A fly's-eye view of landscape heterogeneity

One of the first papers that came to mind when I made a preliminary list of my favourite papers of all time was this one:

Haslett, J. R. (2001). Biodiversity and conservation of Diptera in heterogeneous land mosaics: A fly's eye view. Journal of insect conservation, 5(2), 71-75.

I read this at some point during my PhD and I loved it. Nevertheless I had no idea what to use it for, and as far as I recall I have not yet found an opportunity to cite it (although as I write this I think I do know where to cite it now!). The primary and most fundamental thing that I like about this paper is that it takes seriously the realities of other species and asks what landscapes would look like if we understood them from the perspectives of the species in those landscapes, not just from our own positions as humans using research methods to detect and measure things. On the one hand, it makes us ask "what is the world like for a fly?" This is a question about Umwelt, but also, as this paper develops, a question about scale, and about fractal intermingling of structures. Haslett focuses on syrphid flies, which are pollinators and thus of obvious interest to ecologists interested in pollination processes and how they are structured across landscapes, in turn affecting plant reproduction at landscape scales. Ecologists almost always consider "landscape" to be an inherent scale. Haslett argues that from the perspectives of flies, the size and structure of whatever a landscape is, is different from human ideas in important ways. First of all, flies are very small, compared to humans. Flies also fly, giving them a particular atmospherically-mediated experience of mobility and connectedness.

This paper tries to show that this perspectival shift to the fly's eye view is not just an exercise in philosophical questioning, or merely the remit of ethological studies of behaviour: it tells us concrete things about how ecologies are patterned. Syrphids showed a clear preference for spending time in areas with tall grass rather than grazed meadows. For them, the boundaries between areas of different grass height were perceived as important landscape edges. Although Haslett's study takes place in the Alps on what appear to be fixed pastures, one can easily imagine that the spatial conformation and extent of this landscape distinction must constantly change for syrphid flies in areas where grazers are managed extensively, or are wild. It must also change across seasons, and depend on the mix of perennial and annual meadow plants.

The 2001 paper is partly a review of previous work, including:

Haslett, J. R. (1997). Insect communities and the spatial complexity of mountain habitats. Global Ecology and Biogeography Letters, 49-56.

This paper in particular shows that a variable affecting fly habitat preferences over a landscape is the complexity of the borders of the habitat type. Areas with higher mixing and interdigitation of habitats seem to have higher fly biodiversity.

One reason that it is difficult to cite these two papers is that they are written in a quite different style to how we write ecology papers today, and they seem a little bit hard to "operationalise". Reading the 1997 paper, for example, I am struck that there is a lot of explanation and development of the ideas supporting the analysis, and that the statistical treatment is not as complex or magisterial as we would expect from a "good" paper today. I don't think I would be allowed to include so much explanation and idea development in a paper today. How many times have I been told by reviewers that there should be no argumentation or explanation in the Results section? The results sections in both of these papers are 99% explanation and idea development. In the best possible world, one could take important proposals to think differently, like these two papers, and translate them into a large series of tests and analyses that would satisfy and take advantage of the power of quantitative and experimental thinking.

Thinking from a fly's perspective shows us that landscapes are fractal, as Haslett refers to in 2001 and developed more in his earlier paper:

Haslett, J. R. (1994). Community structure and the fractal dimensions of mountain habitats. Journal of theoretical biology, 167(4), 407-411.

For me, appreciating that the landscape must thus be multiply and simultaneously fractal at different kinds of scales depending on the species perspective in question, has a important intellectual payoff. If there is no inherent scalar nature to the physical structuring of ecosystems, as we implicitly assume most of the time, then our theories of ecological functioning and change (e.g. succession, disturbance), need to recognise this.

Some ecological theories, like landscape ecology or the panarchy model of socio-ecological systems, are all about links across scales. For these theories, we need to ask, have we got the scales, the heterogeneity, and the forms of connection right for all species, or are we just describing what humans perceive? Other ecological theories, like the intermediate disturbance theory, are essentially scale-free. Although the IDH suggests that big and small disturbances have different effects on biodiversity, no scale at which big and small is defined has ever been specified. This is one of the reasons that it is highly ambiguous as a theory and difficult either to falsify, or to support in comparative studies. How would we rethink the IDH, which contains valid empirical observations and so does not deserve to be totally trashed, not as scale-free or scale-ambiguous, but as occurring at multiple, overlapping fractal scales simultaneously?

We might also ask "how big is degradation?" While degradation is very poorly integrated (if at all) with the rest of ecological theory, it is often defined as a reduction in biodiversity or productivity at decadal scales. No mention is ever made about the spatial scale at which it occurs. Just as people have asked if it makes sense to talk about rewilding your garden, we might ask if it makes sense to talk about a square centimetre as being degraded. Can a square centimetre inside a degraded human landscape be not degraded? Can a square centimetre within a non-degraded habitat be degraded? We might in fact understand this as a question of "degraded for whom? For flies? For moss? For humans?" Asking the question this way forces us to ask how degradation would be perceived and experienced, and thus, what exactly it consists of. Through what structures and affordances, and via what aspects of their Umwelts, would a fly or a moss experience a loss of biodiversity or productivity at decadal scales?

On another note, I love flies. Did you know that flies sleep standing upside down on the stems of grasses and other plants?

A fly on a moor in the UK. (c) MR-B