Fish recruitment prediction, using robust supervised classification methods

Improving our ability to predict recruitment is a key element in fisheries management. However, the interactions between population dynamics and different environmental factors are complex and often non-linear, making it difficult to produce robust predictions. ‘Machine-learning’ techniques (in particular, supervised classification methods) have been proposed as useful tools, to overcome such difficulties. In this study, a methodology is proposed to build a robust classifier for fish recruitment prediction with sparse and noisy data. The methodology consists of 4 steps: (1) a semi-automated recruitment discretization method; (2) supervised discretization of predictors; (3) multivariate and non-redundant predictors selection; (4) learning a probabilistic classifier. In terms of fisheries management, the classifier estimated performance has important consequences and, to be useful, the manager needs to know the risk that is being taken when using this number. Probabilistic classifiers such as ‘naive Bayes’, have the advantage that, in addition to the predictions, estimate also the probability of each possible outcome. Anchovy (Engraulis encrasicolus) and hake (Merluccius merluccius) recruitments are used as application examples. ‘Two-intervals’ recruitment discretization accomplishes 70% accuracies and Brier scores of around 0.10, for both anchovy and hake recruitment. In comparison, ‘three-intervals’ recruitment discretization accomplishes 50% accuracies; and Brier scores of around 0.25 for anchovy and 0.30 for hake recruitment. These statistics are the result of validating not only the classifier, but also the previous steps, as a whole methodology.     

Future distribution modelling: A stitch in time is not enough

The last two decades have seen an increasing number of studies assessing the impact of climate change upon biodiversity. A central assumption underpinning research into the potential future habitat of terrestrial biota is that species are presently in equilibrium with their environments and that quantitative climate models adequately represent the distribution of species. Recently, many alarming predictions have emerged concerning the extinction and redistribution of species. Here, we show that even large-scale models of the climatic niche dimensions of species are temporally variable. Distributional models were developed for Salix (willow) species occurring in the province of Ontario, Canada, using three historical climate data sets. Although historical data very accurately represented the distributions of willows, the inherent variability within the models of species based on different periods greatly influenced the direction and magnitude of projected distributional change. We expose a fundamental uncertainty with respect to predicting the responses of species to climate change.     

Modelling carbon and nitrogen dynamics in forest ecosystems of Central Russia under different climate change scenarios and forest management regimes

The individual-based stand-level model EFIMOD was used for large-scale simulations using standard data on forest inventories as model inputs. The model was verified for the case-study of field observations, and possible sources of uncertainties were analysed. The approach developed kept the ability for fine-tuning to account for spatial discontinuity in the simulated area. Several forest management regimes were simulated as well as forest wildfires and climate changes. The greatest carbon and nitrogen accumulations were observed for the regime without cuttings. It was shown that cuttings and wildfires strongly influence the processes of carbon and nitrogen accumulations in both soil and forest vegetation. Modelling also showed that the increase in annual average temperatures resulted in the partial relocation of carbon and nitrogen stocks from soil to plant biomass. However, forest management, particularly harvesting, has a greater effect on the dynamics of forest ecosystems than the prescribed climate change.     

Modelling phytoclimatic versatility as a large scale indicator of adaptive capacity to climate change in forest ecosystems

CLIMPAIR is a new phytoclimatic model, correlative and niche-based, which simultaneously assesses non-linear, non-statistical and dual measurements of proximity/potentiality of a site with respect to a number of climatic ranges of species, defined by convex hulls, within a suitability space. This set of phytoclimatic distances makes it possible to evaluate the degree to which each species is suitable for that site. Considering not only the number of species compatible (expected species richness), but also all those compatible covers presenting a high level of suitability evenness and finally applying an indicator derived from Shannon's classic entropy index to the set of standardized phytoclimatic coordinates in the suitability hyperspace, we can evaluate the phytoclimatic entropy which may be considered as a means of estimating the phytoclimatic versatility of the site. A site with high phytoclimatic entropy would promise versatile future behaviour, characterized by a wide range of possibilities of adaptation to climate change, and hence versatility can be used as an index of resilience and ability of a forest ecosystem to adapt to climate change. The model has been applied to peninsular Spain for 18 forest tree species and 12 climatic variables between the current mean climate (period 1951-1999) and a future climatic scenario (period 2040-2069). The results generally point to a significant decrease in the versatility of forest tree formations in the area studied, which is not homogeneous owing to a dual altitudinal/latitudinal decoupling. The decrease in versatility is greater in Mediterranean biogeographical areas than in Euro-Siberian ones, where in some cases it actually increases. In altitudinal terms, areas at elevations of less than 1500 m tend to become less versatile than areas situated at higher elevations, where versatility increases partly as a result of enrichment of alpine conifer forests with broadleaf species.     

Modeling amphibian energetics, habitat suitability, and movements of western toads, Anaxyrus (=Bufo) boreas, across present and future landscapes

Effective conservation of amphibian populations requires the prediction of how amphibians use and move through a landscape. Amphibians are closely coupled to their physical environment. Thus an approach that uses the physiological attributes of amphibians, together with knowledge of their natural history, should be helpful. We used Niche Mapper™ to model the known movements and habitat use patterns of a population of Western toads (Anaxyrus (=Bufo) boreas) occupying forested habitats in southeastern Idaho. Niche Mapper uses first principles of environmental biophysics to combine features of topography, climate, land cover, and animal features to model microclimates and animal physiology and behavior across landscapes. Niche Mapper reproduced core body temperatures (T_c) and evaporation rates of live toads with average errors of 1.6 ± 0.4 °C and 0.8 ± 0.2 g/h, respectively. For four different habitat types, it reproduced similar mid-summer daily temperature patterns as those measured in the field and calculated evaporation rates (g/h) with an average error rate of 7.2 ± 5.5%. Sensitivity analyses indicate these errors do not significantly affect estimates of food consumption or activity. Using Niche Mapper we predicted the daily habitats used by free-ranging toads; our accuracy for female toads was greater than for male toads (74.2 ± 6.8% and 53.6 ± 15.8%, respectively), reflecting the stronger patterns of habitat selection among females. Using these changing to construct a cost surface, we also reconstructed movement paths that were consistent with field observations. The effect of climate warming on toads depends on the interaction of temperature and atmospheric moisture. If climate change occurs as predicted, results from Niche Mapper suggests that climate warming will increase the physiological cost of landscapes thereby limiting the activity for toads in different habitats.     

Choice of threshold alters projections of species range shifts under climate change

One of the least explored sources of algorithmic uncertainty in bioclimatic envelope models (BEM) is the selection of thresholds to transform modelled probabilities of occurrence (or indices of suitability) into binary predictions of species presence and absence. We investigate the impacts of such thresholds in the specific context of climate change. BEM for European tree species were fitted combining 9 climatic models and emissions scenarios, 7 modelling techniques, and 14 threshold-setting techniques. We quantified sources of uncertainty in projections of turnover, and found that the choice of the modelling technique explained most of the variability (39%), while threshold choice explained 25% of the variability in the results, and their interaction an additional 19%. Choice of future climates explained 9% of total variability among projections. Estimated species range shifts obtained by applying different thresholds and models were grouped by IUCN-based categories of threat. Thresholds had a large impact on the inferred risks of extinction, producing 1.7- to 9.9-fold differences in the proportions of species projected to become threatened by climate change. Results demonstrate that threshold selection has large - albeit often unappreciated - consequences for estimating species range shifts under climate change.     

A concept model to estimate the potential distribution of the Asiatic citrus psyllid (Diaphorina citri Kuwayama) in Australia under climate change—A means for assessing biosecurity risk

Increasing global temperatures as a result of climate change are widely considered inevitable for Australia. Despite this, the specific effects of climate change on Australian agriculture are little studied and the effects on agricultural pests and diseases are virtually unknown. In this paper we consider the impact of climate change on the Asiatic citrus psyllid (Diaphorina citri Kuwayama [Hemiptera: Psyllidae]); one of two known vectors of huanglongbing (citrus greening); a debilitating disease which is caused in Asia by a phloem-limited bacterium ‘Candidatus Liberibacter asiaticus’ (α-Proteobacteria). D. citri does not occur in Australia, but if introduced would pose a major threat to the viability of the Australian citrus industry and to native Citrus species. This paper presents an approach developed to understand how climate change may influence the behaviour, distribution and breeding potential of D. citri. Here we developed and describe an initial dynamic point model of D. citri biology in relation to its citrus host and applied it to a scenario of increasing temperatures, as indicators of climate change, on a continental scale. A comparison between model outputs for the three time frames considered (1990, 2030 and 2070) confirms that increasing temperatures projected under climate change will affect the timing and duration of new citrus growth (flush) necessary for psyllid development throughout Australia. Flushing will start progressively earlier as the temperature increases and be of shorter duration. There will also be a gradual southward expansion of shorter durations of the occurrence of flush. Increasing temperatures will impact on D. citri both directly through alteration of its temperature dependant development cycle and indirectly through the impact on the host flushing cycle. For the whole of Australia, a comparison between model outputs for the three scenarios considered indicates the seasonality of D. citri development will change to match changes in citrus flush initiation. Results indicate that the risk of establishment by D. citri is projected to decrease under increasing temperatures, mainly due to shortened intervals when it can feed on new leaf flushes of the host. However, the spatially heterogeneous results also suggest that regions located on the southern coastline of Australia could become more suitable for D. citri than projected under current temperatures. These results confirm the value of a linked host-pest approach as based on D. citri climatic requirements alone the model would have accounted only for shorter development periods and predicted an increased risk of potential distribution.