“An empirical and mechanistic explanation of abundance-occupancy relationships for a critically endangered nomadic migrant”

Posted on

Matthew H. Webb, Robert Heinsohn, William J. Sutherland, Dejan Stojanovic, Aleks Terauds (Jan 2019)

The DOI will be https://dx.doi.org/10.1086/700595

Abundance-occupancy relationships for a nomadic migrant are explained by variations in food abundance

Don’t assume high abundances of animals equates to high-quality habitat

A swift parrot (Lathamus discolor) in a tree hollow.
(Credit: Henry Cook)

A positive abundance–occupancy relationship (AOR) is a pervasive pattern in macroecology implying species with larger geographic distributions are generally more locally abundant. But for species tracking variable resources in space and time, this is not always the case. This study by the Difficult Bird Research Group provides a rare empirical example of a negative AOR for the nomadic critically endangered swift parrot caused by dynamic pulses in the availability and location of food over seven years (2009–2015). In their breeding range (Tasmania, Australia), local densities of birds increased as food availability decreased, and vice versa.

Although this study was conducted on a highly mobile species over a large geographic area, the results are likely relevant to partial migrants and less mobile species when their resources vary over smaller scales. This highlights the need to carefully consider the appropriate scales of sampling for AOR studies. Similarly, the association between the probability of occurrence and abundance is usually assumed to be positive and constant for species distribution models (SDMs). For the swift parrot, this relationship was positive but non-linear and varied with scale and between years due to differing degrees of spatial aggregation caused by changing food availability. Importantly, the results show that high abundance (or occupancy) does not necessarily equate to high quality habitat.

Nearly fledged swift parrot (Lathamus discolor) chicks.
(Credit: Henry Cook)

Contrasting these two areas of research (i.e. SDMs and AORs) allows better identification of fluctuations in carrying capacity, priority sites, resource bottlenecks, and interpretation of dynamic SDMs. Further, understanding the causal mechanisms of AORs and how they change over time may provide an empirical means to understand changes in population size and range dynamics in variable environments.


The positive abundance-occupancy relationship (AOR) is a pervasive pattern in macroecology. Similarly, the association between occupancy (or probability of occurrence) and abundance is also usually assumed to be positive and in most cases constant. Examples of AORs for nomadic species with variable distributions are extremely rare. Here we examined temporal and spatial trends in the AOR over seven years for a critically endangered nomadic migrant which relies on dynamic pulses in food availability to breed. We predicted a negative temporal relationship, where local mean abundances increase when the number of occupied sites decreases, and a positive relationship between local abundances and the probability of occurrence. We also predicted that these patterns are largely attributable to spatiotemporal variation in food abundance. The temporal AOR was significantly negative and annual food availability was significantly positively correlated with the number of occupied sites, but negatively correlated with abundance. Thus, as food availability decreased, local densities of birds increased, and vice-versa. The abundance – probability of occurrence relationship was positive and non-linear, but varied between years due to differing degrees of spatial aggregation caused by changing food availability. Importantly, high abundance (or occupancy) did not necessarily equate to high quality habitat and may be indicative of resource bottlenecks or exposure to other processes affecting vital rates. Our results provide a rare empirical example that highlights the complexity of AORs for species that target aggregated food resources in dynamic environments.