Résumé
Different models indicate that the evolution of trees and the colonization of land by forests drove significant changes in the carbon cycle and a marked drop in the atmospheric CO2 during the Devonian. However, the contribution of early trees and forests to the terrestrial carbon has never been quantified. As a first step to resolve this challenge, we propose to calculate the above-ground biomass and carbon content stored in a Middle Devonian forest structured like the in situ Givetian forest reported at Riverside Quarry, Gilboa, New York State (Stein et al., 2012). Trees in this forest belong to the Pseudosporochnales. Above ground, they consist of a trunk topped by a crown of branches. Pseudosporochnus, the most widely distributed and best known genus of Pseudosporochnales, is chosen here to represent an archetypal tree of these early forests, and Lorophyton to represent an archetypal early growth stage of these trees (Fairon-Demaret & Li, 1993; Berry & Fairon-Demaret, 2002).This communication presents the architectural modeling of a 3-m high Pseudosporochnus tree using the AmapSim software and its computer simulator (Barczi et al., 2008). The carbon content of the tree and of its components is calculated at any time during growth using the mean carbon density of two Carboniferous plants of comparable structure, Psaronius and Medullosa (Baker & Dimichele,1997). The carbon content of a fully grown tree is calculated to range between 837 and 1,300 g C. At 40% of development and beyond, most of the carbon is contained in the trunk. However, when considering the cumulative amount of carbon, i.e. as if the tree had retained all its branches during growth, most of the carbon is allocated to the branches.Given some assumptions on the density and age of the trees, the carbon content of a Pseudosporochnus forest is calculated to range between 4.3 and 15.5 t C/ha. It is relatively low compared to extant forests, the closest analogs in terms of productivity being either thickets of young trees or environmentally constrained forests.The accuracy of the model is discussed and any conclusion about an adaptation of pseudosporochnalean forests to constrained environments must be taken with great care. Yet, it is interesting to note that previous authors reporting forests around Gilboa suggested that they inhabited either stressed or disturbed environments (Mintz et al., 2010; Stein et al. 2012).