As evident from the title of one of the most commonly used ecology textbooks over the last four decades (Krebs 2009; first edition from 1972), ecology is often defined as the study of the distribution and abundance of species. Numerous studies have been undertaken on this topic, and the overwhelming majority report positive relationships between species distribution and abundance. A plethora of models have been developed aimed at understanding the underlying mechanisms behind this relationship (Borregaard & Rahbek 2010). Running the risk of simplifying this undoubtedly complex issue, there are two main model frameworks that have dominated the field. One view is that the underlying mechanism reflects the resource use (‘niche’) of species and the actual distribution of resources (Brown 1984). Generalists, or species encountering abundant resources, are both locally abundant and widespread. Another view, not necessarily contradicting resource-based models, is that the underlying mechanism is metapopulation dynamics, i.e. processes of colonization and extinction in local populations connected by dispersal (Hanski et al. 1993). Locally abundant species are likely to be successful in colonizing new sites, and species that occupy many sites are likely to be successful in ‘rescuing’ local populations from extinction. Apart from the many different models, there are several methodological obstacles complicating this research topic. Both distribution and abundance may be defined and measured in different ways, and at different spatial scales, influencing how patterns are manifested and understood (e.g. Gaston 2003; Borregaard & Rahbek 2010). Making novel contributions to the study of distribution and abundance is therefore indeed difficult.
A generally overlooked aspect, however, and one that may introduce a new angle to this topic, is how abundance–distribution relationships develop over time. A focus on the temporal component of these relationships enables a coupling to processes in plant communities such as disturbance and succession, and by incorporating a time dimension it creates a direct link to population processes. This is the background context of a study by Guedo & Lamb (2013) in the current issue of Journal of Vegetation Science. Based on a long-term (35-yr) data set from two different communities, open grassland and a grassland-forest ecotone, in a Fescue Prairie in Canada, subjected to different burning management treatments (Fig. 1), Guedo & Lamb were able to reveal how abundance–distribution relationships develop over time. By distinguishing how individual species contributed to these relationships, they open a range of research questions for future studies.
The major conclusion from Guedo & Lamb is that abundance–occupancy relationships vary over time. This is interesting in itself, but as it adds just another dimension of complexity to the topic of distribution–abundance relationships, it may at the same time be discouraging for those who strive for clarity of understanding plant community pattern and process. Is this just another case study showing the immense variation in nature? The results from Guedo & Lamb become more revealing when looking at the details behind the broad picture they describe. They focused on one particular aspect of distribution, namely occupancy, measured as the fraction of large plots (hereafter called ‘sites’) where a species was found. Abundance was measured in two different ways. The measure that gave the clearest results (and which also Guedo & Lamb argued is the most relevant measure) was frequency, which was measured as the fraction of smaller quadrats located within each site that was occupied by a certain species. Thus, frequency can be seen as similar to occupancy, but at a smaller spatial scale.
The results showed that frequency of many species increased over time relative to occupancy, most likely as a result of successional processes following burning treatments. This may be interpreted as a build-up of local populations within a larger-scale framework of an unchanging occupancy. As the number of sites was constrained to eight (Guedo & Lamb, their Fig. 2), this may reflect the fact that colonization outside the studied area was not accounted for. Disregarding this limitation of their study, what is undisputable is that different groups of species contributed differently to the changing abundance–occupancy relationship. In grassland sites, small-statured graminoids and forbs rapidly increased in frequency after burning. A similar, but slower, response was found for species with a more erect stature. In grassland-forest ecotone sites, a similar response was observed among small-statured species, but this response was more transient, since the species dominating before burning, trembling aspen (Populus tremuloides), soon regained its dominance.
The study by Guedo & Lamb informs us that the response of individual species drives changing abundance–occupancy relationship during succession after disturbance. These responses are not idiosyncratic, but related to functional traits of the species. Guedo & Lamb focused on growth form, but a promising area for future studies is to consider also traits related to the regenerative life-cycle phase. Furthermore, species-level responses occur within a time-window that may vary in length, as illustrated by the difference between the grassland and the grassland-forest ecotone sites.
Ever since Lawton (1999) described community ecology as ‘a mess’, there have been attempts to synthesize and conceptually unify this field (e.g. Leibold et al. 2004; Vellend 2010). The results from Guedo & Lamb indicate one avenue for such unification. A community-level pattern such as a positive inter-specific abundance–distribution relationship is ultimately based on the response of individual species to their environment. As species vary, a variation in the shape of these relationships between communities and over time, as shown by Guedo & Lamb, should come as no surprise. However, their study provides some important questions for further analyses of how species response to environmental gradients relates to functional traits (e.g. Kleyer et al. 2012), and how this response is manifested at different spatial scales. The study by Guedo & Lamb also illustrates the value of using time-series longer than the usual few-year studies constrained by ordinary research grants. The next productive step will thus be to use the insights from long-term descriptive studies, such as the one by Guedo & Lamb, to examine species-level mechanisms behind abundance–occupancy relationships.