Submitted to: African Journal of Range and Forest Science
Publication Type: Proceedings
Publication Acceptance Date: 1/15/2003
Publication Date: 6/1/2003
Citation: Northup, B.K., Brown, J.R., Ash, A.J. 2003. Using variability to identify potential state and transition stages of tropical tallgrass sites in northeast australia. African Journal of Range and Forest Science. v. 20. Abstract p. 124. Interpretive Summary: Describing the condition of grazing lands in northeast Australia is difficult. Conceptual state and transition models that describe changes in land condition are used to help producers understand the process. However, these models lack standards for the indicators that define state conditions. Plant communities in this area are naturally variable, making the description of pasture condition difficult. This variability could also serve as a useful monitoring tool. We tested whether pattern changes in herbaceous vegetation might serve as better indicators of condition than pasture-level averages of responses. We collected information during 1993, 1995, and 1998 from experimental pastures on tropical tallgrass sites that were under different grazing pressures. Forage produced by indicator species, and basal area of perennial grasses, were determined by estimation techniques on 100 small rectangular plots along 110-yard, fixed line transects. Averages for these indicators were developed for pastures that were degrading under heavy grazing, and analyzed by traditional statistical techniques. Line diagrams of variation in indicators were also constructed for each transect. We found that pasture averages could not identify condition changes due to high variability, and large sample numbers (n>1000) were required to calculate accurate averages. Line diagrams easily identified changes in amount and distribution of indicators, including: high basal area and abundant desert bluegrass (indicators of good condition); presence of annual grasses and forbs (indicators of over-grazing); and presence of Indian couchgrass (indicator of degradation). Australian tropical grasslands respond quickly to disturbance, and monitoring tools that fail to identify the early stages of damage contribute to degradation. Using spatially oriented measures of variation as indicators could allow rapid identification of condition changes, and improve management decisions.
Technical Abstract: Conceptual state [S(i)] and transition models are used to help determine land condition in northeast Australia, but lack standards for amount(s) of indicators that define state conditions. Endemic variability in the spatial distribution of vegetation makes describing condition changes difficult, but could also serve as a useful monitoring tool. This study tested whether patterns of spatial variability in herbaceous vegetation might serve as better indicators of state condition than species composition responses. Data were collected during 1993, 1995, and 1998 from experimental paddocks, under different grazing pressures, on a tropical tallgrass site. Forage produced by indicator species, and basal area of perennial grasses, were ascertained by BOTANAL procedures on 100, 0.5 m^2 quadrats along fixed transects and standardized across all observations (n=8000). Means from paddocks degrading under heavy grazing were analyzed by univariate statistics, and line diagrams of standardized variation were constructed. Paddock-scale means of commonly measured attributes were ineffective at identifying changes due to high variability (c.v. 50-140%), and large sample numbers (n>1000) were required. Conversely, line diagrams identified changes in amount and distribution of indicators, including: basal area and Bothriochloa ewartiana (S1 indicators, 1995 and 1998); annual grasses and forbs (S2 indicators, 1995); and Bothriochloa pertusa (S3 indicator, 1995 and 1998). Tropical tallgrass sites can respond rapidly to disturbance, and monitoring tools that do not identify early stages of degradation contribute to the damage. Spatially oriented measures of variability could allow more accurate identification of changes in condition than paddock-scale means, and help improve management decisions.