Faculty Publications


Rapidly quantifying reference conditions in modified landscapes

SelectedWorks Author Profiles:

J. Sean Doody

Document Type


Publication Date


Date Issued

October 2008

Date Available

October 2008




Reference conditions remain widely used as a benchmark for ecosystem management, but there remains conjecture about the definition of the reference state. Many techniques used to predict reference conditions are difficult to apply operationally because they are resource-intensive, subjective, or applicable for a limited suite of environmental variables or over a narrow range of environmental variation. We defined the reference state as variation in native vegetation exhibiting relatively little evidence of modification by humans since European settlement. Using data from 462 sites supporting native vegetation in a fragmented landscape in south-eastern Australia, we demonstrated a relatively quick and cost-effective way of objectively predicting reference conditions for various surrogates of biodiversity. We predicted reference values for several variables that are used as biodiversity surrogates (i.e., tree densities by diameter class, trees with hollows, tree regeneration, trees with mistletoe, fallen timber, vegetation cover by vertical stratum, litter cover, cryptogam cover and native plant species richness) using generalized additive models (GAMs) fitted with predictors representing measures of human modification since European settlement (exotic plant cover, number of stumps, evidence of firewood collection, evidence of rabbits, evidence of recent grazing by stock, surrounding land use) and measures of environmental variation (floristic composition, mean annual precipitation, mean annual temperature, solar insolation, aspect, slope). Reference values for each response variable were predicted from these models by holding the significant explanatory variables representing modification since European settlement at their minimum observed values, that is, our definition of the reference state. We demonstrated the importance of independently evaluating predictions of this type using generic ecological models and estimates of reference conditions derived from other sources. (c) 2008 Elsevier Ltd. All rights reserved.


Citation only. Full-text article is available through licensed access provided by the publisher. Members of the USF System may access the full-text of the article through the authenticated link provided.