Titre :	Ecosystem sensitivity to climate change and complex responses
Type de document :	texte imprimé
Auteurs :	James S. Clark, Auteur
Editeur :	Chapman and Hall
Année de publication :	1991
Importance :	p 65-98
Langues :	Anglais (eng)
Catégories :	SCIENCES DE L'ENVIRONNEMENT
Mots-clés :	SENSIBILITE D’ECOSYSTEME  SENSIBILITE SUR LE CHANGEMENT CLIMATIQUE  REACTION FORESTIERE  SAISONNALITE
Résumé :	The lessons of history and existing climate dependencies in ecosystems indicate ranges of sensitivities and complex responses. Broad-scale ecosystem processes, such as NPP, may be most sensitive to small changes in water balance in deserts, grasslands and temperate and conifer forests. Decomposition rates and accumulation rates of detritus may be more sensitive in temperate hardwood forests because decomposition rates slow with increasing latitude to a greater degree than do production rates. Because of their higher litter quality , decomposition rates in hardwood forests are likely to be more sensitive to small climatic shifts than are those in conifer stands. This statement refers only to sensitivity, however, as protracted climate change and changes of large magnitude could have greater effects in boreal conifer forests due to the greater accumulation of organic matter in those areas. Nutrient cycling is sensitive to all aspects of macroclimate, microclimate, seasonality, local vegetation cover and disturbance. Different aspects of nutritient cycles respond in different ways to climate change, therefore it is difficult to speculate on the composite effects in given ecosystems. Fire regime appears sensitive to climate change throughout most of the temperate forests, but it is likely that drier conifer forests will display greater sensitivity to climate change than will mesic forests. As long as we have only the vaguest notion of how climate change will change in any given area, however, it is going to be impossible to predict how ecosystems will change. Even if the boundary conditions used in GCMs could be forecast with high confidence, it would still be exceedingly difficult to predict ecosystem transition at local scales because of the complexities of the climate system and of the ecosystem responses. Prediction of ecosystem change must build from this shaky foundation; ecosystem simulations using GCM output are only as good as the climate assumptions they contain. In fact, they are much worse because they contain their own uncertainties . Perhaps half of the papers published on forest ecosystems in leading journals investigate processes that stand-simulation models must assume we already understand. They contain dozens of functional forms and tens to hundreds of parameters. These problems that result from lack of empirical data and ecosystem understanding are no reson to abondon simulation models. They are cause to question model output and to explore the more fundamental question of climate sensitivity, either through field manipulation or synthesis of climate/ecosystem relationships. We have better potential for assessing ecosystem sensitivity than we do for prediction. There is much existing evidence to suggest which processes are sensitive and where and much coild be done to increase this knowledge.
Numéro du document :	A/MAC
Niveau Bibliographique :	2
Bull1 (Theme principale) :	METEOROLOGIE ,ATMOSPHERE,CLIMATOLOGIE
Bull2 (Theme secondaire) :	ATMOSPHERE,CLIMATOLOGIE-CONSIDERATION GENERALE

Ecosystem sensitivity to climate change and complex responses [texte imprimé] / James S. Clark, Auteur . - Chapman and Hall, 1991 . - p 65-98.
Langues : Anglais (eng)

Catégories :	SCIENCES DE L'ENVIRONNEMENT
Mots-clés :	SENSIBILITE D’ECOSYSTEME  SENSIBILITE SUR LE CHANGEMENT CLIMATIQUE  REACTION FORESTIERE  SAISONNALITE
Résumé :	The lessons of history and existing climate dependencies in ecosystems indicate ranges of sensitivities and complex responses. Broad-scale ecosystem processes, such as NPP, may be most sensitive to small changes in water balance in deserts, grasslands and temperate and conifer forests. Decomposition rates and accumulation rates of detritus may be more sensitive in temperate hardwood forests because decomposition rates slow with increasing latitude to a greater degree than do production rates. Because of their higher litter quality , decomposition rates in hardwood forests are likely to be more sensitive to small climatic shifts than are those in conifer stands. This statement refers only to sensitivity, however, as protracted climate change and changes of large magnitude could have greater effects in boreal conifer forests due to the greater accumulation of organic matter in those areas. Nutrient cycling is sensitive to all aspects of macroclimate, microclimate, seasonality, local vegetation cover and disturbance. Different aspects of nutritient cycles respond in different ways to climate change, therefore it is difficult to speculate on the composite effects in given ecosystems. Fire regime appears sensitive to climate change throughout most of the temperate forests, but it is likely that drier conifer forests will display greater sensitivity to climate change than will mesic forests. As long as we have only the vaguest notion of how climate change will change in any given area, however, it is going to be impossible to predict how ecosystems will change. Even if the boundary conditions used in GCMs could be forecast with high confidence, it would still be exceedingly difficult to predict ecosystem transition at local scales because of the complexities of the climate system and of the ecosystem responses. Prediction of ecosystem change must build from this shaky foundation; ecosystem simulations using GCM output are only as good as the climate assumptions they contain. In fact, they are much worse because they contain their own uncertainties . Perhaps half of the papers published on forest ecosystems in leading journals investigate processes that stand-simulation models must assume we already understand. They contain dozens of functional forms and tens to hundreds of parameters. These problems that result from lack of empirical data and ecosystem understanding are no reson to abondon simulation models. They are cause to question model output and to explore the more fundamental question of climate sensitivity, either through field manipulation or synthesis of climate/ecosystem relationships. We have better potential for assessing ecosystem sensitivity than we do for prediction. There is much existing evidence to suggest which processes are sensitive and where and much coild be done to increase this knowledge.
Numéro du document :	A/MAC
Niveau Bibliographique :	2
Bull1 (Theme principale) :	METEOROLOGIE ,ATMOSPHERE,CLIMATOLOGIE
Bull2 (Theme secondaire) :	ATMOSPHERE,CLIMATOLOGIE-CONSIDERATION GENERALE