Population dynamics in variable environments – impacts of noise colour and synchrony

• Main Author: Lögdberg, Frida
• Format: Doctoral Thesis
• Language: English
• Published: Linköpings universitet, Teoretisk Biologi 2012
• Subjects: Natural Sciences; Naturvetenskap
• Online Access: http://urn.kb.se/resolve?urn=urn:nbn:se:liu:diva-72951; http://nbn-resolving.de/urn:isbn:9789173930017

Environmental variation is an essential part of population dynamics, and two characteristics of such variation—variance and the temporal autocorrelation termed ‘noise colour’—are essential for determining the persistence of a population. In addition, the spatial correlation of local environmental variation between habitat patches (i.e., synchrony) is equally important in subdivided populations connected via dispersal. The research underlying this thesis explored the effects of noise colour and synchrony on population dynamics. The dynamics were studied primarily in single-species models with fast or slow population responses to environmental changes, and several-species systems (i.e., food webs) with different stability properties were also considered. Populations were spatially subdivided with local dynamics in discrete patches, and patch positions were modelled either implicitly or explicitly, with different landscape configurations in the latter case. It has previously been shown that the effect of increased environmental redness on extinction risk in nonspatial models depends on population responsiveness, seen as increased and decreased risks for fast and slow responding populations, respectively. Here, increased redness of noise decreased the extinction risk for fast-responding populations (in accordance with non-spatial studies) in a simple implicit landscape model (Papers I and II). Slow-responding populations in some cases showed a raised extinction risk for intermediate noise colour values (Paper I), which does not agree with earlier results. However, increasing the spatial complexity evened out the differences that were caused by responsiveness (Papers III and IV). Thus, in general, the explicit landscape models displayed a decrease in extinction risk with increasing environmental redness regardless of whether the populations were fast or slow in responding to environmental variation. Still, fast and slow responsiveness of populations differed in relation to the following: overall levels of extinction risk (Papers I, III, and IV), synchrony of population variations (Paper II), colour of population variations (Paper II), and response to landscape structure (Papers III and IV). For fast-responding populations, the degree of synchrony of population variations was similar to the synchrony of environmental noise (Paper II). Local populations of a model organism that responded slowly to environmental variation were more synchronized than the environmental variation itself, and the largest shift between the environment and the populations was seen for intermediate red noise colours (Paper II). This indicated that dispersal-induced population synchrony could be enhanced by reddened noise. Landscape configuration proved to be important for the general levels of extinction risk. This effect was most pronounced for fast-responding populations (Papers III and IV) and became even more distinct when distance-dependent synchrony was added between the environmental variations (Paper IV). Adding explicit landscapes led to an decrease in the differences between fast- and slow-responding populations, when considering the influence of noise colour on extinction risk. Also, landscape configuration affected the importance of degree of synchrony through its impact on distances between patches, which resulted in configurations where extinction risk depended solely on noise colour. The effects on stability exerted by populations embedded in food webs were investigated in an implicit landscape model (Paper V). Three types of food webs with different properties of inherent stability all showed a decrease in stability at increased environmental variance and increased redness of environmental variation. In conclusion, the single-species models showed that the survival conditions of populations that were near extinction were improved by all of the following: decreased synchrony, reddening of noise, and aggregation of patches. The results of the web simulations indicate that we need better understanding of how findings obtained using single-species models can be used to reveal the effects of noise colour on species communities. From a management perspective, altering landscape structure may compensate for increased extinction risks caused by changed noise colour of environmental variation, which is a predicted outcome of climate change.