Dr Ian Renner

Citizen science programs, and particularly atlas schemes based on opportunistic biological records, are very important sources of data for species distribution models and conservation. Nevertheless, these data are prone to bias, particularly when they come from less popular or hard to detect/identify species, such as insects. With such biased data, it is important to evaluate the stability of the model predictions. In recent years, point process models (PPMs) have shown their strength as a unifying framework to fit presence-only species distribution models with many advantages in model implementation and interpretation; PPMs are closely connected to methods already in widespread use in ecology such as MaxEnt and to logistic regression and benefit from being more transparent about resource selection and absence handling. Moreover, there is a well-developed set of tools to fit these models and assess various features of the underlying model, including model stability. However, such tools are currently unavailable when point process models are fitted with a lasso penalty, which has been shown to improve predictive performance. Based on the French citizen science program ¿Stag beetle Quest¿, we propose new methods to assess model stability in this context. The ultimate goal was to develop a set of functions to analyze PPM models with lasso penalties fitted with presence-only data. To assess model stability, we randomly sampled different subsets of locations with varying size from the whole dataset and used the proposed tools to compare fitted intensities and model coefficients. All the developed measures are complementary and can be used to identify at what number of point locations the model stabilizes, which will be dependent on the dataset. Our work presents a new toolbox to explore questions around model stability based on the number of locations in the context of point process models with a lasso penalty and confirms once more the use of the point process modeling framework as a flexible and unifying framework to fit presence-only species distribution models.

Presence-only data are widely used for species distribution modelling, and point process regression models are a flexible tool that has considerable potential for this problem, when data arise as point events. In this paper, we review point process models, some of their advantages and some common methods of fitting them to presence-only data. Advantages include (and are not limited to) clarification of what the response variable is that is modelled; a framework for choosing the number and location of quadrature points (commonly referred to as pseudo-absences or 'background points') objectively; clarity of model assumptions and tools for checking them; models to handle spatial dependence between points when it is present; and ways forward regarding difficult issues such as accounting for sampling bias. Point process models are related to some common approaches to presence-only species distribution modelling, which means that a variety of different software tools can be used to fit these models, including maxent or generalised linear modelling software.

Modeling the spatial distribution of a species is a fundamental problem in ecology. A number of modeling methods have been developed, an extremely popular one being MAXENT, a maximum entropy modeling approach. In this article, we show that MAXENT is equivalent to a Poisson regression model and hence is related to a Poisson point process model, differing only in the intercept term, which is scale-dependent in MAXENT. We illustrate a number of improvements to MAXENT that follow from these relations. In particular, a point process model approach facilitates methods for choosing the appropriate spatial resolution, assessing model adequacy, and choosing the LASSO penalty parameter, all currently unavailable to MAXENT. The equivalence result represents a significant step in the unification of the species distribution modeling literature. © 2013, The International Biometric Society.

Presence-only data, where information is available concerning species presence but not species absence, are subject to bias due to observers being more likely to visit and record sightings at some locations than others (hereafter "observer bias"). In this paper, we describe and evaluate a model-based approach to accounting for observer bias directly - by modelling presence locations as a function of known observer bias variables (such as accessibility variables) in addition to environmental variables, then conditioning on a common level of bias to make predictions of species occurrence free of such observer bias. We implement this idea using point process models with a LASSO penalty, a new presence-only method related to maximum entropy modelling, that implicitly addresses the "pseudo-absence problem" of where to locate pseudo-absences (and how many). The proposed method of bias-correction is evaluated using systematically collected presence/absence data for 62 plant species endemic to the Blue Mountains near Sydney, Australia. It is shown that modelling and controlling for observer bias significantly improves the accuracy of predictions made using presence-only data, and usually improves predictions as compared to pseudo-absence or "inventory" methods of bias correction based on absences from non-target species. Future research will consider the potential for improving the proposed bias-correction approach by estimating the observer bias simultaneously across multiple species. © 2013 Warton et al.