A provocative headline is a double-edged sword. Why do I indicate that one of the most rapidly developing fields of animal ecology should still be regarded as a soft science? When it comes to individual space use rest assured that I’m thrilled by the substantial leaps forward in some parts of the theory of animal whereabouts. On the other hand, I also have critical comments. In my view there is still too strong disconnection between some general properties of movement-related animal behaviour and theoretical representations of this behaviour in models.
Both in my book and in previous blog posts I have repeatedly pointed out the unfortunate fact that contemporary models in the field often referred to as “movement ecology” have matured into two quite distinct premise foundations. On one hand we see a broadened recognition of scale-free movement as a quite general property and on the other hand also a broadened recognition that spatial memory is influencing movement under common ecological conditions. The former insight is often referred to a “Lévy walk-like movement”, while the latter focus regards strong and recent developments of the classic theory related to home range behaviour, philopatric dispersal, and similar memory map-dependent properties. Unfortunately, these two directions of theoretical development – scaling and memory aspects – still tend to be progressing quite independently of each-other. Lévy walk is inherently void of spatial memory influence, and site fidelity models are generally building on a scale-specific framework with spatial memory influence as a model extension of this framework. Hence, a scientific field consisting of disparate theories when it comes to the other camp’s main system premises has to fail. Thus, I dare to call animal movement theory unnecessarily soft – until a unification is reached.
To summarize, from a bird’s view of the current state I dare to categorize the science of animal movement as soft science due to lack of coherence between on one hand modelling the sliding scale from scale-specific to scale-free space use, and on the other hand modelling the variable strength of directed returns to familiar patches – on the basis of a common theoretical framework.
When I during the early 90’s started locking my focus on these two aspects of animal behaviour the interplay between multi-scaled space use and the simultaneous expression of site fidelity, both fields of research were in respective camps surrounded by confusion and controversies. Conferences, workshops and books on animal space use were typically surrounded by a myriad of ideas and concepts in the emergent field of landscape ecology. Thus, as far as I recall the scene back then, nobody else seemed to be tempted to make the science of animal space use even messier by studying model coherence between scaling and spatial memory. The general attitude was that “scale-free movement” – if it was at all recognized – might be interesting as a concept, but the theory had a far way to go before it could reach general acceptance with consequences for statistical and dynamical modelling and ecological analysis.On the other hand, “the Lévy camp” typically disregarded the site fidelity aspect of movement all together, except for occasional and brief reference to the concept as a challenge for the future.
Progress in respective camps have obviously been strongly propelled by large databases of animal movement that has been collected from modern GPS technology. Thus, theory and empirical data have been brought closer together. Wildlife ecology, computer modelling and advanced statistical analysis is thriving together in many strong research groups. Still, I’m waiting for bolder steps towards stronger unification between scaling and memory.
Wildlife ecologists have much to look forward to from such a leap towards better model realism. Consider all the interesting aspects that could be more easily studied and tested when the respective hypotheses are based on a more coherent theory of animal space use. Respective quantification of parameters connected to scaling and memory will then be founded on models with stronger predictive power. It is tempting lo start listing a long string of examples of potential theory applications at this point, but first things first. Conferences and workshops on the scaling/memory issue lay in the cards!