Statistical Modeling of Litter Decomposition from Integrated Cellulase Activity

Abstract

Efforts to model plant litter decomposition have centered around three approaches. On large spatiotemporal scales, decomposition rates have been estimated from climate parameters, typically temperature and potential evapotranspiration (e.g., Insam 1990). At the ecosystem scale, litter degradation models based on empirical fits to mass loss curves have been widely used (Webster and Benfield 1986, Boulton and Boon 1991). The simplest of these is the first-order exponential decay model. Rate constants for this model are routinely reported in litter decomposition papers, but at the same time, this approach is generally considered unrealistic. More precise twoand three-stage models have been developed that recognize different stages in decomposition, but their predictive value is limited. More general models have been produced by relating decomposition directly to litter chemistry. In this approach, both simple parameters, e.g., nitrogen and lignin, and compound parameters, e.g., C:N, fignin: N, and lignocellulose index, defined as lignin/(lignin + holocellulose), have been used as indices of biodegradability (Melillo et al. 1982, Taylor et al. 1989). While qualitatively predictive, these models do not consider the role of edaphic and climatic factors in the regulation of biotic activity, so cannot predict rates. In this report, we describe a new process-level modeling approach focused on microdecomposer activity. It offers three advantages: (1) ease of application, (2) realism, and (3) potential for linkage to larger scale models. The approach is an empirical one: the potential activities of key enzymes directly involved in the degradation of plant structural components are used as indices of microdecomposer activity. Previous studies have established correlations between cellulose activity and mass loss from decomposing leaf litter (Sinsabaugh et al. 1991, Sinsabaugh 1993). However, such correlations do not permit mass loss estimation because of the temporal difference between mass loss, a cumulative parameter, and enzyme