Conservation Biology: Predicting Birds' Responses to Forest Fragmentation

Center for Conservation Biology, Department of Biological Sciences, Stanford University, 371 Serra Mall, Stanford, CA 94305, USA.
Current Biology (Impact Factor: 9.57). 11/2007; 17(19):R838-40. DOI: 10.1016/j.cub.2007.07.037
Source: PubMed


Understanding species' ecological responses to habitat fragmentation is critical for biodiversity conservation, especially in tropical forests. A detailed recent study has shown that changes in the abundances of bird species following fragmentation may be dramatic and unpredictable.

Download full-text


Available from: Cagan H Sekercioglu, Oct 09, 2015
221 Reads
  • Source
    • "Also other forms of habitat alteration, e.g. fragmentation, intensification of farmland production or urban expansion pose a threat to seed disperser communities worldwide (Sekercioglu & Sodhi 2007; Tscharntke et al. 2008). Increasing land-use intensity, e.g. by enhanced use of pesticides and fertilizer, has led to the "
    [Show abstract] [Hide abstract]
    ABSTRACT: Seed dispersing animals, ranging from small insects to large mammals, provide a crucial service for a large number of plant species worldwide. However, a decline in dispersers due to direct and indirect threats leads to disruptions of seed dispersal processes. As disperser species are differently susceptible to these threats, consequences for ecosystems are hard to predict. Impacts range from hampered regeneration of plant species to shifts in communities and a decline in ecosystem function. Here, we review these threats as well as expected consequences for communities and for the entire ecosystem. We further introduce options to protect dispersers and consider future research directions.
    Basic and Applied Ecology 03/2012; 13(2):109-115. DOI:10.1016/j.baae.2012.02.006 · 1.94 Impact Factor
  • Source
    • "Forest fragmentation poses a substantial threat to global biodiversity and may cause cascading impacts on a wide range of ecosystem functions and services (Wu et al., 2003; Millennium Ecosystem Assessment, 2005). Fragmentation occurs when tracts of continuous forest are broken up into smaller pieces as a result of land-use change (Chalfoun et al., 2002; Franklin et al., 2002; Fahrig, 2003; Watson et al., 2004; Sekercioglu & Sodhi, 2007), creating new edges between forest and other vegetation types, disconnecting patches from adjacent, continuous habitat and reducing patch sizes (Collinge, 1996; Fahrig, 2003; Saura & Carballal, 2004), thereby disrupting the movement patterns of many organisms and isolating populations (Redpath, 1995). Isolated populations in fragments are more sensitive to stochastic events, which can lead to population decline or extinction (Driscoll & Weir, 2005; Arroyo-Rodríguez et al., 2007), and fragmentation is also likely to constrain the ability of many species to move in response to changing climatic conditions (Collingham & Huntley, 2000; Ewers & Didham, 2006). "
    [Show abstract] [Hide abstract]
    ABSTRACT: Aim Few studies have attempted to assess the overall impact of fragmentation at the landscape scale. We quantify the impacts of fragmentation on plant diversity by assessing patterns of community composition in relation to a range of fragmentation measures.Location The investigation was undertaken in two regions of New Zealand – a relatively unfragmented area of lowland rain forest in south Westland and a highly fragmented montane forest on the eastern slopes of the Southern Alps.Methods We calculated an index of community similarity (Bray–Curtis) between forest plots we regarded as potentially affected by fragmentation and control forest plots located deep inside continuous forest areas. Using a multiple nonlinear regression technique that incorporates spatial autocorrelation effects, we analysed plant community composition in relation to measures of fragmentation at the patch and landscape levels. From the resulting regression equation, we predicted community composition for every forest pixel on land-cover maps of the study areas and used these maps to calculate a landscape-level estimate of compositional change, which we term ‘BioFrag’. BioFrag has a value of one if fragmentation has no detectable effect on communities within a landscape, and tends towards zero if fragmentation has a strong effect.Results We detected a weak, but significant, impact of fragmentation metrics operating at both the patch and landscape levels. Observed values of BioFrag ranged from 0.68 to 0.90, suggesting that patterns of fragmentation have medium to weak impacts on forest plant communities in New Zealand. BioFrag values varied in meaningful ways among landscapes and between the ground-cover and tree and shrub communities.Main conclusions BioFrag advances methods that describe spatial patterns of forest cover by incorporating the exact spatial patterns of observed species responses to fragmentation operating at multiple spatial scales. BioFrag can be applied to any landscape and ecological community across the globe and represents a significant step towards developing a biologically relevant, landscape-scale index of habitat fragmentation.
    04/2010; 19(5):741 - 754. DOI:10.1111/j.1466-8238.2010.00542.x
  • Source
    • "such thresholds is inconclusive and different values are often suggested (Lindenmayer & Luck 2005). These inconsistencies are attributed to the high level of variability in systems and species-specific traits that also have a significant role in determining extinction probabilities (Mac Nally, Bennett & Horrocks 2000; Vos et al. 2001; Sekercioglu & Sodhi 2007). Consequently , there has been little success in translating such advances into real-world prediction solutions for land managers (Mac Nally et al. 2000; Nassauer & Opdam 2008). "
    [Show abstract] [Hide abstract]
    ABSTRACT: Human land-use has a profound influence on wildlife populations; habitat loss can directly decrease population size and carrying capacity, and isolation of the remaining populations can increase their extinction probability. Landscape ecology as a discipline has worked towards creating general rules for the way species respond to landscape change. These rules include, for example, estimates of thresholds at which populations respond more severely to landscape level variables, or general theories as to which species will be more susceptible to landscape change. The demand for these generalisations is driven by the need for inexpensive, rapid and effective methods to manage problems caused by landscape change. The question as to whether general rules are accurate or useful solicits mixed responses from scientists and conservation managers. The most cited reason for this mixed response is the empirical inconsistencies in the way species respond to landscape change. In this thesis I suggest that general rules must be tested in an a priori fashion to directly assess their utility and assist in their translation from theory to practical tool. My primary aim is to test general rules in landscape ecology through creating a priori models; these models are based on ecological theories and existing species and landscape information. My secondary aim is to enhance the understanding of landscape level habitat fragmentation problems for birds in South East Queensland, Australia. I address these aims within four main data chapters as summarised below, where Chapter 1 is a broad introduction to the topic. Chapter 2 asks the question: can general rules and threshold theory be used to predict bird species patch occupancy in a fragmented landscape? I create a simple decision tree model based on threshold theories in landscape ecology, and use this to predict presence or absence of 17 forest bird species in a largely agricultural landscape. This decision tree is broadly based on theoretical patch area and connectivity threshold estimates, and incorporates basic species specific information (such as habitat suitability and mobility). I test this model using a presence/absence survey data set. The process of assessing for which species the model did not work is revealing: I show that the accuracy of ‘present’ predictions is somewhat compromised for habitat specialist species and ‘absent’ predictions are compromised for generalist species. Through creating the ‘optimal’ decision tree models for these species I show that these inaccuracies are likely to arise from vegetation mapping problems, including the lack of a ‘habitat quality’ measure. The study therefore highlights the need for high quality vegetation maps to carry out effective planning. For the majority of species I achieve reasonable predictive success. This study provides hope that general rules have some predictive ability in landscape ecology, and highlights the value of testing models to assess why, and for which species general rules may or may not work. In Chapter 3, I assess the utility of basic ecological principles for predicting the relative value of vegetation patches for specific bird species, focusing on a highly altered urban landscape. I create a model based on the mechanisms expected to be driving species abundance within urban landscapes where most sensitive bird species are likely to be already lost. The model states that a bird species will be more abundance in areas where the vegetation structure matches a species foraging height requirements; however, this effect will be moderated by the landscape context of the patch. From this model I create an index to quantify and rank the predicted value of patches for 30 species of interest in unmanaged and revegetated urban sites, in Brisbane city, Australia. I test the model using bird abundance data, and show that it achieved a reasonable level of predictive accuracy. The model presented within this study is significant as it has relatively low complexity and limited data requirements, yet provides a means to assess how altering the landscape context and vegetation structure within a patch may enhance the abundance of bird species of interest. With further development, the relative simplicity of the model should make it easy to use for land managers. In Chapter 4 I aim to examine how landscape features influence spatial genetic relatedness patterns at a fine, within-population scale on bird species with different life-history traits. I argue that individual level movement characteristics (particularly dispersal routes) in a variable landscape will drive these spatial genetic patterns; thus I create an a priori model based on this theory to make more specific quantifiable predictions of relatedness patterns. I use animal movement theory to deduce these movement characteristics (particularly the strength of avoidance of habitat boundaries) for species with different life-history traits, and apply the model for two closely related passerine bird species which co-occur within South East Queensland (the yellow-throated scrubwren, Sericornis citreogularis, a habitat specialist; and the white-browed scrubwren, Sericornis frontalis, a habitat generalist). I test these models using data on pairwise genetic distances between individuals of each species. The key outcome of this study is that the genetic data supports my predictions that individual level movement characteristics are a mechanistic driver of within-population spatial genetic patterns. For the habitat specialist bird species, the genetic data supported a model which incorporated a strong avoidance response to habitat boundaries and for the generalist species no response to habitat boundaries. This study takes a novel approach to an individual-based genetics study, making specific quantifiable predictions of how a species may be impacted by different landscape features. This research could have significant implications for conservation management, particularly for understanding and managing population responses to a changing landscape, and the early stages of fragmentation. In Chapter 5 I address the question of whether urban revegetation is more successful if it is used to extend the area of existing vegetation, or enhance connectivity in the landscape. This study is novel; for instead of assessing the factors influencing the extinction of a species in a patch, I assess the factors influencing colonisation. Using bird survey data, I use hierarchical partitioning and model selection approaches to determine the relative effect of connectivity and patch area on bird species richness and abundance in revegetated patches. The key finding was that connectivity provided better model fit for bird species richness, and total patch area and connectivity was better for mean bird abundance. My results suggest that the conservation goals of revegetation efforts, particularly in an urban landscape, must be considered when planning a revegetation program. Using revegetation to increase patch area may be the most effective approach for ensuring species persistence over time (i.e. abundance). However, to attract more species into an area enhancing the total area connected in the landscape may be a better approach. In this thesis I explicitly test general rules and theories in landscape ecology within a priori predictive models. Through their generality, the models I develop are potentially suitable for application in other ecosystems. The process of synthesising these models in a simple form, and testing them in a real landscape was revealing. I was able to examine where some general rules do not work, and also where they may not apply or need adjusting. I strived to create models that are easy to use and understand, particularly within Chapters 2 and 4, by trading off simplicity and accuracy. The models produce accurate results to the point that they are arguably valuable tools for landscape managers. This is achieved without compromising their accessibility, and so the research has the potential to transcend the gap between science and real world utility.
Show more