It seems that the final publication of our special issue on species distribution models and the niche is drawing nearer, and to bridge the time for the (admittedly, probably sizable) group of those are checking their RSS feeds daily in expectation of this event, I would like to highlight three other papers of this special issue that have recently become available online. All of them are concerned with inference of process-based or process-oriented distribution models, and together with the other two papers I already blogged about, we cover a relatively wide range of different approaches, applications and issues around the topic of adopting a more process-based modeling approach to understand species distributions.
The first paper “How to understand species’ niches and range dynamics: a demographic research agenda for biogeography” led by Frank Schurr, is a review and perspective on what Frank and Jörn have termed “Dynamic range models (DRMs)”, that is, dynamic hierarchical statistical models for predicting species distributions that are essentially composed of three levels 1) a demographic model that relates environmental predictors to growth 2) a dispersal model 3) an observation model. As opposed to hybrid SDMs, DRMs allow estimating all these processes in parallel, which promises to avoid some of the problems of (hybrid) SDMs with source-sink dynamics and non-equilibrium situations.
Figure 1: The demographic basis of Hutchinsonian niches, range dynamics and biogeographical data. From Schurr et al., in press. Copyright see publisher.
The second paper, “Parameter and uncertainty estimation for process-oriented population and distribution models: data, statistics and the niche” led by Glenn Marion covers a very similar topic, namely hierarchical SDMs (or state-space models), but with the emphasis shifted more towards the technical issues for estimating models with different complexity and different data types. We do so by discussing three case studies, starting with a simple hierarchical SDM that is inferred from static data, over the case of a plant invasion described by several snapshots in time, up to a dynamic example that uses heterogeneous data.
Finally, in “A physiological analogy of the niche for projecting the potential distribution of plants”, Steve Higgins et al. demonstrate that it is possible and actually surprisingly successful (at least I was somewhat surprised that this worked so well) to replace purely correlative models by physiological models of the niche and fit those to distribution data. I thought that this is a very encouraging example of how models that include the abundant prior ecological/physiological information we have in their structure can still be fitted in very much the same way that we fit correlative models at the moment.
theoretical ecology 
Pingback: Special Issue: The ecological niche as a window to biodiversity « theoretical ecology