Spread of introduced species in heterogeneous landscapes: effects of dispersal, habitat quality, and fragmentation

Understanding the mechanisms fostering the spread of organisms, particularly invasive species and disease agents, is a topic of major concern within ecology.  The dynamics of their spread are influenced by their dispersal ability, as well as characteristics of the landscape that they move across, such as its quality and heterogeneity (which can in turn influence dispersal).  Since humans are facilitating the introduction of new species to areas outside their original home ranges, while concurrently increasing landscape fragmentation and influencing the quality of the remaining habitat, understanding how these factors interact to affect the persistence and spread of these potentially invasive species is important from both a conservation and theoretical perspective.

To study how heterogeneity in habitat quality and habitat fragmentation across a landscape influence the spread of introduced organisms and how availability of suitable habitat influences invasion success, Dewhirst and Lutscher (2009) used integrodifference models to simulate movement of an introduced species that exhibits separate growth and dispersal phases (which is a common life cycle strategy in many organisms, including plants).  Dispersal was described by a dispersal kernel, where the probability of moving from one point to another was dependent only on distance between the two points, but the average dispersal distance depended on the initial location of the dispersing individual, as well as the variance of the dispersal kernel. Dispersal ability with respect to fragmentation was altered by changing the variance of the dispersal kernel.  Habitat quality at a position was modeled by changing the intrinsic rate of population increase (r) associated with that location, with patches represented either high or low quality habitat.

The authors found that when there was small variance of the dispersal kernel, the population could persist, because individuals from good patches would mostly stay in the good patches where they could exhibit positive population growth.  When the individuals could disperse very far, then their probability of landing in a good habitat was proportional to the total amount good habitat in the landscape. When the authors use these models to analytically solve for the minimum proportion of ‘good’ habitat necessary for the successful invasion of the species (persistence & spread), they found that it was a saturating function related to the variance of the dispersal kernel, and the type of dispersal kernel also strongly influenced the shape of that relationship.  Thus, invasion threshold increased with increasing dispersal (or fragmentation) ability.  When they incorporated an Allee effect into the model, they found the opposite pattern, where the minimum proportion of good that was habitat that was necessary for successful invasion decreased with variance (i.e., dispersal ability or habitat fragmentation).  This can be explained by the fact that as fragmentation increases, distances between good locations decrease, thereby allowing enough individuals to move to a new good patch to lead to a positive Allee effect, thus facilitating persistence in that new patch.  Additionally, for the exponential square-root kernel (tails of dispersal kernel are not exponentially bounded), then invasions can accelerate.  This result is similar to previous results found on homogeneous landscapes (Kot et al. 1996), and thus habitat fragmentation cannot stop species invasions.  Overall, results indicate that different management strategies should be employed to stop invasions depending on whether an Allee effect is known to be present in the population dynamics of the introduced species.  If there is no Allee effect, then fine-scale fragmentation of suitable habitat could reduce its spread, while if there is an Allee effect, then large patches of unsuitable habitat would be required.

Dewhirst, S. & Lutscher, F. Dispersal in heterogeneous habitats: thresholds, spatial scales, and approximate rates of spread. Ecology (2009).