Graeme Newell

Predicting landscape-genetic consequences of habitat loss, fragmentation and mobility for multiple species of woodland birds.

Inference concerning the impact of habitat fragmentation on dispersal and gene flow is a key theme in landscape genetics. Recently, the ability of established approaches to identify reliably the differential effects of landscape structure (e.g. land-cover composition, remnant vegetation configuration and extent) on the mobility of organisms has been questioned. More explicit methods of predicting and testing for such effects must move beyond post hoc explanations for single landscapes and species. Here, we document a process for making a priori predictions, using existing spatial and ecological data and expert opinion, of the effects of landscape structure on genetic structure of multiple species across replicated landscape blocks. We compare the results of two common methods for estimating the influence of landscape structure on effective distance: least-cost path analysis and isolation-by-resistance. We present a series of alternative models of genetic connectivity in the study area, represented by different landscape resistance surfaces for calculating effective distance, and identify appropriate null models. The process is applied to ten species of sympatric woodland-dependant birds. For each species, we rank a priori the expectation of fit of genetic response to the models according to the expected response of birds to loss of structural connectivity and landscape-scale tree-cover. These rankings (our hypotheses) are presented for testing with empirical genetic data in a subsequent contribution. We propose that this replicated landscape, multi-species approach offers a robust method for identifying the likely effects of landscape fragmentation on dispersal.

On the selection of thresholds for predicting species occurrence with presence-only data.

Abstract Presence‐only data present challenges for selecting thresholds to transform species distribution modeling results into binary outputs. In this article, we compare two recently published threshold selection methods (maxSSS and maxF pb) and examine the effectiveness of the threshold‐based prevalence estimation approach. Six virtual species with varying prevalence were simulated within a real landscape in southeastern Australia. Presence‐only models were built with DOMAIN, generalized linear model, Maxent, and Random Forest. Thresholds were selected with two methods maxSSS and maxF pb with four presence‐only datasets with different ratios of the number of known presences to the number of random points (KP–RP ratio). Sensitivity, specificity, true skill statistic, and F measure were used to evaluate the performance of the results. Species prevalence was estimated as the ratio of the number of predicted presences to the total number of points in the evaluation dataset. Thresholds selected with maxF pb varied as the KP–RP ratio of the threshold selection datasets changed. Datasets with the KP–RP ratio around 1 generally produced better results than scores distant from 1. Results produced by We conclude that maxFpb had specificity too low for very common species using Random Forest and Maxent models. In contrast, maxSSS produced consistent results whichever dataset was used. The estimation of prevalence was almost always biased, and the bias was very large for DOMAIN and Random Forest predictions. We conclude that maxF pb is affected by the KP–RP ratio of the threshold selection datasets, but maxSSS is almost unaffected by this ratio. Unbiased estimations of prevalence are difficult to be determined using the threshold‐based approach.

Raising the bar for systematic conservation planning

Systematic conservation planning (SCP) represents a significant step toward cost-effective, transparent allocation of resources for biodiversity conservation. However, research demonstrates important consequences of uncertainties in SCP and of basing methods on simplified circumstances involving few real-world complexities. Current research often relies on single case studies with unknown forms and amounts of uncertainty as well as low statistical power for generalizing results. Consequently, conservation managers have little evidence for the true performance of conservation planning methods in their own complex, uncertain applications. To build effective and reliable methods in SCP, there is a need for more challenging and integrated testing of their robustness to uncertainty and complexity, and much greater emphasis on generalization to real-world situations.

Diet selection by the swamp wallaby (Wallabia bicolor): feeding strategies under conditions of changed food availability

Abstract Mammals can have either generalized (mixed) or specialized diets. We expected swamp wallabies (Wallabia bicolor) to have mixed diets, and predicted a negative relationship between the selection of a food type and its relative availability (negative frequency dependence). We collected data on diets in a native Eucalyptus forest where the relative availability of food types (defined as 5 plant functional groups: ferns, forbs, monocots, shrubs, and trees) had been altered by timber harvesting. A comparison of diets between individuals living in 2 habitat types (unharvested forest and 5-year-old regenerating areas) showed that in both habitats forbs were the major dietary component, although moderate amounts of shrubs and monocots also were consumed. Trees and ferns were eaten less at unharvested sites, and more at 5-year-old sites. Nonmetric multidimensional scaling followed by a multiresponse permutation procedure demonstrated a substantial difference in diet composition between the habitats (multiresponse permutation procedure: A = 0.20, P < 0.001), but when analyzed using an index of diet selection, the difference was smaller (A = 0.05, P = 0.04). Three alternative analyses demonstrated negative frequency dependence in many cases, a result generally consistent with a mixed feeding strategy. With the exception of tree foliage, selection was positively correlated with the relative availability of at least 1 other food type, and largely uncorrelated with 3 forage quality variables (nitrogen, water, and dry matter digestibility). Additional data at a finer resolution and in different seasons are required to test the generality of these conclusions.

The effect of sample size on the accuracy of species distribution models: considering both presences and pseudo-absences or background sites

Most high-performing species distribution modelling techniques require both presences, and either absences or pseudo-absences or background points. In this paper, we explore the effect of sample size, towards developing improved strategies for modelling. We generated 1800 virtual species with three levels of prevalence using ten modelling techniques, while varying the number of training presences (NTP) and the number of random points (NRP representing pseudo-absences or background sites). For five of the ten modelling techniques we built two versions of models: one with an equal total weight (ETW) setting where the total weight for pseudo-absence is equivalent to the total weight for presence, and another with an unequal total weight (UTW) setting where the total weight for pseudo-absence is not required to be equal to the total weight for presence. We compared two strategies for NRP: a small multiplier strategy (i.e. setting NRP at a few times as large as NTP), and a large number strategy (i.e. using numerous random points). We produced ensemble models (by averaging the predictions from 30 models built with the same set of training presences and different sets of random points in equivalent numbers) for three NTP magnitudes and two NRP strategies. We found that model accuracy altered as NRP increased with four distinct patterns of performance: increasing, decreasing, arch-shaped and horizontal. In most cases ETW improved model performance. Ensemble models had higher accuracy than the corresponding single models, and this improvement was pronounced when NTP was low. We conclude that a large NRP is not always an appropriate strategy. The best choice for NRP will depend on the modelling techniques used, species prevalence and NTP. We recommend building ensemble models instead of single models, using the small multiplier strategy for NRP with ETW, especially when only a small number of species presence records are available.