Written by Laura Graham – Postgraduate Research Student
Urban Conservation and Ecological Modelling School of Geography University Of Nottingham
In an attempt to move away from the idea that nature belongs strictly in the countryside, conservation of wildlife in urban areas is gaining increased attention from both policy-makers and ecologists. In the UK, for example, the Lawton report and the Natural Environment White paper both highlight the importance of urban green space within larger habitat networks. Accordingly, it is crucial that biodiversity is fully and effectively integrated into the planning process. It is recognised by ecological theorists, and as such adopted into planning and conservation policy recommendations, that a landscape-scale approach to biodiversity assessment is necessary. I also argue that extinction debt—that there is a time lag between habitat destruction and the effect on species—should be factored into ecological impact assessments. One way of incorporating large spatial and temporal extents into ecological assessments is to take a metapopulation approach.
A metapopulation in ecology is a collection of populations of a species living in discrete habitat patches in a landscape. These populations are separated by what is assumed to be uniformly unsuitable non-habitat, known as matrix. This is a fragmented landscape, which can either be naturally fragmented, such as in island systems, or artificially fragmented, for example the remnant habitat caused by urbanisation. Although local populations tend to have a finite time to extinction, metapopulations are kept in a stable state by local colonisations and extinctions of the species, unless the level of fragmentation exceeds some threshold at which point the habitat can no longer support the species and it becomes extinct.
My PhD research aims to test the use of a specific metapopulation model, the incidence function model (IFM) for investigating biodiversity aspects of urban planning and policy. The IFM simulates local colonisations and extinctions over a finite time period in order to investigate species persistence. The colonisations are simulated by a probability based on patch connectivity, and the extinction probability is based on patch area (see Figure 1). I take a scenario based approach, using the adapted model, to explore the ecological implications of policy decisions. Case studies shall include assessing the impacts of proposed developments on biodiversity at landscape level; assessing impacts of proposed green infrastructure targets using hypothetical scenarios; identifying suitable locations for conservation interventions. I shall also investigate whether the IFM can be used as part of the process to evaluate biodiversity offsetting proposals.
The first part of my research involved collecting species distribution data with which to calibrate the model. As the spatial extent of my study area (Nottingham City and surrounding peri-urban fringe) is large, and I use a suite of indicator species, data collection by me alone would be infeasible. This is where the NBN Gateway comes in. With a huge amount of species records covering a large amount of space and time, these data give a good picture of species distributions over time. I identified two issues with using these data for IFM calibration; the early stages of my research involved developing methods to overcome these issues. The first of these issues was that the IFM requires species occupancy information at habitat patch level, but to get the species and spatial coverage required the finest resolution of data I can use from NBN Gateway is 1 x 1 km.
To address this issue, I have developed and evaluated methods of downscaling data to patch level. The second of these issues was that the IFM requires species presence–absence data; however the data held in the NBN Gateway is mostly presence only. To overcome this issue, I have identified grid cells which I consider 'well surveyed' using a method based on the species-accumulation curve and calibrated the IFM based only on those grid cells.
Current outputs of my research which may be of use to the recording community are maps showing the spatial bias in recording. These may help identify areas to target for future recording effort. At present, these maps have been prepared for my study area for birds, bats and herptiles. The method can however be easily applied to any other geographic area or taxonomic group. The future output of my research will be a model for comparing the relative impacts of policy and planning scenarios on biodiversity.
Find out more about Research applications of NBN data here and examples of use of NBN data for research here