Habitat connectivity and fragmented nuthatch populations in agricultural landscapes

In agricultural landscapes, the habitat of many species is subject to fragmentation. When the habitat of a species is fragmented and the distances between patches of habitat are large relative to the movement distances of the species, it can be expected that the degree of habitat connectivity affects processes at population and individual level. In this thesis, I report on a study of effects of habitat fragmentation and opportunities to mitigate these effects by planning ecological networks. The objective of the study is to contribute to an improved knowledge about the effects of fragmentation and defragmentation of habitat on populations, in particular effects of differences in the degree of habitat connectivity on colonization and habitat selection. The main question of the research was: do networks of patches contribute to population sustainability of species in fragmented habitat?The development of planning for nature in the Netherlands is sketched in the second chapter. It is illustrated with the spatial concepts for the rural areas that landscape planning became landscape ecological based. After this chapter, I addressed three questions that were derived from the main question.The first question was: what variables can measure the degree of connectivity of habitat patches and are the differences in the degree of connectivity related to the colonization probability of patches? Therefore, habitat patches and the distances between these patches were modelled as networks. In landscapes with fragmented habitat for a certain species, these networks appear as so-called nonconnected networks consisting of disjointed subsets of patches. Between these subsets, exchange of individuals happens seldom of never.We derived parameters that measure the degree of connectivity of the patches in those networks. The parameters can deal with the size (the number of elements) and the spatial configuration of these subsets. One of the parameters was used to investigate the relationship between the degree of connectivity measured at different spatial scales and colonization of unoccupied patches by the nuthatch Sitta europaea in three regions in the Netherlands.To vary the spatial scale, I used threshold distances as maximum dispersal distances for which the degree of habitat connectivity was calculated. Habitat patches are assumed to be connected when the distances between the patches are less than this threshold distance. The degree of habitat connectivity measured for threshold distances of approximately 2.4 to 3 km best explains the differences in the colonization probability of unoccupied patches. These threshold distances give an indication of the distances covered by dispersing nuthatches that led to successful colonizations. Moreover, I could give an indication of the range of threshold distances where effects of constrained dispersal can be expected in the three regions.The second question was: is habitat selection limited in landscapes with fragmented habitat? Therefore, effects of the degree of habitat connectivity on the selection of territories were investigated. Based on a spatially explicit individual-based model, it could be hypothesized that habitat selection is limited when the degree of connectivity is low. This hypothesis was tested with empirical data of nuthatches in four regions in the Netherlands. One of the regions can be considered as the reference region with contiguous habitat where dispersal is not constrained. The habitat quality for nuthatches could be measured by the mean trunk diameter of oaks and beeches.We concluded that selection of territories is limited in fragmented habitat compared to selection in contiguous habitat. The quality of the occupied territories in fragmented habitat is lower than in contiguous habitat. This is especially the case when the population level is low. We showed that a lower average breeding success can be found in territories with low degree of connectivity. The results indicate the absence of a negative feedback between population level and the average breeding success in fragmented habitat, which contributes to the increased extinction probability of populations. Among other factors, limited habitat selection in fragmented habitat may thus result in a lower population density than in contiguous habitat.The degree of habitat connectivity can increase due to the allocation of new habitat. This may mitigate the effects of fragmentation. The third question was: how can networks of patches be optimally allocated in agricultural landscapes that both meets the requirements for population sustainability and takes into consideration the suitability of the land for competing land uses? We developed two spatial allocation models that plan new habitat considering ecological guidelines of minimum patch sizes and maximum threshold distances and the suitability of the land for competing land uses. The model MENTOR adds new patches that may act as "stepping stones" between reserve sites. The model ENLARGE enlarges existing sites. We showed that both the allocation of stepping stones and the enlargement of existing sites provide a higher percentage of occupied habitat. An interesting question for further research is under which conditions either the allocation of stepping stones or the enlargement of existing sites is preferred as strategy for conservation planning in human-dominated landscapes.The results of the research provide evidence that the degree of habitat connectivity determines both the colonization probability of unoccupied patches and the selection of habitat. They also give an indication at what spatial scale the degree of habitat connectivity affects these processes as observed for nuthatches. When through networks of patches the degree of habitat connectivity can be enhanced, positive effects on population sustainability can be expected. This thesis contributes to an improved problem detection of effects of habitat fragmentation and explores opportunities for defragmentation of habitat and optimization of land use allocation in human-dominated landscapes. With the knowledge about the effects of fragmentation and defragmentation, this study may be a step forward to enhance and preserve biodiversity.