Thermodynamic entropy fluxes reflect ecosystem characteristics and succession

Ecosystems are open systems that constantly exchange energy and entropy with their surroundings, giving rise to continuous evolution in ecosystem's internal structures and functions. For an ecosystem to evolve towards more developed stages and to sustain its functions, I hypothesize that (1) an ecosystem's net thermodynamic entropy budget must be negative (i.e., exporting entropy) and (2) the rate of entropy production should increase with succession towards a possible maximum (i.e., maximum entropy production or MEP) before a decline may occur. To test these hypotheses, energy flux data collected from Ameriflux and Canada Flux using eddy tower covariance method at seven sites (including three groups of forested vs. deforested ecosystems) were compared by converting energy into entropy fluxes. Based on energy budgets and corresponding absolute temperatures for each energy components, net entropy production rate for each ecosystem (Delta S-e) was calculated every half-an-hour for multiple years (2004-2009, which varied among the sites). All the ecosystems examined clearly displayed predominantly negative Delta S-e, indicating higher entropy outputs than inputs. The ratio of entropy outputs to inputs (SR) ranged from 1.23 to 1.53 (averaged 1.35 +/- 0.10) based on averaged daily fluxes across all the ecosystems, with higher SRs at midday (1.46-2.23, averaged 1.80 +/- 0.24) and lower SRs at midnight (1.09-1.22, averaged 1.13 +/- 0.04). Ecosystems having higher entropy production rates also generally produced a higher SR, indicating a more efficient energy dissipation. The overall mean Delta S-e ranged from 0.258 W/m(2)K in the deforested site in Saskatchewan, Canada to 0.609 W/m(2)K in the forested ecosystem in Flagstaff, Arizona. The deforested sites had, on average, 27% and 18% lower Delta S-e in Saskatchewan and Flagstaff, respectively, than their respective forested sites. Two forested sites at the Duke Forests, North Carolina had higher vertical bar Delta S-e vertical bar than the grass field. However, the forest site at a later stage of succession had slightly lower vertical bar Delta S-e vertical bar than the planted forest. This may suggest a gray area in defining how an ecosystem evolves towards the state of MEP and the definition of climax. There is also a possibility of retrogression after a forest reaching a possible climax. This study shows that the MEP hypothesis may predict the re-forestation of disturbed sites, but the hypothesis needs to be further tested for succession once a forest cover has been re-established. (C) 2014 Elsevier B.V. All rights reserved.