Bridging gaps in fragmented marshland : applying landscape ecology for bird conservation
An important part of the natural values in The Netherlands is based on the fact that the country has a unique geographical position in temperate lowland Europe at the mouth of the rivers Rhine and Meuse. This creates a number of interesting gradient situations e.g. between saltwater and freshwater systems, between eutrophic and oligotrophic systems, and between tidal, streaming and stagnant waters. This position offers excellent conditions for a wide variety of wetland systems: river and clay marshlands with open water and macrophyte vegetation like reedlands, peat marshland with bogs, fens and mires, as well as estuaries with saltmarshes. This marshland diversity resulted in a very diverse avifauna with many species occurring in high densities. The diversity and quantity of these wetland ecosystems, however, is severely threatened and consequently the number and the distribution of typical marshland bird species decreased (Den Boer 2000). Acidification, euthrophication and desiccation, cultivation and unfavourable management practices all contributed to a decline of total area and an increasing degree of fragmentation of the remaining habitat for marshland birds (Figure 1.1). The Dutch Nature Policy Plan, published in 1990 (NPP 1990), mentioned fragmentation as one of the most important threats to biodiversity. A nation-wide ecological network of nature areas was proposed as the solution for this problem. It is vital for the success of this strategy that the implementation is underpinned by ecological knowledge of the underlying processes in this network system.The process of habitat fragmentation leads to landscapes with dispersed small populations within an inhospitable matrix. Small populations are likely to go extinct by stochastic demographic fluctuations. Consequently, the viability of these small populations depends on the likelihood that they will be recolonised by individuals from elsewhere. Crucial is whether the distances most individuals are likely to cover between years, as a result of the dispersal process, are large enough compared to the interpatch distances within a landscape. A set of subpopulations (or metapopulation), may be viable, even when all subpopulations are small in size under the condition that local extinction and recolonisation rates are balanced.Fragmentation can lead to a conservation problem, which asks for spatial solutions. In a multifunctional society, like The Netherlands, finding and implementing effective solutions are part of a spatial planning process. The way from problem definition to the actual implementation of a plan can be regarded as a cyclic planning process (Figure 1.2) with successive phases on problem detection, exploring solutions, development of landscape scenarios, designing an actual plan and plan evaluation. Often, ecological knowledge is poorly used in such planning processes. It is probably due to the fact that the knowledge is not tailored for the different phases of the planning cycle. This asks for generalisations, from case studies to a variety of landscapes. It also asks for aggregation of knowledge on single species to knowledge on multispecies level.It is a challenge for landscape ecology to develop and support such an approach and link ecology and spatial planning and, to my opinion, knowledge systems should play a major role (Figure 1.3).This thesis is an attempt to elaborate this landscape ecological line of thought for the problem of fragmented marshland and conservation of bird species. It tries to bridge two 'gaps'. Firstly, the gap (in literal sense) between remaining pieces of marshland. What is an effective spatial strategy for the persistence of marshland birds in The Netherlands? How should we 'bridge' the gaps between our remaining habitats? The second gap is metaphorical and refers to the transfer of ecological knowledge gathered in case studies and at the species level into tools and instruments for application in nature management and policy. In conclusion, the central questions of this thesis are: (1) under what spatial conditions do marshland birds demonstrate negative effects of fragmentation and (2) how to utilise ecological knowledge for practical tools in conservation?In Chapter 2 distribution data of six typical marshland passerines occurring in heavily fragmented landscapes in Gelderland and Sealand Flanders are analysed. Some of the species are common, like the reed warbler and the reed bunting (occurring in >50% of the habitat patches), others are scarce or rare (sedge warbler, great reed warbler, penduline tit and bluethroat). After correcting for patch size and quality, in most cases occurrence is significantly explained by patch connectivity, a measure for the spatial configuration of the patch. The response of species to fragmentation is variable, but the analyses confirm that in many landscapes in The Netherlands habitat fragmentation negatively influences occurrence of bird populations. By linking the fraction of occupied patches to ecologically scaled landscape indices (ESLI) a diagnostic tool was set up. The ESLI 'average patch carrying capacity' is a good estimator of the fraction of occupied patches and that the 50% occupation threshold (assumed to be a viability threshold according to Vos et al. 2001) will be reached at an average patch carrying capacity of 2-5 territories. This diagnostic instrument is practical and quickly applicable because no field data are required, the essential information concerning carrying capacity can be collected from literature.Spatially explicit population models, mathematical models simulating population dynamics of a species with subdivided populations, are useful research tools to explore and understand the behaviour of populations in a fragmented landscape. These models require lots of information for parameterisation of the essential parameters, the most important source are field data. Particularly quantitative data on the dispersal and emigration rates are essential, though collecting these data is time-consuming and difficult. In Chapter 3 the results of a field study on the great reed warbler are presented. The study area consisted of reed beds along lakes, total length around 15 kilometres. During 1994-2000 over 1100 individuals were colour banded and more than 250 individuals were resighted in one or more following years. By analysing records of marked individuals, survival, dispersal and emigration rates were quantified. Most individuals show displacements within the study area. Survival is estimated by analysing the capture-recapture data using the program MARK and depends on age, sex and year. Emigration rates are estimated by dividing the study area in six equal parts and analysing the differences in local survival between the parts and various combinations. The resulting function describes the relation between the emigration rate and the size of an area. The demographic data and the quantification of dispersal are subsequently used for the parameterisation of a metapopulation model. This model is decribed and used in the following chapters.Chapter 4 demonstrates the influence of large environmental impacts on the dynamics of a metapopulation. In the last decades a number of passerines wintering in the Sahel zone suffers from the droughts. The sedge warbler is one of those species. After drought years (like in the mid-seventies and eighties) the population shows a decline of more than 50%. After these collapses the population recovers, but not in all marshlands. An analysis of the population trends in a large number of marshlands in The Netherlands indicated that populations in large marshlands recovered quickly, but that small and relatively isolated populations, like in the eastern and southern parts of the country, did not recover at all. Exercises with a metapopulation model support the hypothesis that this lack of resilience is due to fragmentation effects. Populations of fragmented areas are particularly vulnerable for catastrophic events, probably because certain thresholds for population viability are linked with population size.Chapter 5 elaborates on the role of low quality habitats (sinks) in a metapopulation. Persistence of sink populations depends on nearby source populations. However, do sink areas also contribute to the persistence of source areas? In a riverine landscape in the province of Gelderland a number of large marshlands are located close to each other, these constitute source areas for the reed warbler. In a radius of 20-30 kilometres around these marshlands hundreds of small marshlands occur in a landscape dominated by agriculture. A field study demonstrated that in many of these small reed elements reed warblers occur. A regression analysis shows that the habitat quality of these habitat patches is poor and that they most likely are to be considered sink areas. Furthermore, spatial parameters are very important in explaining the probability of occurrence, the abundance and also extinction and recolonisation rates of the habitat patches. In isolated patches the probability of occurrence is low, densities are low, the probability of extinction is high and recolonisation probabilities are low. Thus reed warbler populations in these small elements show metapopulation dynamics. Using a spatially explicit population model proves that sink areas do contribute to the stability and persistence of source areas. A recovery after a population decline is much faster in source areas embedded in a network with sink patches than in source areas with no sinks around. This influence however, depends on the spatial conditions. For reed warblers mainly the sink areas within 2-5 kilometres of a source patch are of importance, and the total number of individuals supported by the sinks should be at least 25% of the source population size.In Chapter 6 I analyse distribution data of the bittern in The Netherlands. The bittern is a scarce breeding bird, occurring in less than 50% of all suitable reed dominated marshlands. The species is rare in the eastern, north-eastern and southern parts of the country, although suitable habitat patches occur. A regression model is used to predict probability of occurrence of the bittern in a patch given the size, abiotic (soil) conditions of the marshland and the connectivity. The results indicate that low probability of occurrence of the bittern in certain regions can be explained by unfavourable spatial conditions. The predictions per habitat patch are subsequently used as a calibration set for a simple expert model, SCAN, based on connectivity. The model uses a connectivity value for grid cells of 250x250 meters as an indicator of spatial conditions. The SCAN output gives satisfactory results and correlates well with the regression model predictions. It is considered a useful instrument for problem detection, although the challenge is to find generic rules to translate the model output into probability of occurrence or persistence measures.In Chapters 2-6 a number of methods is introduced for the analysis of species occurrence and persistence in fragmented landscapes: developing landscape indices (Chapter 2), metapopulation models (Chapter 4 and 5) and regression models (Chapter 2 and 6). Can we develop an instrument that integrates the various methods and is applicable on landscape rather than on species level? In Chapter 7 a rule-based system is described that combines the merits of all previously mentioned methods. Central issue is the key-patch approach. A key patch is defined as a habitat patch within a habitat network of such a size that the probability of extinction of the local population is less than 5% in 100 years. It is the basis of a persistent metapopulation. The predictive models resulting from the various regression analyses are used to indicate key population standards for a number of species. With help of metapopulation models for two ecoprofiles, a marshland passerine 'reed warbler' and a marshland heron 'bittern' these key population standards are extrapolated to landscape configurations. Literature data on various species groups are used to test the validity of the standards. Three standards are proposed, for long-lived large vertebrates (example: bittern, otter): 20 individuals, for middle-long lived, medium sized vertebrates (example: great reed warbler): 40 individuals and for short-lived, small vertebrates (reed warbler, voles): 100 individuals. These standards are implemented in a GIS and rule-based system, called LARCH, and this is considered a useful evaluation tool for spatial scenarios and plans.In Chapter 8 I conclude that the occurrence and persistence of marshland bird species in The Netherlands is negatively influenced by the current degree of habitat fragmentation. I also discuss the value of the presented instruments and tools for application in the different phases of the planning cycle.What spatial strategy is effective to bridge the gaps between isolated marshlands? The results of this thesis indicate that the primary option is to enlarge existing marshlands, creating at least five key populations for most of the marshland birds. This requires extension of the size of most existing marshlands by nature restoration, aiming at totals of 5000-10 000 ha. Next, a marshland 'backbone' should be created with restoration of medium-sized marshlands along several axes. By this way marshland bird populations in the periphery are better 'connected' with the core areas and will show a higher, regional, persistence. It will also enhance the probability of occurrence and saturation in existing marshland areas and thus is a cost-effective measure.
| Main Author: | |
|---|---|
| Other Authors: | |
| Format: | Doctoral thesis biblioteca |
| Language: | English |
| Subjects: | birds, dispersal, fragmentation, marshes, netherlands, populations, terrestrial ecosystems, wetlands, fragmentatie, landschapsecologie, moerassen, moerasvogels, natuur, nederland, populaties, terrestrische ecosystemen, versnippering, verspreiding, vogels, |
| Online Access: | https://research.wur.nl/en/publications/bridging-gaps-in-fragmented-marshland-applying-landscape-ecology- |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| Summary: | An important part of the natural values in The Netherlands is based on the fact that the country has a unique geographical position in temperate lowland Europe at the mouth of the rivers Rhine and Meuse. This creates a number of interesting gradient situations e.g. between saltwater and freshwater systems, between eutrophic and oligotrophic systems, and between tidal, streaming and stagnant waters. This position offers excellent conditions for a wide variety of wetland systems: river and clay marshlands with open water and macrophyte vegetation like reedlands, peat marshland with bogs, fens and mires, as well as estuaries with saltmarshes. This marshland diversity resulted in a very diverse avifauna with many species occurring in high densities. The diversity and quantity of these wetland ecosystems, however, is severely threatened and consequently the number and the distribution of typical marshland bird species decreased (Den Boer 2000). Acidification, euthrophication and desiccation, cultivation and unfavourable management practices all contributed to a decline of total area and an increasing degree of fragmentation of the remaining habitat for marshland birds (Figure 1.1). The Dutch Nature Policy Plan, published in 1990 (NPP 1990), mentioned fragmentation as one of the most important threats to biodiversity. A nation-wide ecological network of nature areas was proposed as the solution for this problem. It is vital for the success of this strategy that the implementation is underpinned by ecological knowledge of the underlying processes in this network system.The process of habitat fragmentation leads to landscapes with dispersed small populations within an inhospitable matrix. Small populations are likely to go extinct by stochastic demographic fluctuations. Consequently, the viability of these small populations depends on the likelihood that they will be recolonised by individuals from elsewhere. Crucial is whether the distances most individuals are likely to cover between years, as a result of the dispersal process, are large enough compared to the interpatch distances within a landscape. A set of subpopulations (or metapopulation), may be viable, even when all subpopulations are small in size under the condition that local extinction and recolonisation rates are balanced.Fragmentation can lead to a conservation problem, which asks for spatial solutions. In a multifunctional society, like The Netherlands, finding and implementing effective solutions are part of a spatial planning process. The way from problem definition to the actual implementation of a plan can be regarded as a cyclic planning process (Figure 1.2) with successive phases on problem detection, exploring solutions, development of landscape scenarios, designing an actual plan and plan evaluation. Often, ecological knowledge is poorly used in such planning processes. It is probably due to the fact that the knowledge is not tailored for the different phases of the planning cycle. This asks for generalisations, from case studies to a variety of landscapes. It also asks for aggregation of knowledge on single species to knowledge on multispecies level.It is a challenge for landscape ecology to develop and support such an approach and link ecology and spatial planning and, to my opinion, knowledge systems should play a major role (Figure 1.3).This thesis is an attempt to elaborate this landscape ecological line of thought for the problem of fragmented marshland and conservation of bird species. It tries to bridge two 'gaps'. Firstly, the gap (in literal sense) between remaining pieces of marshland. What is an effective spatial strategy for the persistence of marshland birds in The Netherlands? How should we 'bridge' the gaps between our remaining habitats? The second gap is metaphorical and refers to the transfer of ecological knowledge gathered in case studies and at the species level into tools and instruments for application in nature management and policy. In conclusion, the central questions of this thesis are: (1) under what spatial conditions do marshland birds demonstrate negative effects of fragmentation and (2) how to utilise ecological knowledge for practical tools in conservation?In Chapter 2 distribution data of six typical marshland passerines occurring in heavily fragmented landscapes in Gelderland and Sealand Flanders are analysed. Some of the species are common, like the reed warbler and the reed bunting (occurring in >50% of the habitat patches), others are scarce or rare (sedge warbler, great reed warbler, penduline tit and bluethroat). After correcting for patch size and quality, in most cases occurrence is significantly explained by patch connectivity, a measure for the spatial configuration of the patch. The response of species to fragmentation is variable, but the analyses confirm that in many landscapes in The Netherlands habitat fragmentation negatively influences occurrence of bird populations. By linking the fraction of occupied patches to ecologically scaled landscape indices (ESLI) a diagnostic tool was set up. The ESLI 'average patch carrying capacity' is a good estimator of the fraction of occupied patches and that the 50% occupation threshold (assumed to be a viability threshold according to Vos et al. 2001) will be reached at an average patch carrying capacity of 2-5 territories. This diagnostic instrument is practical and quickly applicable because no field data are required, the essential information concerning carrying capacity can be collected from literature.Spatially explicit population models, mathematical models simulating population dynamics of a species with subdivided populations, are useful research tools to explore and understand the behaviour of populations in a fragmented landscape. These models require lots of information for parameterisation of the essential parameters, the most important source are field data. Particularly quantitative data on the dispersal and emigration rates are essential, though collecting these data is time-consuming and difficult. In Chapter 3 the results of a field study on the great reed warbler are presented. The study area consisted of reed beds along lakes, total length around 15 kilometres. During 1994-2000 over 1100 individuals were colour banded and more than 250 individuals were resighted in one or more following years. By analysing records of marked individuals, survival, dispersal and emigration rates were quantified. Most individuals show displacements within the study area. Survival is estimated by analysing the capture-recapture data using the program MARK and depends on age, sex and year. Emigration rates are estimated by dividing the study area in six equal parts and analysing the differences in local survival between the parts and various combinations. The resulting function describes the relation between the emigration rate and the size of an area. The demographic data and the quantification of dispersal are subsequently used for the parameterisation of a metapopulation model. This model is decribed and used in the following chapters.Chapter 4 demonstrates the influence of large environmental impacts on the dynamics of a metapopulation. In the last decades a number of passerines wintering in the Sahel zone suffers from the droughts. The sedge warbler is one of those species. After drought years (like in the mid-seventies and eighties) the population shows a decline of more than 50%. After these collapses the population recovers, but not in all marshlands. An analysis of the population trends in a large number of marshlands in The Netherlands indicated that populations in large marshlands recovered quickly, but that small and relatively isolated populations, like in the eastern and southern parts of the country, did not recover at all. Exercises with a metapopulation model support the hypothesis that this lack of resilience is due to fragmentation effects. Populations of fragmented areas are particularly vulnerable for catastrophic events, probably because certain thresholds for population viability are linked with population size.Chapter 5 elaborates on the role of low quality habitats (sinks) in a metapopulation. Persistence of sink populations depends on nearby source populations. However, do sink areas also contribute to the persistence of source areas? In a riverine landscape in the province of Gelderland a number of large marshlands are located close to each other, these constitute source areas for the reed warbler. In a radius of 20-30 kilometres around these marshlands hundreds of small marshlands occur in a landscape dominated by agriculture. A field study demonstrated that in many of these small reed elements reed warblers occur. A regression analysis shows that the habitat quality of these habitat patches is poor and that they most likely are to be considered sink areas. Furthermore, spatial parameters are very important in explaining the probability of occurrence, the abundance and also extinction and recolonisation rates of the habitat patches. In isolated patches the probability of occurrence is low, densities are low, the probability of extinction is high and recolonisation probabilities are low. Thus reed warbler populations in these small elements show metapopulation dynamics. Using a spatially explicit population model proves that sink areas do contribute to the stability and persistence of source areas. A recovery after a population decline is much faster in source areas embedded in a network with sink patches than in source areas with no sinks around. This influence however, depends on the spatial conditions. For reed warblers mainly the sink areas within 2-5 kilometres of a source patch are of importance, and the total number of individuals supported by the sinks should be at least 25% of the source population size.In Chapter 6 I analyse distribution data of the bittern in The Netherlands. The bittern is a scarce breeding bird, occurring in less than 50% of all suitable reed dominated marshlands. The species is rare in the eastern, north-eastern and southern parts of the country, although suitable habitat patches occur. A regression model is used to predict probability of occurrence of the bittern in a patch given the size, abiotic (soil) conditions of the marshland and the connectivity. The results indicate that low probability of occurrence of the bittern in certain regions can be explained by unfavourable spatial conditions. The predictions per habitat patch are subsequently used as a calibration set for a simple expert model, SCAN, based on connectivity. The model uses a connectivity value for grid cells of 250x250 meters as an indicator of spatial conditions. The SCAN output gives satisfactory results and correlates well with the regression model predictions. It is considered a useful instrument for problem detection, although the challenge is to find generic rules to translate the model output into probability of occurrence or persistence measures.In Chapters 2-6 a number of methods is introduced for the analysis of species occurrence and persistence in fragmented landscapes: developing landscape indices (Chapter 2), metapopulation models (Chapter 4 and 5) and regression models (Chapter 2 and 6). Can we develop an instrument that integrates the various methods and is applicable on landscape rather than on species level? In Chapter 7 a rule-based system is described that combines the merits of all previously mentioned methods. Central issue is the key-patch approach. A key patch is defined as a habitat patch within a habitat network of such a size that the probability of extinction of the local population is less than 5% in 100 years. It is the basis of a persistent metapopulation. The predictive models resulting from the various regression analyses are used to indicate key population standards for a number of species. With help of metapopulation models for two ecoprofiles, a marshland passerine 'reed warbler' and a marshland heron 'bittern' these key population standards are extrapolated to landscape configurations. Literature data on various species groups are used to test the validity of the standards. Three standards are proposed, for long-lived large vertebrates (example: bittern, otter): 20 individuals, for middle-long lived, medium sized vertebrates (example: great reed warbler): 40 individuals and for short-lived, small vertebrates (reed warbler, voles): 100 individuals. These standards are implemented in a GIS and rule-based system, called LARCH, and this is considered a useful evaluation tool for spatial scenarios and plans.In Chapter 8 I conclude that the occurrence and persistence of marshland bird species in The Netherlands is negatively influenced by the current degree of habitat fragmentation. I also discuss the value of the presented instruments and tools for application in the different phases of the planning cycle.What spatial strategy is effective to bridge the gaps between isolated marshlands? The results of this thesis indicate that the primary option is to enlarge existing marshlands, creating at least five key populations for most of the marshland birds. This requires extension of the size of most existing marshlands by nature restoration, aiming at totals of 5000-10 000 ha. Next, a marshland 'backbone' should be created with restoration of medium-sized marshlands along several axes. By this way marshland bird populations in the periphery are better 'connected' with the core areas and will show a higher, regional, persistence. It will also enhance the probability of occurrence and saturation in existing marshland areas and thus is a cost-effective measure. |
|---|