Landscape planning for biodiversity conservation in agricultural regions: A case study from the Wheatbelt of Western Australia
Biodiversity Technical Paper, No. 2
Robert J. Lambeck, CSIRO Division of Wildlife and Ecology
Commonwealth of Australia, 1999
ISBN 0 6422 1423 9
Appendix 1: Landscape planning for nature conservation at the bioregional scale: Extension of the focal-species approach
The protection of biodiversity in agricultural landscapes requires specification of the types of habitat required, the amount of that habitat, its configuration and how it should be managed. For the majority of Australia's agricultural regions this will require the construction of new habitat as well as the better management of that which remains. The focal-species approach presented in this report provides a procedure for specifying the type, quantity and configuration of habitat and the management regimes required to meet the needs of the biota that occur in a specified area. However, the study found that a solution derived in one landscape could not simply be transferred to another location and that management at a sub-catchment scale was inadequate for meeting the long-term population requirements of the more vulnerable species in the landscape. This raises two critical questions:
- Under what circumstances can a conservation solution derived from a single location be legitimately applied to other locations?
- At what spatial scale does a conservation strategy need to be applied in order to ensure the long term persistence of the species that we wish to retain?
Both of these issues need to be considered in a broader regional context. This section provides a schema for extending the focal-species approach to a bioregional scale, and presents a framework for developing effective conservation strategies for the whole of the agricultural region.
Landscapes differ from each other in both their biophysical and anthropogenic characteristics. Natural variations between regions means that different landscapes support a different biota and hence will have different design and management requirements. In recognition of this fact, regionalisations such as the Interim Biogeographic Regionalisation of Australia (IBRA) have been developed with a view to identifying areas that are sufficiently homogeneous that consistent management practices can be applied.
In agricultural regions, however, the natural biophysical parameters that influence the distribution and abundance of the biota have been significantly modified by patterns of human land use. Landscapes which previously may have supported similar suites of species now retain significantly different biotas simply due to different degrees of human impact.
If we wish to develop conservation recommendations which have relevance beyond the location from which they are derived it will be necessary to partition the existing biophysical regionalisations into landscape 'types' which reflect variation in human land-use patterns. Such a partitioning will enable the identification of those areas that are similar, not only in their natural biophysical patterns, but also in their anthropogenic patterns. These regions would then represent relatively homogeneous Conservation Management Zones (CMZs). It will then be possible to apply conservation recommendations derived from within any CMZ to the remainder of that zone or to other zones with similar characteristics.
By subdividing the existing regions which cover the wheat-sheep zone into a manageable number of CMZs it will be possible to strategically select a limited number of locations to which the focal-species approach can be applied. These analyses will produce conservation specifications for each of the sub-regions resulting in a set of conservation guidelines which would cover the whole of the agricultural region.
Existing regionalisation such as IBRA (Thackway & Creswell 1994) or the Natural Resource Zones developed by the Western Australian EPA and Agriculture WA (Allison et al. 1993) can serve as an initial framework for defining areas that are broadly homogeneous with regard to natural biophysical parameters. Each of these regions can then be analysed further to identify portions of the landscape, such as catchments or sub-catchments that are also similar in terms of clearing pattern and clearing history. Clearing patterns can be derived from maps of remnant vegetation produced under the Bureau of Resource Science's Australian Agricultural land Cover Change Project. GIS procedures can then be used to calculate a range of measures of landscape geometry. Ordination procedures can then be applied to identify a set of landscape 'types', or conservation management zones, for which planning recommendations can be developed using the focal-species approach.
Having identified a set of conservation management zones it will be possible to develop regional design and management specifications for each of these by strategically applying the focal-species approach to representative subsets of each zone. Regional application of the focal-species approach follows the same sequence of steps as was the case for a local exercise.
These steps include:
- identification of threatening processes
- identification of species considered at risk as a result of those threats
- ranking of species in terms of their vulnerability to those threats
- determination of the magnitude of the response required to ameliorate the threat to a level where it will protect the most sensitive species.
Each of the above steps are considered here in greater detail to determine how they may differ between a small-scale application versus a regional assessment.
For much of the agricultural region the types of threats will be generic but their relative importance and the species that they affect may vary between locations. These threats will generally include habitat loss and isolation; habitat degradation as a result of grazing and soil compaction by exotic herbivores, weed invasion and fluxes of fertilisers and other farm chemicals into remnants; predation by cats and foxes; rising saline water tables; and inappropriate fire regimes. Some additional threats such as plant or animal diseases may be occur in particular areas. Expansion of the assessment process to a regional scale may also introduce additional threats if the bigger area encompasses new land uses such as urban development or mining. Consultation with relevant experts and local land-holders will quickly identify the known threats within any given region.
Ideally, the identification of species that are likely to be at risk of local extinction would be based on a comprehensive survey and monitoring process that would provide indications of declining population trajectories. In the absence of such information, the identification of potentially vulnerable species will require the use of expert opinion, anecdotal observations and expectations based on knowledge from elsewhere. In most cases it is possible to quickly generate lists of potentially vulnerable species for a region by consulting local interested land-holders and people with a personal or professional interest in the biota of the region.
In the absence of reliable information about population trends it will be necessary to make some assumptions about which types of species will be most vulnerable to the different types of threats. As was the case for the local application of the focal-species approach, it will be useful to consider two broad categories of threats: those requiring habitat reconstruction and those requiring better management of processes within remnants.
As a general rule, certain species of birds will be most likely to suffer from habitat loss and isolation and hence are most suited for specifying the spatial characteristics of reconstructed habitat. These birds will tend to be habitat specialists that require relatively large habitat patches and can persist only as metapopulations distributed over a number of remnants. Species from other taxonomic groups will be less likely to require the landscape to be redesigned. Rather, they will need improved management of other threatening processes. For example, small to medium sized mammals will generally be vulnerable to predation as will some reptiles and ground-dwelling or ground-nesting birds. Reptiles and invertebrates will be sensitive to altered microhabitats attributable to stock impacts or weed invasion. Plants are also likely to be affected by stock and weeds as well as by inappropriate fire regimes, altered nutrient and chemical levels and, in some areas, diseases such as Phytophthora.
In the absence of information to the contrary, it may be necessary to rely on the spatial requirements of birds to develop the design criteria for a landscape. Species from other taxa can subsequently be used to specify the necessary microhabitat characteristics and management regimes that need to be applied to that landscape. An advantage of using birds to develop design specifications is that it is possible to make some predictions about the species most likely to be at risk and to rapidly survey an area to assess their status and determine the minimum patch characteristics that are required for their occurrence.
Where species from other taxonomic groups are known to be area limited it will be necessary to develop appropriate survey procedures to determine the minimum spatial attributes of the habitat types in which they occur.
Species considered vulnerable may differ between local and regional assessments. Species thought to be at risk locally because they are poorly represented at a particular location may be well represented elsewhere in the region. Consequently they may not be considered vulnerable when viewed from a regional perspective. Conversely, a species that is common in one location may not be seen as being at risk there, but may be deemed vulnerable when considered at a regional scale if this is the only population of that species. Regional assessment will therefore provide a more reliable indication of the status of the biota.
Species that have been identified as being at risk are categorised according to the threat(s) that are responsible for their vulnerability. Within each of these categories, species are ranked according to their relative sensitivity. This process is equivalent to that described for the smaller scale case study.
Similar types of responses will generally be required to manage threatening processes at both local and regional scales. However, the magnitude of the response may change as the spatial scale increases. Not only will theses changes occur in an absolute sense as would be expected as a result of increasing the area over which they are applied, but also in terms of the amount of effort required per unit area.
The effect of changing scale on landscape design
Increasing the spatial scale at which the focal-species approach is applied will alter the sampling strategy that is required to assess the minimum requirements of area-limited and dispersal-limited species. When an exercise is conducted at a local scale it is possible to sample all remnants to provide relatively precise information about the status of various species on remnants with different characteristics. A regional scale application, on the other hand, will require the area to be sub-sampled. Minimum values of habitat area and inter-patch distance obtained from this sample can then be extended to the remainder of the region defined by the regionalisation process described above.
Increasing the scale at which the focal-species approach is applied may reduce the amount of reconstruction required per unit area to meet the needs of the biota in the selected region. This is because reconstruction of habitat within a local catchment will rarely be sufficient to support viable populations of species that occur in low densities. Even if a whole local catchment was revegetated, such species would still be unlikely to occur in sufficient numbers to be viable in their own right. By applying the solution at a larger scale, local land holders will only have to contribute towards a conservation outcome rather than delivering it within their own catchment. This may potentially reduce the amount of effort required per unit area to address a conservation objective.
The effect of changing scale on management responses
Some threatening processes that primarily impact on individual remnants may be managed in much the same way regardless of the area over which the exercise is undertaken. For example, the recommendation to exclude stock from remnant vegetation will hold at both local and regional scales and the management response will be much the same. Similarly, recommendations for weed control will be unlikely to change as the area to be managed increases. However, the intensity of management for some other threats may diminish as the spatial scale increases. An increase in the area over which baiting for feral predators is conducted may enable a reduction in the density and frequency of baiting. A characteristic of local predator control is that the removal of predators from one location results in a rapid influx of new individuals into vacated territories, necessitating an intensive and frequent baiting regime. By baiting over larger areas, the likelihood of predators reaching the central parts of a management area will diminish and the intensity of baiting required around high priority areas may also decline.
If the assumption that the spatial requirements of area-limited and dispersal-limited bird species will encompass the spatial requirements of all other species can be sustained, then application of the focal-species approach at a regional scale could greatly enhance our capacity to rapidly develop landscape designs for nature conservation. This was found to be the case in the Wallatin Creek study and there may be a justifiable basis for making such an assumption throughout much of the agricultural area of Australia. By and large, large mammals tend not to be at risk, while many of the small to medium sized mammals which remain are primarily threatened by predation. Current patterns of habitat occupancy of these smaller mammals in the presence of predators are likely to be significantly different from their distribution and abundance if predation pressures were removed. It is therefore inappropriate to use the characteristics of their currently occupied patches to specify minimum landscape requirements for mammals.
It is also unlikely that we will be able to routinely use the requirements of reptiles, invertebrates or plants to design landscapes for nature conservation at a regional scale. These taxa tend to be most sensitive to loss of microhabitat as a result of grazing by stock, weed invasion and inappropriate fire regimes. Once broad landscape designs are developed it may be possible to use the requirements of these taxa to specify the microhabitat characteristics that need to be incorporated into any habitat reconstruction and the management regimes that need to be applied.
This argument should not be used as an excuse for ignoring mammals, reptiles, invertebrates and plants when considering design. Where expert knowledge is available to indicate the spatial requirements of any of these taxa, they can be incorporated into the focal-species approach to determine whether they exceed the requirements of other species. If such assumptions are to be made, it is essential that monitoring strategies be designed and implemented in order to test them.
It appears then, that in many circumstances, birds may be the group most sensitive to changes in the spatial attributes of habitat at a regional scale. If this is the case, it will increase the tractability of designing landscapes for nature conservation, as birds can be rapidly surveyed to determine presence or absence with a reasonable degree of reliability. Such surveys, conducted over a few weeks in a single breeding season could be used to identify minimum landscape parameters for reconstruction at relatively low cost over a short period of time, provided maps of dominant vegetation types are available. By strategically undertaking bird surveys within the different conservation management zones discussed previously, it will be possible to generate landscape designs for significant portions of the agricultural zone in a relatively short period of time.
Guidelines for management of processes within remnants based on the needs of focal species will be more difficult to obtain rapidly, but it is clear that, in our highly fragmented landscapes, the absence of habitat is likely to be the greatest threat to much of our biota. An ability to rapidly develop parameters for habitat reconstruction at a regional scale that specifically address the spatial needs of the biota in that landscape represents a significant advance in our capacity to manage for nature conservation in agricultural landscapes. Management of processes can be directed, in the short term, by general principles. Longer term investigations which categorise species according to their responses to other threats, and which assess the relative sensitivity of these species to those threats, will subsequently provide more reliable strategies for within-remnant management.
This section has considered only the regional implications of expanding nature conservation strategies to a regional scale. It is important to also consider procedures for integrating the recommendations that this process generates with other land-use objectives. While it is important that the management of biodiversity be undertaken at a regional scale, this scale may not be appropriate for meeting other objectives. For example, the management of hydrological problems may be better managed within catchment boundaries and production objectives may still be best managed within individual farm boundaries.
Given that catchment groups are already widely established throughout the agricultural regions of Australia, it would be inappropriate to expect community groups to undertake formal planning for hydrological management and production at scales larger than their local catchment. It is seen that the procedures described above would be undertaken by appropriate experts (with assistance from community groups where there is sufficient interest) in order to provide the parameters for reconstruction that could be incorporated by land holders into their current planning process.
Links to another web site
Opens a pop-up window