A multiscale methodological approach for monitoring the effectiveness of grassland management

Virágh, K. and Horváth, A. and Bartha, S. and Somodi, I. (2008) A multiscale methodological approach for monitoring the effectiveness of grassland management. Community Ecology, 9 (2). pp. 237-246. ISSN 1585-8553

Text comec.9.2008.2.13.pdf Restricted to Repository staff only until 31 December 2028. Download (422kB)

Abstract

Conservation treatments often take place at the scale of vegetation stands and affect within-stand heterogeneity and coexistence patterns of species first. Therefore, it is important to capture changes in these characteristics of vegetation to assess response to treatments early. We propose a method based on Juhász-Nagy’s information theory models, which is capable of describing fine-scale spatial structure of plant communities and characterizes temporal processes as a function of spatial pattern. The proposed multiscale approach handles structural complexity and its dependence on spatial scales with the help of a few coenological descriptors and helps to reveal how fine-scale vegetation pattern affects dynamics. The information statistical functions used in our study (species combination diversity, FD and associatum, As ) characterize the scale-dependent variability of multispecies coexistence (structural complexity) and multispecies spatial dependence (the degree of spatial organization). The maxima of these functions and the related characteristic areas (plot sizes) can be used to construct an abstract coenostate space, where spatiotemporal processes (degradation, regeneration) can be followed. We demonstrate the usefulness of the proposed methods for detecting degradation and monitoring vegetation changes in different stands (18 seminatural and 13 slightly degraded stands) of Brachypodium pinnatum dominated wooded steppe meadows in Hungary. The information theory measures captured changes of fine-scale vegetation patterns that remained unexplored by species richness and Shannon diversity. The maximum values of information statistical measures and the related characteristic areas detected differences between seminatural and slightly degraded stands. In the coenostate space, seminatural stands appeared to be less variable compared to degraded ones. Seminatural stands from various geographic locations were less dispersed in this space, i.e., less heterogeneous than degraded ones. The two regions of the coenostate-space defined by the set of seminatural and degraded stands were significantly different. Furthermore, we conclude that the region containing seminatural stands can be regarded as a reference region in this abstract space. Temporal variation of seminatural and degraded stands was also clearly different. Therefore, we recommend the approach for exploring the actual dynamic states of vegetation stands to be treated and for following consequences of treatments in order to determine effectiveness of the conservation action.

Actions (login required)

Edit Item