14. Habitat connectivity

Summary

Figure 14 (i).  Change in habitat connectivity for selected broad habitats in the wider countryside, 1990 to 2007

 

 

 

Assessment of change in habitat connectivity for selected broad habitats

	Long term	Since 1998	Latest year
Broad-leaved, mixed and yew woodland			N/A
Neutral grassland			N/A

 

• In this indicator, connectivity is a measure of the size and distribution of patches of woodland and neutral grassland, and the relative ease with which typical species of each can move through the landscape between the patches. Maintaining and improving connectivity is important in ensuring the long-term survival of biodiversity in a fragmented landscape, especially under a changing climate.
• The indicator shows little or no overall change in the degree of connectivity for broad-leaved, mixed and yew woodland between 1990 and 2007. Although not shown, over the same period there has been an increase in the area of broad-leaved woodland. The lack of a corresponding statistically significant change in connectivity may be related to changes in woodland pattern, changes in the wider landscape, or both.
• The indicator shows an increase in the degree of connectivity for neutral grassland between 1990 and 2007, although the overall change between 1998 and 2007 is not statistically significant. Between 1990 and 2007 there has been an increase in the overall habitat area. The increase in connectivity is most likely to be related to an increase in habitat area, but there may also be effects from changes in habitat pattern in the wider landscape.
• The indicator is based on an analysis of changes in land cover recorded in Countryside Survey – a detailed periodic audit of a statistically representative sample of land across Great Britain. Expert opinion was used to assess the relative likelihood of movement, by species characteristic of each habitat, between habitat patches, across different intervening land cover types found in the survey.
• Further analysis is required to better explain the causes of the changes in connectivity (which may be due to changes in the extent of the habitat or changes around the habitat blocks). Until this analysis has been undertaken, the current information is insufficient for an assessment of change to be made, despite the statistically significant increase seen in connectivity in neutral grassland habitat.

Indicator description

The indicator shows changes in the degree of habitat connectivity in Great Britain for two Broad Habitats:

1. Broadleaved, mixed and yew woodland.  This includes all woodland with a canopy cover of at least 25 per cent, where more than 80 per cent of the canopy trees are broad-leaved species or yew trees.
2. Neutral grassland, which includes all grassland on neutral soils including both unimproved and semi-improved grassland.

The indicator provides a measure of functional connectivity of these two habitats in the wider landscape, i.e. the relative likelihood of species typical of the habitat being able to move within and between habitat patches.  The calculations underlying the connectivity indicator take into account the area of habitat patches, how isolated they are, which habitats are next to each other, and the ease with which species are able to move through the surrounding landscape.  The influence of habitat quality on species is only partially covered by this indicator.

Figure 14 (ii).  Confidence intervals for the indicator measures.

The assessment of the indicator is undertaken by evaluation of the 95 per cent confidence intervals between the first and latest surveys undertaken (1990 and 2007), and between the two most recent surveys (1998 and 2007).  These are shown in Figure 14 (ii). 

The indicator shows a non-significant declining trend in the connectivity of Broad-leaved, Mixed and Yew Woodland in Great Britain and an increasing trend in the connectivity of Neutral Grassland in Great Britain.  The trend for Neutral Grassland was significant between 1990 and 2007 but not in the short term between 1998 and 2007. 

Relevance

Habitat fragmentation is the process whereby natural and semi-natural habitats are separated into a number of smaller patches, through land use change and urban development.  Habitat loss and fragmentation reduces the size of populations and hinders the movement of individuals among increasingly isolated populations, threatening their long-term viability.  The process can be cumulative over time, but may be reversed through habitat management, restoration and recreation.  Many of the habitats in the UK landscape are already highly fragmented.

The effects of habitat fragmentation can be compounded by changes in land use between the patches.  The importance of these changes depends on which habitats are next to each other (edge effects) and the ease with which species can move through the intervening landscape (permeability).

Habitat loss and fragmentation was identified by the Millennium Ecosystem Assessment as one of five direct drivers of biodiversity loss.  The impacts of habitat loss and fragmentation are recognised within the Convention on Biological Diversity and the European Union Habitats Directive.

In 2007 Defra published a set of climate change adaptation principles (Mitchell et al 2007).  These set out actions that can be taken to protect biodiversity now, whilst increasing the chance that biodiversity will be able to adapt to future changes.  These principles include taking action to increase area of habitats available for biodiversity and improving the opportunity for wildlife to disperse through the countryside.

Background

The development of the connectivity indicator is an innovative use of available data and limited information about how species move through landscapes.  The connectivity indicator has been calculated for two Broad Habitat types (Broad-leaved, Mixed and Yew Woodland, and Neutral Grassland) using field survey data collected consistently from 591 Countryside Survey 1km² sample squares in Great Britain, in 1990, 1998 and 2007. 

The two habitats were selected because of their widespread and dispersed occurrence across Great Britain, allowing changes to be detected at the 1km² scale of the sample squares.  Woodland and grassland are also at opposite ends of the continuum of ecological succession; woodland tends to be long-lasting and stable, whereas grassland is more dynamic.  Selecting these habitats has allowed the approach to be tested on data which is representative of the Countryside Survey dataset as a whole and relevant to environmental policy making. 

Woodland in the Countryside Survey is defined as having a 25 per cent canopy cover, so thinned or recently replanted stands, newly created woodland, or woods with low tree density are not included.  This means that connectivity of this habitat may be underestimated.  However, felling and replanting of previously wooded areas may represent an interruption in connectivity for some species, which the indicator can detect. 

A further complication encountered in calculating this indicator relates to the definition of the woodland BAP Broad Habitats.  Broad Habitats did not exist when the 1990 Countryside Survey was undertaken and habitats were assigned for analysis in 1998 using detailed attributes.  Considerable work has been undertaken to achieve consistency across Broad habitats in the Countryside Survey data for 1990, 1998 and 2007 including re-allocation of both 1998 and 1990 data and field based correction of 1998 polygons by 2007 surveyors.  Despite that, where sufficiently detailed data was not available for re-allocation in 1990, it may be that polygons which were allocated to Coniferous Woodland in 1998 and 2007 were allocated to Broad-leaved, Mixed and Yew Woodland in 1990.  The definition of 80 per cent cover of broad-leaved and yew species now used for Broad-leaved, Mixed and Yew Woodland in Countryside Survey means that mixed coniferous / broad-leaved woodlands will be recorded as Coniferous Woodland.  The potential extent of this issue requires further exploration but may lead to inconsistency in the indicator between 1990 and other survey dates, although relatively small proportions of woodland surveyed in Countryside Survey (~5 per cent) are mixed conifer/broad-leaf.

The definition for the Neutral Grassland broad habitat type includes not just Priority Habitats (upland and lowland hay meadows) but also other semi-improved habitats with species more typical of improved grassland.  The increase in connectivity observed may mask differences in trends for more tightly defined habitats which fit within the broad habitat definition. 

As the indicator is based on Countryside Survey data, the extent to which change in the indicator represents real change in connectivity across Great Britain will be dependent on the spatial and temporal resolution of Countryside Survey data.  The connectivity values calculated range from zero (no habitat in the sample squares) to one hundred (habitat completely covering the sample squares, with complete intra-patch connectivity).  Due to the highly fragmented nature of many British habitats, it is expected that most connectivity values will be very low. 

The method used to calculate the indicator accounts for the influence of the habitats which are adjacent to a patch of woodland or grassland.  These ‘edge impacts’ (i.e. effects of adjoining land cover) are taken into account by considering the surrounding land cover of each patch.  A strip of habitat may be ‘removed’, to reflect that, from the point of view of the species modelled, the area of habitat actually used is smaller than that mapped by the fieldwork.

The method then calculates the probability of movement of a hypothetical species typical of a particular habitat, both within and between the habitat patches.  These potential movements are weighted by:

• the area of the source and destination patches, with more potential movements from (or to) larger patches;
• an expectation that the majority of movements will be to nearer existing patches (implemented by applying a negative exponential dispersal curve);
• and an expectation that there will be greater potential movement through ecologically similar landscape features, as opposed to more restricted movement through habitats which are less suitable for the species (calculated through a least-cost distance measure).

The weightings used to implement these movements were derived from an analysis of expert opinion, as comprehensive ecological requirements for all species of these habitats are not available.

A reduction in habitat area may lead to a reduction in the size of a species’ population and so to an increased risk that a species will cease to exist within a group of patches.  An increase in ecological isolation may also cause a reduction in the movement of individuals between isolated patches.  These movements may maintain genetic diversity, rescue declining populations, or re-establish populations in patches from which they have previously been lost; they are important in maintaining the dynamics and viability of species populations at a local or regional level.

Further development planned

The indicator methodology has been developed by Forest Research in collaboration with the Centre for Ecology and Hydrology, using Countryside Survey data. The results of this work provide a significant step forward in our ability to understand and describe habitat fragmentation and connectivity. However, the indicator is currently presented as unassessed pending further investigation of the underlying data to better interpret the observed changes. This further investigation is a priority for future work. Additional further development is planned that would allow disaggregation of the data for individual country reporting; and further refinement through the addition of new empirical data on species movement through different landscape types.

Details of development of the habitat connectivity are available in Watts et al (2008).

References

Watts, K. and Handley, P. (2010). Developing a functional connectivity indicator to detect change in fragmented landscapes. Ecological Indicators 10, 552-557.

Watts, K., Handley, P., Eycott, A.E., Peace, A., Marzano, M., Scholefield, P. and Norton, L. (2010a). Habitat Connectivity – Developing an indicator for UK and country level reporting  Phase 2: Production of the indicator - Defra contract  WC0716. Forest Research & Centre for Ecology and Hydrology.

Watts, K., Eycott, A.E., Handley, P., Ray, D., Humphrey, J.W. and Quine, C. (2010b). Targeting and evaluating biodiversity conservation action within fragmented landscapes: An approach based on generic focal species and least-cost networks. Landscape Ecology 25(9), 1305-1318

Eycott, A.E., Watts, K., Brandt, G., Buyung-Ali, L.M., Bowler, D., Stewart, G.B. and Pullin, A.S. (2010). Which matrix features affect species movement? Systematic Review No. 43. Collaboration for Environmental Evidence, Bangor, Wales, UK.

Eycott, A.E., Watts, K., Brandt, G., Buyung-Ali, L.M., Bowler, D.E., Stewart, G.B. and Pullin, A.S. (2008). What is the evidence for the development of connectivity to improve species movement, as an adaptation to climate change? Defra contract WC0705. Forest Research & Centre for Evidence-Based Conservation.

Watts, K., Handley, P., Scholefield, P. and Norton, L. (2008). Habitat Connectivity – Developing an indicator for UK and country level reporting. Phase 1: Pilot Study - Defra contract WC0704. Forest Research & Centre for Ecology and Hydrology.

Mitchell, R.J., Morecroft, M., Mountford, O., Acreman, M., Ross, L.C., Maclean, I.M.D., Crick, H.Q.P., Rehfisch, M.M., Piper, J. and Wilson, E. (2007). England Biodiversity Strategy: Towards Adaptation to Climate Change. Contract report CRO327 for Defra.

Watts, K., Humphrey, J.W., Griffiths, M., Quine, C.P. and Ray, D. (2005). Evaluating Biodiversity in Fragmented Landscapes: Principles. Forestry Commission Information Note No.073. Forestry Commission, Edinburgh.

For further information about changes in woodland and grassland habitats, see also Indicator 2 – Plant Diversity.

Web links for further information