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
Chapter 3 - Integrating biodiversity conservation with other land uses (continued)
Stakeholder groups were all understandably cautious about entering into an integrated management process for which the outcomes were unclear. It is clearly unreasonable to expect stakeholder participation in a process for which the level of commitment of all stakeholders and the type of outcomes are not clear at the outset. It was therefore considered necessary to develop a prototype model for the catchment based on preliminary information. This prototype was then presented to the various stakeholder groups in the catchment to demonstrate its capabilities and to enable them to assess its appropriateness for the management of their catchment. Having viewed the prototype, the various groups were then able to ensure that the appropriate issues and objectives were being adequately addressed by the process.
Each stakeholder group had the opportunity to specify the issues, landuses, and guidelines that they considered appropriate for the region from their own perspective. Unfortunately, because of the incompleteness of the available information for both land and nature conservation, and for alternative and more sustainable production land-uses, the recommended outcomes must still be considered preliminary. Further development of guidelines will be required before an acceptable catchment plan can be generated and implemented.
Responses by stakeholders to the preliminary solutions were positive. There was a general perception that the requirements for addressing nature conservation objectives were within the scope of what they were prepared to undertake and that these actions would also make a positive contribution to managing hydrological issues and to enhancing the amenity value of the landscape. It must be recognised, however, that land-owners in the Wallatin Catchment are among the most progressive in the wheatbelt when it comes to the consideration of nature conservation as a legitimate land-use objective.
Further development of hydrological guidelines will be necessary before the landholders can assess their acceptability. The strategy used in the current exercise was one of widespread alley planting. More strategic solutions if available, would be viewed more favourably by the group.
Identification of the Wallatin Catchment as a Focus Catchment under the State Government Salinity Action Strategy will ensure that further advice is made available to the catchment group. The procedures developed in this study will provide a basis for catchment planning under this strategy. Ideally, the future development of this approach should include consideration of a wider array of land-uses such as the commercial use of indigenous plants and animals, including wildflower enterprises and the production of oil from Eucalyptus and Melaleuca species. Future planning should also consider the increasing demand for 'cleaner and greener' produce. However, it will be necessary to convince land-holders that these new activities will be commercially viable.
A prerequisite for reliable land-use allocation is the availability of accurate maps which reliably reflect land capability. The map sets available for this exercise were derived from a number of different sources. Landform maps were produced by W. M. McArthur using air photo interpretation and ground survey. These maps were produced with the objective of partitioning the landscape into units which reflected a combination of attributes including vegetation type, soil type, and position in the landscape. Remnant vegetation was mapped from aerial photographs and digitised for incorporation into a geographic information system by Agriculture WA and CSIRO Division of Wildlife and Ecology. Vegetation communities within remnants were mapped using Landsat MSS and Landsat TM satellite images (Hobbs et al. 1989; Lambeck & Wallace 1993) combined with air photo interpretation and ground surveys to validate and, where necessary, correct the satellite data. Soil types outside of the remnants were mapped by individual farmers to reflect suitability for agricultural land uses using procedures described in Hawkins (1990).
In practice, many of the above map sets should be correlated to some degree. The distribution of soil types should reflect the landforms in the area and vegetation should show some correspondence to both soils and landforms. However, the extent of these relationships between map sets appeared to vary throughout the study area reflecting differences in the interpretation of the landscape by different people. This was particularly evident where individual farmers were responsible for developing soil maps of their properties. For some farms, the soil was mapped at a very fine resolution, reflecting small scale variation in soil types. Other farmers appeared more pragmatic in their mapping, primarily capturing differences that best reflected land capability for current farming practice. Neither of these approaches is more correct than the other. They simply reflect differences between individuals in their perception of the purpose of the mapping exercise. In addition, attempts to create discrete boundaries between soil types which grade into each other is a subjective process. No two individuals could derive the same map and probably no one individual could produce the same map twice. As a consequence, soil boundaries often did not match on opposite sides of a farm boundary and soil boundaries failed to match vegetation boundaries at the edges of remnants. In addition, if remnant vegetation as mapped by Government agencies does not match the maps generated by landholders, further errors can be introduced into the exercise and the reliability of the outcomes will diminish correspondingly.
Because of the discrepancies between vegetation and soil maps it was necessary, in this exercise, to use landform maps as a surrogate for vegetation types as this was the only data set that was available both inside and outside of the remnants. However, it is clear from satellite mapping that the match between vegetation types and mapped landforms is also relatively poor. Consequently the results generated by the land allocation exercises can only be considered as a general guide to what can be done rather than a prescription that must be implemented exactly as mapped.
For this procedure to more effectively reflect the needs for agriculture and the needs of the plants and animals in the landscape, it will be necessary to ensure greater correspondence between data sets. This should be based on standardised data capture protocols and appropriate validation of maps.
Hydrological guidelines for this exercise were based on general landscape principles. As a consequence, there are no criteria for assessing the likely impact of the specified actions on water table levels. If such predictive capacities are required it will be necessary to develop hydrological models based on reliable maps of geomorphological features and high resolution digital elevation models. Failure to develop effective strategies for managing hydrology will not only compromise the agricultural values of the catchment, but will also threaten much of the remnant vegetation that is essential for protecting the remaining biota.
The ramifications of these shortcoming in data quality are significant. The reliability of the results from any planning process will only be as good as the quality of the information that is used. Even the best models and procedures cannot convert poor information into reliable solutions. If planning seeks to provide reliable outcomes, then it must be based on reliable inputs. The absence of standards and consistency in data collection, the lack of agreement about the type of data that should be collected, and the lack of resources for gathering the information even if such agreement can be reached, present major problems for any regional planning process. Planning based on poor quality data will result in actions which do not deliver the required outcomes and will alienate stakeholders who enter into the planning process in good faith with an expectation that their problems will be addressed. However, the absence of quality information should not provide an excuse for not undertaking actions that are urgently needed. Responses to data limitations for hydrological objectives should be treated in a similar way to those recommended for nature conservation: where there is an urgent need to act and the necessary information is not available, then general principles should be employed. Where the situation is less urgent, attempts should be made to acquire the information required for a more strategic approach.
Regional approaches to planning can help to reduce these data limitations. By identifying areas that are sufficiently similar in regard to natural and human-induced pattern it will be possible to develop recommendations based on quality data acquired for a subset of the region and legitimately apply those results beyond the location where they were generated.