Part III Global Scale Processes Flan chap13 2004 8 24 17 26 page 215 1 Flan chap13 2004 8 24 17 26 page 216 2 13 Terrestrial Ecosystems and Interannual Variability in the Global Atmospheric Budgets of 13CO2 and 12CO2 Introduction Several recent ecosystem studies provide evidence that plant discrimination against 13 C during photosynthesis is highly variable on synoptic to interannual timescales Critical axes of variation include the response to drought stress in temperate and tropical forests Ekblad and H gberg 2001 Bowling et al 2002 Mortazavi and Chanton 2002 Ometto et al 2002 Fessenden and Ehleringer 2003 and shifts in the abundance of C3 and C4 vegetation both in response to land use Vandam et al 1997 Townsend et al 2002 and climate Tieszen et al 1997 Still et al 2003b The purpose of this chapter is to place this variability in the context of the global carbon cycle First I present equations describing the global atmospheric budget for 13 CO2 and 12 CO2 in graphical form using Robin Hood diagrams Robin Hood diagrams provide an intuitive means for comparing isotopic fluxes from terrestrial ecosystems with those from the oceans and fossil fuels and for partitioning ocean and land carbon sinks using atmospheric 13 C and CO2 observations Enting et al 1994 Second I critically review the mechanisms by which terrestrial ecosystems are thought to impart isotopic anomalies to the atmosphere on interannual timescales A question that has generated considerable debate in the last few years is what fraction of the interannual variation of 13 C in the global atmosphere should be attributed to changes in land and ocean carbon sinks and what fraction should be attributed to changes in plant discrimination and other processes During the 1990s the growth rate of 13 C was highly variable with the most rapid declines occurring during the middle to latter stage of the 1997 1998 El Nino event Fig 13 1 Battle et al 2000 217 Flan chap13 2004 8 24 17 26 page 217 3 218 13 Terrestrial Ecosystems and Interannual Variability of 13 CO2 and 12 CO2 370 CO2 ppm yr 1 A B 365 360 355 7 80 13C 7 85 7 90 7 95 8 00 2 1 0 0 05 0 00 0 05 01 00 20 99 20 98 19 97 19 96 19 95 19 94 19 93 19 19 19 92 0 10 91 13C rate of change yr 1 D 3 19 C CO2 rate of change ppm yr 1 8 05 Year Figure 13 1 A Global CO2 concentrations derived from observations from the NOAA Carbon Monitoring Diagnostics Laboratory flask measurements http www cmdl noaa gov ccg Battle et al 2000 The global mean was constructed by removing a mean seasonal cycle averaging all stations within 6 latitude zones 0 30 30 60 and 60 90 in each hemisphere weighting each latitude zone by surface area and smoothing with a 1 year box car running mean filter B Same as A but for 13 C Battle et al 2000 C The atmospheric growth rate of CO2 was highly variable during the 1990s and reached a maximum during late 1997 and early 1998 during the strong El Nino event observed at that time D Changes in atmospheric 13 C mirror changes in CO2 shown in C but have an opposite sign A flux with a 13 C of 25 explains most of the variability whereas a flux with a 13 C of 9 explains only a small fraction This suggests that terrestrial ecosystems contributed to most of the variability during this period Flan chap13 2004 8 24 17 26 page 218 4 Robin Hood Diagrams 219 In this context several recent modeling analyses have highlighted a number of independent mechanisms that may or may not work in parallel to explain the atmospheric anomalies apart from sources and sinks At least three of the proposed mechanisms are connected to the shifting patterns of precipitation that accompany El Nino events During El Nino less rain falls on land particularly in tropical ecosystems Ropelewski and Halpert 1996 This decrease causes global scale drought stress lowers leaf internal CO2 concentrations and causes a temporary decrease in global plant discrimination Randerson et al 2002 Scholze et al 2003 It is also likely that precipitation shifts between C3 and C4 ecosystems causing a change in the relative contribution of these plant functional types to global gross primary production GPP and thus global discrimination Scholze et al 2003 Finally global drought stress also changes the distribution and abundance of fires Langenfelds et al 2002 Van der Werf et al 2004 including their isotopic composition which is not well known No contemporary modeling framework has simultaneously considered all of the mechanisms that have been proposed and it is unlikely that any single mechanism is exclusive In this context I conclude by offering several directions for experimental inquiry that could guide model development and ultimately seem necessary for a predictive understanding of interannual variability in discrimination at large spatial scales A Graphical Means to Understanding the Global Atmospheric Budget Robin Hood Diagrams Overview Here I provide a step by step description of each vector in a Robin Hood diagram of the global atmospheric carbon budget Fig 13 2 The budget and figure is constructed arbitrarily for the year 1990 This vector approach is referred to as a Robin Hood diagram because of the abundance of arrows Inez Fung personal communication 2000 and the overall process of solving for land and ocean carbon sinks using 13 CO2 and 12 CO2 is known as a double deconvolution Heimann and Keeling 1989 In the conventional representation of the global budget observational changes labeled with a 1 in Fig 13 2 represent some combination of fossil fuel emissions 6 land disequilibria forcing 5 a terrestrial carbon sink 4 ocean disequilibria forcing 3 and an ocean carbon sink 2 In a more general representation of the budget Fig 13 2B both gross primary production 7 and a return flux from the biosphere to the atmosphere 8 are separately considered Randerson et al 2002 In this form it can be seen ab affect GPP and ensuing disequilibria that anomalies in discrimination defined below have the potential to affect interannual variability in Flan chap13 2004 8 24 17 26 page 219 5 13 Terrestrial Ecosystems and Interannual Variability of 13 CO2 and 12 CO2 220 Robin hood diagrams 0 Corresponding algebra A 1 Conventional form net flux approximation d aCa eb eo Nb a ab Rb a a f Ff No a ao Go a a dt 4 5 6 2 3 2 50 100 Pg C yr 1 150 1 Observations 2 Ocean Sink 3 Ocean Diseq 4 Land Sink 5 Land Diseq 6 Fossil fuels 1 0 1 2 6 4 5 3 0 4 0 B 5 1 50 7 500 d C Nb Ff No dt 4 6 2 3 6 100 6 7 2 3 7 150 8 General form with gross terrestrial