duration and warmer summers overwhelm the positive effect of warmer winters, producing projected population declines. Ultimately, the contrasting population-level impacts of climate change across seasons emphasize the critical need to examine the entire annual climate cycle to understand potential long-term population consequences of climate change.

Forest tree species distribution for Europe 2000-2020: mapping potential and realized distributions using spatiotemporal machine learning

This article describes a data-driven framework based on spatiotemporal machine learning to produce distribution maps for 16 tree species (Abies alba Mill., Castanea sativa Mill., Corylus avellana L., Fagus sylvatica L., Olea europaea L., Picea abies L. H. Karst., Pinus halepensis Mill., Pinus nigra J. F. Arnold, Pinus pinea L., Pinus sylvestris L., Prunus avium L., Quercus cerris L., Quercus ilex L., Quercus robur L., Quercus suber L. and Salix caprea L.) at high spatial resolution (30 m). Tree occurrence data for a total of three million of points was used to train different algorithms: random forest, gradient-boosted trees, generalized linear models, k-nearest neighbors, CART and an artificial neural network. A stack of 305 coarse and high resolution covariates representing spectral reflectance, different biophysical conditions and biotic competition was used as predictors for realized distributions, while potential distribution was modelled with environmental predictors only. Logloss and computing time were used to select the three best algorithms to tune and train an ensemble model based on stacking with a logistic regressor as a meta-learner. An ensemble model was trained for each species: probability and model uncertainty maps of realized distribution were produced for each species using a time window of 4 years for a total of six distribution maps per species, while for potential distributions only one map per species was produced. Results of spatial cross validation show that the ensemble model consistently outperformed or performed as good as the best individual model in both potential and realized distribution tasks, with potential distribution models achieving higher predictive performances (TSS = 0.898, R2 logloss = 0.857) than realized distribution ones on average (TSS = 0.874, R2 logloss = 0.839). Ensemble models for Q. suber achieved the best performances in both potential (TSS = 0.968, R2 logloss = 0.952) and realized (TSS = 0.959, R2 logloss = 0.949) distribution, while P. sylvestris (TSS = 0.731, 0.785, R2 logloss = 0.585, 0.670, respectively, for potential and realized distribution) and P. nigra (TSS = 0.658, 0.686, R2 logloss = 0.623, 0.664) achieved the worst. Importance of predictor variables differed across species and models, with the green band for summer and the Normalized Difference Vegetation Index (NDVI) for fall for realized distribution and the diffuse irradiation and precipitation of the driest quarter (BIO17) being the most frequent and important for potential distribution. On average, fine-resolution models outperformed coarse resolution models (250 m) for realized distribution (TSS = +6.5%, R2 logloss = +7.5%). The framework shows how combining continuous and consistent Earth Observation time series data with state of the art machine learning can be used to derive dynamic distribution maps. The produced predictions can be used to quantify temporal trends of potential forest degradation and species composition change.