Chapter 9

9.1 Ten Years of GCTE Research:Apprehending Complexity

The general assumption, in the first phase of elevated carbondioxide (CO2) research, was that increasing atmosphericCO2 concentration would act as a fertilizer for plant systems.Under this “large-plant-scenario” (Fig. 9.1), increased photo-synthesis at elevated CO2 would lead to higher plant biom-ass which would induce a negative atmospheric feedbackloop that would offset anthropogenic carbon (C) emissions.

A second general assumption in early research was thatthe effects of elevated CO2 on plant community structurecould be predicated using a C-based perspective with clear“winners” and “losers” based on species’ ability to take ad-vantage of the increased atmospheric CO2 concentration(e.g., plant with C3 vs. C4 photosynthetic pathways). Thereality unveiled by a decade of research proved more com-plex and the “large plant scenario” largely untenable.

9.1.1 Effects of CO2 on Plant Diversity ThroughAlterations of the Physical Environment

A large number of studies on plants growing in pots,monoliths and in situ, now yield a broad perspective of

responses to elevated CO2 (e.g., Bazzaz 1990; Poorter 1993;Körner 2001; Navas 1998; Woodward 2002; Poorter andNavas 2003). Reviews have used meta-analysis to iden-tify statistically significant trends (Curtis and Wang 1998;Wand et al. 1999; Kerstiens 2001). Pot studies have gen-erally confirmed the C-based perspective of the differ-ential response of functional groups of plant species toelevated CO2: plants with the C3 photosynthetic pathwaygenerally respond more that C4 plants (due to greaterstimulation of net C fixation), fast-growing plants morethan slow-growing plants (due to stronger C sinks), andN-fixing plants more than non-fixers (because the highC cost of N fixation is overcome). However, an impor-tant revelation of in situ experiments, both within andoutside of the GCTE network, is that this C-based per-spective gives an incomplete understanding of how plantsrespond to elevated CO2 in multi-species communities.Apparently C-based functional groupings of plant re-sponse to CO2 break down (Poorter and Navas 2003)while the effects of CO2 on ecosystem water balance andnutrient cycling often play a larger role on plant com-munity structure than previously suspected (Fig. 9.1).

Elevated CO2 can modify soil nutrient availability,nutrient uptake capacity, and nutrient use efficiency(Bassirirad et al. 1996; Hungate 1999; Körner 2001) all ofwhich are likely to modify plant community structure in

Plant Biodiversity and Responses to Elevated Carbon Dioxide

Catherine Potvin · F. Stuart Chapin III · Andrew Gonzalez · Paul Leadley · Peter Reich · Jacques Roy

Fig. 9.1.Simple cartoon illustrating theearly and current views of theeffect of elevated CO2 on plantecosystems. Current knowl-edge suggest that elevated CO2is likely to affect, minimally,photosynthetic carbon uptake,plant growth and allocation,N and water uptake leading todifferential above- and below-ground competition. Moreoverresponse to elevated CO2 is notconsistent within functionalgroups. Here FG1 and FG2 referto two hypothetic functionalgroups

126 CHAPTER 9 · Plant Biodiversity and Responses to Elevated Carbon Dioxide

nutrient limited ecosystems. Yet, the effects of CO2 onplant community structure acting indirectly throughnutrient availability are poorly studied. On the otherhand, several studies indicate that soil water status is of-ten key in determining species-specific responses to el-evated CO2 in situ (e.g., Owensby et al. 1993; Volk et al.2000; Morgan et al. 2004), although there are also a num-ber of studies where this is not the case (e.g., Arp et al.1993; Zavaleta et al. 2003a,b; Nowak et al. 2004). The ef-fects of CO2 on soil water balance may help explain whymany of the expected direct CO2 effects on diversity, suchas the encroachment of C3 woody plants into C4 savannas(Walker 2001) or the increase in competitive ability ofC3 species compared to C4 species in grasslands, are notconsistently observed (Morgan et al. 2004). Species-spe-cific responses in calcareous grasslands may also largelydepend on CO2 effects on water balance. In study of CO2effects on calcareous grassland, Leadley et al. (1999) foundthat Carex flacca had a greater growth response to el-evated CO2 than the dominant grass Bromus erectus. Volket al. (2000) provided compelling evidence that the largedifferences in response of Carex and Bromus to elevatedCO2 may be explained by the effects of elevated CO2 onsoil water content. More generally, Stocklin and Körner(1999) identified the functional group of mesophytic spe-cies as the most responsive to increased CO2 in mono-liths of these calcareous grassland communities. Appar-ently, reductions in stomatal conductance at elevated CO2that lead to modifications in soil water status can play akey role altering plant community structure.

Climate apparently also modulates community re-sponses to elevated CO2. Roy and collaborators (unpub-lished) compared the CO2 response of Mediterranean

annual herbaceous species over several growing seasons.Their experiment documented large inter-annual vari-ability in the response of individual species to CO2. Theresponse of most annual species to elevated CO2 switchedfrom being highly negative in some years to being highlypositive in other years. For example, in the communitycomposed of 2 grasses, 2 composites and 2 legumes, thebiomass response of the annual grass Bromus madritensis(Fig. 9.2) changed from a 70% reduction to a 300% in-crease depending on the years. Because the planting wasrepeated similarly every year for 5 years in greenhousestracking outside temperature and precipitation, the re-sults suggest that climatic conditions might drive theoutcome of this elevated CO2 experiment. Similarly, thepresence of an interaction between climatic variabilityand response to CO2 for perennial herbaceous commu-nities was reported by Owensby et al. (1996) for a Kansastallgrass prairie. On wet years, there was no significant bio-mass response to CO2, while on dry years, the C4 speciesAndropogon gerardii was stimulated and drove a posi-tive community biomass response.

Roy (unpublished) further noted that CO2 interactedwith the diversity level of the community: year to yearvariability was less in the communities with the highernumber of species. Several experiments both in the fieldand in microcosms suggest that increasing diversity in-creases the stability of ecosystem properties (Hooperet al. 2005). Tilman and Downing (1994) reported forexample that, in the Minnesota grassland, plots with ahigh diversity resisted drought better then plots with lowdiversity. An issue that remains largely unanswered isthe extent to which elevated CO2 and diversity would in-teract to either increase or decrease the stability of eco-

Fig. 9.2.CO2 response of the annualgrass Bromus madritensis (to-tal aboveground biomass atsenescence) when grown for5 successive years in communi-ties of various complexity es-tablished by planting youngseedlings of annual or perenialgrasses, composites and le-gumes in a predefined pattern(annual grasses: G, compos-ites: C, legumes: L, perennialgrasses: Gp). Each bar is theaverage of one to four 0.5 m2

model ecosystems at each CO2level, with 360 to 30 individu-als of B. madritensis per modelecosystem depending on diver-sity level. Two greenhouseswere run at 350 µmol CO2 mol–1

and two at 700 µmol CO2 mol–1.Average annual temperatureand relative humidity in thegreenhouses for the five yearsof the experiment are given inthe small panels above themain figure

127

system properties. Clearly responses to elevated CO2 aremodulated by climate, water and nutrient availability withdifferent feedback loops:

� Changes in atmospheric CO2 alter the relative avail-ability of resources (C, water, and nutrients), which inturn alters competitive interactions among plantspotentially leading to changes in diversity or domi-nance patterns.

� Environmental variability affects species-specific re-sponses to elevated CO2 while community diversityand composition in turn influences both the inter-annual variability and the direction of the responseto elevated CO2.

9.2 Temporal Variation and Responseto Elevated CO2

9.2.1 Reproductive and Evolutionary Aspectsof the Response to Elevated CO2

Adding to complexity, researchers recognized that theresponse of vegetative biomass to elevated CO2 is notsufficient to understand long-term community responses.CO2 induced differences in number and quality of seedsshould have strong impact on the diversity of plant com-munities over the long-term. In a meta-analysis cover-ing 79 species, Jablonski et al. (2002) reported a stimula-tion of the number of reproductive organs or their weightby about 20%. Crop species were more responsive (+25%)than wild species (+4%) despite a similar stimulation ofvegetative biomass. Variability is very high between gen-era (–26% in Ipomea sp to +30% in Brassica sp.) as wellas within genera (from +2% to +60% in Raphanus sp.).Individual seed size was marginally increased but seednitrogen content was reduced by about 14% in both wildspecies and crops. In a single experiment in the field with17 species, Thürig et al. (2003) found a significant increasein the number of flowering shoots (+24%) and seeds(+29%) with very large responses in some species (flow-ering shoot +236% in Trifolium pratense and +1 000% inCarex flacca). In perhaps the first field experiment tocompare tree species reproductive output, Stiling et al.(2004) found a four-fold increase in acorn density fortwo oak species and no change in the third one.

Furthermore, rising atmospheric CO2 concentrationsmay alter the genetic structure of plant populationsthrough directional selection; i.e., within a population,those individuals that have the greatest response to CO2in terms of growth and reproduction will tend to excludethose individuals of the same species with weaker re-sponses. If such evolutionary responses occur responsesof plant species observed in short-term experiments maynot be representative of the response of plants to long-term, slowly rising atmospheric CO2 concentrations. Sev-

eral studies have demonstrated genetic variation in re-sponses to CO2 concentration (Wang et al. 1994; Stewartand Potvin 1996; Schmid et al. 1996; Roumet et al. 2002;Spinnler et al. 2003; Mohan et al. 2004; Bidart-Bouzat2004; Ward and Kelly 2004) suggesting that elevated CO2may drive rapid directional selection. However only afew studies experimentally tested the evolutionary re-sponses to controlled selection under various CO2 levels.Brassicae juncea did not show genetic adaptation to acombined CO2 and temperature increase, possibly dueto inbreeding depression in a stressful environment(Potvin and Toussignant 1996). Derner et al. (2004) foundthat the increase in CO2 responsiveness from the first tothe third generation of wheat was not driven by geneticchanges but by CO2-induced changes in seed quality.Andalo et al. (1996) reported a decrease in seed germi-nation of Arabidopsis thaliana when grown in two suc-cessive generations under elevated CO2. Ward et al.(2000), also on A. thaliana, is the only report of a selec-tive effect of CO2 but genetic adaptation was found onlyunder selection at low Pleistocene-like CO2 concentra-tion. Evolutionary responses at elevated CO2 resulted ina shortened life cycle that did not allow for increased bio-mass accumulation. It appears that:

� Variations in the response of growth, allocation, phe-nology, or ontogeny (Kinugasa et al. 2003; Ward et al.2000; LaDeau and Clark 2001) can explain the ob-served large interspecific differences in reproductiveresponses.

� In principle, it is critical to incorporate the effects ofevolutionary processes in our predictions of ecosystemsresponses to global change (Ward and Kelly 2004). How-ever, too few studies exist to give us a clear idea of theevolutionary potential in the face of elevated CO2.

9.2.2 Communities at Equilibrium VersusDynamic Systems

Community dynamics also has to be taken into accountto determine whole systems response to elevated CO2.Changes in the relative contribution of individual spe-cies under elevated CO2 could alter community produc-tion with consequent changes in community composi-tion (Körner 2001; Potvin and Vasseur 1997; Reynoldset al. 2001; Chapin et al. 2000). Several experiments havefocused on understanding the impact of elevated CO2on community composition. We use three long-term CO2enrichment studies conducted on grasslands to illustratethis point: the calcaeous grassland experiment of Swit-zerland (Leadley et al. 1999; Niklaus and Körner 2004),the grazed pasture of Southern Québec, Canada (Potvinand Vasseur 1997; Vasseur and Potvin 1997) and the an-nual grassland of the Jasper Ridge Biological Preserve inCalifornia (Zavaleta et al. 2003a,b). Floristically the rich-

9.2 · Temporal Variation and Response to Elevated CO2

128 CHAPTER 9 · Plant Biodiversity and Responses to Elevated Carbon Dioxide

est of those grassland sites was the Swiss grassland witha total of 70 vascular species (Leadley et al. 1999), followedby the Californian site with a total of 43 species (Zavaletaet al. 2003a) and then the Québec site about 10 species(Potvin and Vasseur 1997). CO2 enrichment lasted sixyears in Switzerland, five years in Québec while Zavaletaet al. (2003a,b) reported three years of CO2 enrichment.These three experiments provide a unique opportunityto examine the extent to which elevated CO2 might alterspecies composition with a possible feed-back loop oncommunity biomass accumulation.

In the Swiss grassland, the number of species presentin the plots increased through time, 1996 vs. 1999, butexposure to elevated CO2 had no significant effect onspecies richness. In contrast elevated CO2 significantlyenhanced evenness estimated as the modified Hillratio E* (Niklaus and Körner 2004). At the same time,community productivity increased under elevated CO2 andthe increase in aboveground biomass could be predictedby precipitation. The highest biomass enhancement wasobserved after five years of enrichment (+31%) and theeffect seemed to decrease thereafter (+18% enhancementon year 6). At the Québec field site, between 1992 and 1996,total species number was reduced by 10% under elevatedCO2, 47% in the ambient open-tops and 31% in the controlplots (Potvin unpublished). Through time evenness, mea-sured as Simpson’s index, decreased under ambient CO2while it increased under elevated CO2 (Potvin and Vasseur1997). Destructive harvest, carried-out at the end of thefive years of CO2 enrichment, showed that biomass accu-mulation was not statistically different between the high(360 g m–2 ±125) and the ambient (486 g m–2 ±152) open-tops nor the control plots (435 g m–2 ±118) (Potvin un-published). Three years of elevated CO2 over Californiagrassland led to a reduction in species number of 8%(Zavaleta et al. 2003a) however no information was givenon evenness. The Californian study examined the effectof N deposition, precipitation and warming on diversity.Interestingly the effect of these different global changedrivers on diversity was purely additive. The two enrich-ments (N and CO2) reduced diversity, while precipitationincreased diversity and warming had no effect. Norby et al.(this volume) provides an exhaustive analysis of the re-sults from the Jasper Ridge Global Change Experiment(JRGCE). In the JRGCE, elevated CO2 led to increased plantbiomass when acting alone but to a reduction in NPP whenapplied in conjunction with other global change drivers:temperature, N deposition and precipitation.

Together these results suggest that elevated CO2 willelicit changes in biodiversity, whether negative or posi-tive, and that these changes may or may not be associ-ated with changes in community biomass. This observa-tion is supported by other studies. An experiment com-bining elevated CO2 with sward management of temper-ate grassland monoliths (Teyssonneyre et al. 2002)showed that cutting frequency modified the biodiversi-

ty response of the monoliths. At low cutting frequencythe decline in species diversity was partly alleviated byelevated CO2, a results with much resemblance to theQuébec experiment. Furthermore, elevated CO2 may in-teract with a number of other components of globalchange including N deposition, changes in precipitation,and temperature. CO2 is thought to stimulate produc-tion and C storage primarily in areas of nitrogen depo-sition (Schimel 1995; Nadelhoffer et al. 1999). In thesenutrient-rich environments, enhanced productivity isoften associated with reduced species diversity (Tilmanand Downing 1994). However, in the JRGCE, Zavelettaet al. (2003a,b) found no interactions between CO2, N,precipitation and temperature on plant diversity. Thediversity responses of any combination of treatmentscould be predicted from the sum of the responses to in-dividual treatments. Additive responses of this type wouldmake prediction of CO2 responses in the future much moretractable, but additional multifactor experiments will beessential for testing the generality of this pattern.

The three field experiments discussed above differfundamentally from those using growth chambers, inwhich a fixed set of species is exposed to elevated CO2,because the communities under study were moving tar-gets where species richness, identity and dominance were

Fig. 9.3. Predicted changes in community biomass due to changesin dominance and identity of dominant species for the Québec ex-periment. The model assumes an arbitrary time scale leading tochanges in dominance in the community. Predicted biomass wascalculated as PB = Saba + Sb(bi + bj + bk) where Sa is the area occu-pied by the dominant species, Sb is the area occupied by each of theother species, ba is the biomass per area of the dominant speciesand bi–k the biomass per area of the three companion species. Attime 1, community biomass was modeled for even stands whereSimpson’s index, D = 4. Dominance increased through time to amaximum, D = 1.53, when 80% of the area was occupied by a singlespecies. The model was run four times with either Phleum pratense,Plantago major, Taraxacum officinale, or Poa pratensis as dominantspecies. The scenario in which Poa becomes dominant is analog toobservations from the Farnham elevated CO2 experiments

129

constantly changing. As such they might provide a morerealistic appraisal of the effect of elevated CO2. The studyof dynamic communities raises the possibility that bio-mass response to elevated CO2 could be hard to detect inearly successional sites if large changes in biomass asso-ciated with a shift in species composition are taking place.A simple model was built to examine the effect of shiftsin dominance of the four most common species on com-munity biomass at the Québec site. The simulation showsthat both changes in dominance pattern and in the iden-tity of the dominant species strongly affect communitybiomass (Fig. 9.3). Community biomass decreased by47%, when Taraxacum became dominant, while it in-creased by 23%, when Phleum became dominant. Biom-ass in the high CO2 chambers was 25% lower then that ofthe ambient CO2 chambers after five years of enrichment.The simulation shows that succession-mediated changesin biomass could either mask or amplify changes in bio-mass due to CO2 enrichment. Leadley et al. (1999) re-ported vast differences in responsiveness to elevated CO2between Lotus corniculatus, Bromus erectus and Carexflacca and the other species during the first three yearsof results for the Swiss grassland experiment. In fact, theauthors explain the observed lag in the response of above-ground biomass by the growth rate of these species; im-plicitly suggesting that successional pattern is importantfor an understanding of biomass responses to CO2. Bolkeret al. (1995), using an individual-based forest model, simi-larly reported that changes caused by successional se-quence were larger than direct CO2 responses. In theirsimulation, they identify red oak as a species drivingmany of the patterns observed. Two years of CO2 enrich-ment in the Nevada Desert FACE Facility (NDFF) re-ported a 42% reduction in density for the native annualspecies coupled with a 40% increase in abovegroundbiomass. Bromus madritensis, an invasive exotic species,became a higher proportion of total plant density at el-evated CO2 (Smith et al. 2000). Those changes in commu-nity biomass associated with a shift in composition illus-trate the challenge facing ecologists in predicting the ef-fect of elevated CO2 on dynamic communities (Díaz 2001).

An important legacy of the last ten years of elevatedCO2 research is the recognition that natural communi-ties must been studied as dynamic systems. Studies ofthe dynamic responses of communities to elevated CO2indicate that:

� Elevated CO2 might elicit changes in community di-versity both in term of species richness or evenness.In dynamic communities, species identity and domi-nance patterns can drive the response to CO2.

� Community biomass can change regardless of the ef-fect of CO2 due to succession alone. As both effectscan be of similar magnitude, it may be difficult to par-tition the effect of CO2 and succession on transientcommunities.

� Although the effects of elevated CO2 on biodiversitymay be less pronounced than those of other globalchange drivers such as land use, climate, N deposition,and biotic exchange, elevated CO2 will likely interactwith these drivers in ways that differ among biomes(Sala et al. 2000).

9.3 Biodiversity Loss and Response to Elevated CO2

9.3.1 Species Diversity and Response to Elevated CO2

The studies reported in the previous section suggest thatelevated CO2 has the potential, at least under certain cir-cumstances, to alter biodiversity. The interaction betweenCO2 and diversity must therefore be examined in thecontext of the diversity-productivity relationship. TheGCTE decade has witnessed an enormous effort by ecolo-gists to try and disentangle the links existing betweendiversity and ecosystem functioning (e.g., Chapin et al.2000; Tilman et al. 2001; Spehn et al. 2005). As pointedout by Hooper et al. (2005) biodiversity has often beenconfounded with species richness although other com-ponents of biodiversity such as evenness, compositionand presence/absence of key species all exert strong ef-fects on ecosystem properties. Loreau and Hector (2001)suggested that a diversity effect might be due to eithersampling or complementarity. Sampling refers to the like-lihood that a key species be present in a community whilecomplementarity describes interactions among species.Research on elevated CO2 suggests that the response toelevated CO2 is largely mediated by species identity andcompetitive interactions (see Körner et al. this volum;,Hanley et al. 2004; Navas et al. 2002; Ramsier et al. 2005).It is thus likely that elevated CO2 could modify the diver-sity/productivity relationship. Three projects explicitlyaddressed this question by manipulating CO2 concen-tration and plant diversity: the Swiss calcareous grass-land study (Stocker et al. 1999; Niklaus et al. 2001), Roy’sannual herbaceous communities (see Sect. 9.2.1) and theMinnesota, USA, BioCON project (Reich et al. 2001a,b,2004) (Fig. 9.4).

To answer the question: “How does species loss influ-ence biomass and its response to elevated CO2?” theBioCON project relied on plots planted with 16, 9, 4 and1 species while the Swiss experiment involved plots of 31,15 and 5 species. In the BioCON project, total biomass ofplots with 9, 4, and 1 species was, respectively 2, 13, and50% less than that of plots with 16 species (Reich et al.2001a,b; Reich et al. 2004). The enhancement of biomassresulting from elevated CO2 decreased with decliningdiversity. In ambient plots, the average stimulation oftotal biomass in response to elevated CO2 was 22%, 18%,10% and 7% in, respectively, the 16-, 9-, 4- and 1-speciesplots (Reich et al. 2001a,b). Differences in diversity ac-counted for a five-fold difference in the effect of CO2 en-

9.3 · Biodiversity Loss and Response to Elevated CO2

130 CHAPTER 9 · Plant Biodiversity and Responses to Elevated Carbon Dioxide

richment Whereas elevated CO2 increased biomass by258 g m–2 in the most diverse plots, it increased biomassby only 47 g m–2 in the monocultures. Likewise, the Swissexperiment found a significant effect of diversity on bio-mass; this effect, however, declined with time (Niklaus et al.2001). Elevated CO2 had a significant but highly variableeffect on the biomass of high-diversity plots but no sucheffect on the medium and low-diversity communities. Con-versely, Roy and collaborators found a significant increasein plant biomass production at elevated CO2 in Mediterra-nean grassland communities, but the addition of specieswithin functional groups did not alter this response.

Two complementary BioCON experiments allow fur-ther partitioning of the effect of functional groups andspecies within functional groups on CO2 responses. Theseexperiments include plots with (i) 1 or 4 species fromsingle functional groups (C4 grasses, C3 grasses, N-fix-ing legumes and non-N-fixing herbaceous species) and(ii) four species varying from 4 to 1 in functional groups.All these plots were replicated across four CO2 andN levels. Biomass increment in response to elevated CO2,enriched N, or their combination was greater in 4-spe-cies than in 1-species plots (Fig. 9.5). Furthermore, plotsplanted with each forb species grown in monoculture had,on average, 44% less total biomass than plots with allfour forbs species. Across all four functional groups,single functional group plots with one instead of fourspecies had 29% less biomass on average (Reich et al.2004). Decreasing the number of groups from 3 or 4 to 1or 2 resulted in 20% lower total biomass on average, evenwith species richness constant at 4. Total biomass re-sponses to CO2 and N reflect effects on fine root biomass(Fig. 9.5). For instance, on average, in plots with four func-tional groups, elevated CO2 increased fine root biomassby 191 g m–2, but as the number of groups decreasedto 3, 2, and 1, elevated CO2 increased fine root biomassby 91, 90, and only 14 g m–2, respectively.

In the BioCON project, functional group diversity wasnot a prerequisite for species richness to have significanteffects and this effect was seen in disparate functionalgroups. This contrasts with the Mediterranean grasslandstudy that did not find evidence that reductions in thenumber of functional groups altered the growth responseto elevated CO2 In the Swiss experiment, species respond-ing positively to CO2 enrichment did not follow commonfunctional type classification. Thus while elevated CO2might well influence community composition, it is diffi-cult to predict changes. Together, the BioCON, Swissgrassland, and Mediterranean grassland experimentsprovide insights on how a species poor world would re-spond to elevated CO2:

� Decreases in diversity can lead to lower total plantbiomass and to reduced responsiveness to elevatedCO2. This diversity effect is highly variable rangingfrom very pronounced (BioCON), relatively small(Swiss grassland) to nil (Mediterranean grassland).

� In the BioCON project, holding species richness con-stant, plots with decreasing functional group richnesshad less biomass and smaller enhancements of biom-ass in response to CO2, when compared with plots withgreater functional group richness. This was not thecase in the Mediterranean grassland study.

9.3.2 Ecosystem C Fluxes in a Species-Poor World

The previous section suggests that species-poor commu-nities could accumulate less biomass and be less respon-sive to elevated CO2 with the magnitude of responsesmodulated by resource availability. This raises a ques-tion: Would a species-poor world become a less efficientC sink? If this were the case, the initial assumption ofelevated CO2 research, coined earlier as “the large-plant-

Fig. 9.4. Effects of species richness and functional group richness on biomass accumulation and response to elevated CO2. a Total biom-ass and b change in total biomass (above- and below-ground, 0–20 cm depth) (+1 S.E.) for plots planted with one functional group andeither 1 or 4 species, grown at four combinations of ambient (368 µmol mol–1) and elevated (560 µmol mol–1) CO2, and ambient or enriched N(4 g N m–2 yr–1). Data averaged over two harvests in each year from 1998 to 2001. c Change in total biomass and d in fine root biomass in plotswith 4 species, and either 1, 2, 3 or 4 functional groups, grown at four combinations of ambient and elevated CO2, and ambient and enriched N

131

scenario”, would be invalidated. In fact, we know littleabout the effects of species loss on the capacity of eco-systems to take up C in the long term. So, where does thefuture of research on elevated CO2 and diversity lie?

First, if we want to understand the consequences ofspecies loss on the ability of ecosystem to sequester C,we need to develop better ways to estimate C cycling.Although it is widely acknowledged that C sequestrationand net primary production (NPP) are not equivalent(Körner 1995; Karnosky 2003) most experiments exam-ining the diversity/productivity relationship have usedeither NPP or biomass accumulation to report on eco-system function. Catovsky et al. (2002) emphasize theurgent need to move beyond estimates of NPP towardbetter estimations of net ecosystem exchange (NEP). Itis important to understand the diversity effects on plantbiomass, on litter decomposition as well as on soil respi-ration (both autotrophic and heterotrophic) in order toestimate appropriately C fluxes in and out of the system.

In the only published measurements of the effects of plantdiversity on NEP, Stocker et al. (1999) found that reduc-tions in plant diversity did not alter whole ecosystem CO2fluxes in calcareous grasslands at elevated or ambient CO2concentrations. At the ecosystem level, Chambers et al.(2004) modeled the response to elevated CO2 of tree bio-mass and wood litter for a Central Amazon forest. Animportant conclusion of their study is that coarse litteris important in determining the forest carbon balanceand should be considered as a C flux. They argue thatunder elevated CO2, the gains in carbon due to increasedgrowth would roughly balanced by the increased stockof coarse litter. Using microcosms, Van Voris et al. (1980)proposed that the complexity of CO2 efflux time series,i.e., the number of cyclic components required to com-pose a time series, might provide useful information re-garding ecosystem C cycling. In laboratory microcosmsof an old-field ecosystem, they demonstrated that thecomplexity of CO2 flux time series was negatively related

Fig. 9.5. Independent effects of species richness (panels a and b) and functional group richness (panels c and d) on biomass accumula-tion and response to elevated CO2. a Total biomass (aboveground plus belowground, 0–20 cm depth)(+1 S.E.) for plots planted with onefunctional group (F = 1) and either 1 or 4 species, grown at four combinations of ambient (368 µmol mol–1) and elevated (560 µmol mol–

1) concentrations of CO2, and ambient N and enriched N (4 g N m–2 yr–1). Data averaged over two harvests in each year from 1998 to 2001.b For the same plots, the change in total biomass (compared with ambient levels of both CO2 and N) in response to elevated CO2 alone (atambient N), to enriched N alone (at ambient CO2), and to the combination of elevated CO2 and enriched N, pooled across years. c Changein total biomass in response to elevated CO2 or enriched N alone, and to the combination of elevated CO2 and enriched N, for plots with4 species, and either 1, 2, 3 or 4 functional groups, grown at four combinations of ambient and elevated CO2, and ambient and enriched N.d Change in fine root biomass in response to elevated CO2 or enriched N alone, and to the combination of elevated CO2 and enriched N, ineach year, for same plots as in C

9.3 · Biodiversity Loss and Response to Elevated CO2

132 CHAPTER 9 · Plant Biodiversity and Responses to Elevated Carbon Dioxide

to the amount of calcium lost in the leachate, an indica-tor of system disturbance in their experiment. This re-sult suggests that the analysis of CO2 fluxes may be a use-ful integrative measure of the effects of disturbance (e.g.,species loss) on the C cycle.

Second, we need to better understand the impact oncommunity disassembly on C cycling. Habitat destruc-tion and climate change are and will continue to be themajor drivers of species loss and subsequent ecosystemresponses (e.g., Lawton and May 1995; Thomas et al.2004a). For example, using an area-based extrapolationof climate envelopes, Thomas et al. (2004b) suggested thatclimate change might lead to up to 35% of the taxa being“committed to extinction” in their study regions over thenext 50 years. To date, experiments addressing the func-tional consequences of species loss have typically beenperformed over small spatial and temporal scales rela-tive to the size, generation times and possible extinctionrates of the organisms involved (e.g., Sect. 9.3.1, Tilmanet al. 1997; Hector et al. 1999; Lepš et al. 2001). Becauseextinction in response to habitat destruction and climatechange is a non-random process dominated by non-equi-librium population dynamics (Kareiva and Wennergren1995), we need to better understand the long-term effectsof community disassembly.

Community disassembly and its affects on C uptakeand on the response to elevated CO2 have received rela-tively little attention (Lawton 1994; Sala et al. 1996; Grime1998; Gonzalez and Chaneton 2002). For example, theidentity of species extinctions (e.g., rare vs. dominant;Grime 1998) and the rate at which such extinctions oc-cur could generate non-linear ecosystem responses(Lamont 1995). Furthermore, long-term temporalchanges in the abundances and biomass of remnantpopulations (e.g., Laurance et al. 1997; Didham et al. 1998)have been largely missing from biodiversity-C cyclingstudies conducted to date. The first Ecotron experiment(Naeem et al. 1994) reported that higher diversity com-munities consumed more CO2 than lower diversity com-munities. It was suggested, however, that the experimen-tal design could not separate diversity and compositioneffects (Huston 1997), but an additional shortcoming wasthe limited duration of the experiment. McGrady-Steedet al. (1997) studied microbial communities that exhib-ited greater inter-community variation in CO2 flux at lowdiversity than at high diversity. However, this effectseemed to be due to differences in species compositionamong low diversity communities (Morin and McGrady-Steed 2004), and not to diversity effects per se.

The results obtained to date do not demonstratewhether a species-poor world will become a less efficientC sink or not. We might however be able to flip the ques-tion around and propose management strategies to in-crease C storage at the landscape level. GCTE is comingto an end at the same time as the Kyoto Protocol is com-ing into effect. The Nations of the world are now seeking

ways to promote the uptake of C through land-usechanges while ensuring the stability of these sinks.Hooper et al. (2005) summarized our knowledge to dateby indicating that community composition is at least asimportant as species richness to determine ecosystemproperties. Therefore management strategies that incor-porate biodiversity will have to pay attention to the iden-tity of species and their interactions (Hiremath and Ewel2001). Furthermore estimates of the potential forC sequestration need to incorporate a precise assessmentnot only of C stocks but also of C losses. For example,C accounting methodologies proposed for afforestationand reforestation projects under the Clean DevelopmentMechanism (www.unfccc.com) need to inventory coarsewoody debris as well as living biomass.

Many questions remain for the next generation of glo-bal change research:

� How is the relationship between net primary produc-tivity and net ecosystem productivity affected by bio-diversity, elevated CO2 and other global change driv-ers?

� Can we predict which species are most likely to belost by habitat fragmentation or climate change?

� How will the process of species loss, driven by habitatfragmentation and climate change, affect CO2 fluxesand the response of ecosystems to elevated CO2?

� In what ways, if any, will future species loss alter theC source and sink capacities of different ecosystems?

9.4 Conclusions

In situ studies of communities suggest that the interac-tions between biodiversity, atmospheric CO2 concentra-tion, and C cycling are very complex. Rising CO2 willprobably act on plant diversity through a number of veryindirect pathways; e.g., by altering the relative avail-ability of resources such as water and nutrients therebyaltering competitive interactions among plants. Second,changes in plant community structure due to rising CO2concentrations, particularly in highly dynamic systems,may be as or more important in determining bio-mass responses than the direct effect of elevated CO2.Finally, species loss resulting from global change mightreduce productivity as well as responsiveness to elevatedCO2, although the response will largely depend on spe-cies identity.

Over all, simple scenarios invoking a clear and simplebiospheric loop between atmospheric CO2 and ter-restrial biomass due to CO2 fertilization need to be aban-doned and replaced by an understanding of a complexsystem involving both biotic and abiotic feed-back loops.Our ability to either manage or predict the relation-ship between diversity and ecosystem function is stilllimited.

133

References

Andalo C, Godelle B, Lefranc M, Mousseau M-A, Hill-Bottraud I(1996) Elevated CO2 decreases seed germination in Arabidopsisthaliana. Glob Ch Biol 2:129–135

Arp WJ, Drake BG, Pockman WT, Curtis PS, Whigham DF (1993)Interactions between C3 and C4 saltmarsh plant species duringfour years of exposure to elevated CO2. Vegetatio 104/105:133–144

Bassirirad H, Thomas RB, Reynolds JF, Strain BR (1996) Differen-tial responses of root uptake kinetics of NH4

+ and NO3– to en-

riched atmospheric CO2 concentration in field-grown loblollypine. Plant, Cell and Environment 19:367–371

Bazzaz FA (1990) The response of natural ecosystems to rising glo-bal CO2 levels. Ann Rev Ecol Syst 21:167–196

Bidart-Bouzat MG (2004) Herbivory modifies the lifetime fitnessresponse of Arabidopsis thaliana to elevated CO2. Ecology85(2):297–303

Bolker BM, Pacala SW, Bazzaz FA, Canham CD, Levin SA (1995) Spe-cies-diversity and ecosystem response to carbon dioxide fertili-zation – Conclusions from a temperate forest model. Glob ChBiol 1(5):373–381

Catovsky S, Bradford MA, Hector A (2002) Biodiversity and ecosys-tem productivity: implications for carbon storage. Oikos97(3):443–448

Chambers JQ, Higuchi N, Texeira LM, dos Santos J, Laurance SG,Trumbore SE (2004) Response of tree biomass and wood litterto disturbance in Central Amazon forest. Oecologia 141(40):596–611

Chapin III FS, Zavaleta ES, Eviner VT, Naylor RL, Vitousek PM,Reynolds HL, Hooper DU, Lavorel S, Sala OE, Hobie SE, MackMC, Díaz S (2000). Consequences of changing biodiversity.Nature 405 (11):234–242

Curtis PS, Wang X (1998) A meta-analysis of elevated CO2 effects onwoody plant mass, form, and physiology. Oecologia 113:299–313

Derner JD, Tischler CR, Polley HW, Johnson HB (2004) Intergenera-tional above- and belowground responses of spring wheat (Triti-cum aestivum L.) to elevated CO2. Basic Appl Ecol 5(2):145–152

Díaz S (2001) Complex interactions between plant diversity, suc-cession and elevated CO2. TREE 16(12) 667

Didham RK, Hammond PM, Lawton JH, Eggleton P, Stork, NE (1998)Beetle species responses to tropical forest fragmentation. EcolMonog 68:295–323

Gonzalez A, Chaneton E (2002) Heterotroph species extinction,abundance and biomass dynamics in a fragmented microeco-system. J Anim Ecol 71:594–602

Grime JP (1998) Benefits of plant diversity to ecosystems: immedi-ate, filter and founder effects. J Ecol 86:902–910

Hanley ME, Tropimov S, Taylor G (2004) Species-level effects moreimportant than functional group-level responses to elevated CO2:evidence from simulated curves. Functional ecology 18:304–313

Hector A, Schmid B, Beierkuhnlein C (1999) Plant diversity andproductivity experiments in European grasslands. Science 286:1123–1127

Hiremath AJ, Ewel JJ (2001) Ecosystem nutrient use efficiency, pro-ductivity, and nutrient accrual in model tropical communitiesEcosystems 4:669–682

Hooper DU, Chapin III FS, Ewel JJ, Hector A, Inchausti P, Lavorel S,Lawton JH, Lodge DM, Loreau M, Naeem S, Schmid B, Setala H,Symstad AJ, Vandermeer J, Wardle DA (2005) Effects of biodi-versity on ecosystem functionaing: a consensus of currentknowledge. Ecological Monographs 7591) 3–35

Hungate BA (1999) Ecosystem responses to rising atmospheric CO2:feedbacks through the nitrogen cycle. In: Luo Y, Mooney HA,eds. Carbon Dioxide and Environmental Stress. San Diego, USA:Academic Press, 265–285

Huston MA (1997) Hidden treatments in ecological experiments:re-evaluating the ecosystem function of biodiversity. Oecologia110:449–460

Jablonski LM, Wang XZ, Curtis PS (2002) Plant reproduction underelevated CO2 conditions: a meta-analysis of reports on 79 cropand wild species. New Phytol 156(1):9–26

Kareiva P, Wennergren U (1995) Connecting landscape patterns toecosystem and population processes. Nature 373:299–302

Karnosky DF (2003) Impacts of elevated atmospheric CO2 on for-est trees and forest ecosystems: Knowledge gaps. Environmentinternational 29:161–169

Kerstiens G (2001) Meta-analysis of the interaction between shade-tolerance, light environment and growth response of woodyspecies to elevated CO2. Act Oecol 22:61–69

Kinugasa T, Hikosaka K, Hirose T (2003) Reproductive allocationof an annual, Xanthium canadense, at an elevated carbon diox-ide concentration. Oecologia 137(1):1–9

Körner Ch (1995) Toward a better experimental basis for upscalingplant responses to elevated CO2 and climate warming. Plant Celland Environment 18:1101–1110

Körner C (2001) Biosphere responses to CO2 enrichment. Ecol Appl10:1590–1619

Körner C, Morgan J, Norby R (2007) CO2 fertilization: when, where,and how much. In: Canadell J, Pataki D, Pitelka L (eds) Globalchange and terrestrial ecosystems. Springer-Verlag, Heidelberg,this volume

LaDeau SL, Clark JS (2001) Rising CO2 levels and the fecundity offorest trees. Science 292(5514):95–98

Lamont BB (1995) Testing the effect of ecosystem composition struc-ture on its functioning. Oikos 74:283–295

Laurance WF, Laurance SG, Ferreira LV, Rankin-de Merona JM,Gascon C, Lovejoy TE (1997) Biomass Collapse in AmazonianForest Fragments. Science 278:1117–1118

Lawton JH (1994) What do species do in ecosystems? Oikos 71:367–374Lawton JH, May R (1995) Extinction Rates. Oxford University Press,

OxfordLeadley PW, Niklaus PA, Stocker R, Körner C (1999) A field study of

the effects of elevated CO2 on plant biomass and communitystructure in a calcareous grassland. Oecologia 118(1):39–49

Lepš J, Brown VK, Díaz Len TA, Gormsen D, Hedlund K, Kailová J,Korthals GW, Mortimer SR, Rodriguez-Barrueco C, Roy J, SantaRegina I, van Dijk C van der Putten WH (2001) Separating thechance effect from other diversity effects in the functioning ofplant communities. Oikos 92:123–134

Loreau M, Hector A (2001) Partitioning selection and complemen-tarity in biodiversity experiments. Nature 412:72–76

McGrady-Steed J, Harris PM, Morin PJ (1997) Biodiversity regulatesecosystem predictability. Nature 390:162–164

Mohan JE, Clark JS, Schlesinger WH (2004) Genetic variation ingermination, growth, and survivorship of red maple in responseto subambient through elevated atmospheric CO2. Glob Ch Biol10(2):233–247

Morgan JA, Mosier AR, Milchunas DG, LeCain DR, Nelson JA, Par-ton WJ (2004) CO2 enhances productivity, alters species com-position, and reduces digestibility of short grass steppe vegeta-tion. Ecol Appl 14:208–219

Morin PJ, McGrady-Steed J (2004) Biodiversity and ecosystem function-ing in aquatic microbial systems: a new analysis of temporal varia-tion and species richness-predictability relations. Oikos 104:458–466

Nadelhoffer KJ, Emmertt BA, Gundersen P, Kjonaas OJ, KoopmansCJ, Schleppi P, Tietema A, Wright RF (1999) Nitrogen depositionmakes a minor contribution to carbon sequestration in temper-ate forests. Nature 398:145–148

Naeem S, Thompson LJ, Lawler SP, Lawton JH, Woodfin RM (1994)Declining biodiversity can alter the performance of ecosystems.Nature 368:734–736

Navas ML (1998) Individual species performance and response ofmulti-specific communities to elevated CO2: a review. Funct Ecol12:721–727

Navas ML, Garnier E, Austin MP, Viaud A, Gifford RM (2002) Seek-ing a sound index of competitive intensity: Application to thestudy of biomass production under elevated CO2 along a nitro-gen gradient. Austral Ecology 27:463–473

Niklaus PA, Körner Ch (2004) Synthesis of a six-year study of cal-careous grassland responses to in situ CO2 enrichment. Eco-logical Monographs 74(3) 491–511

Niklaus PA, Leadley PW, Schmid B, Körner C (2001) A long-termfield study on biodiversity X elevated CO2 interactions in grass-land. Ecol Monog 7:341–356

Nowak RS, Zitzer SF, Babcock D, Smith-Longozo V, Charlet TN, ColemanJS, Seemann JR, Smith SD (2004) Elevated atmospheric CO2 doesnot conserve soil water in the Mojave desert. Ecology 85:93–99

References

134 CHAPTER 9 · Plant Biodiversity and Responses to Elevated Carbon Dioxide

Owensby CE, Coyne PI, Ham JM, Auen LM (1993) Biomass produc-tion in a tallgrass prairie ecosystem exposed to ambient andelevated levels of CO2. Ecol Appl 3:644–653

Owensby C, Ham JM, Knapp AK, Rice CW, Coyne PI, Auen LM (1996)Ecosystem-level responses of tallgrass prairie to elevated CO2.Carbon dioxide and terrestrial ecosystems. GW. Koch and HAMooney. San Diego, Academic Press, pp 147–162

Poorter H (1993) Interspecific variation in the growth response of plantsto an elevated ambient CO2 concentration. Vegetatio 104/105:77–97

Poorter H, et al. (1998) Do slow-growing species and nutrient-stressed plantsrespond relatively strongly to elevated CO2? Glob Ch Biol 4:693–697

Poorter H, Navas ML (2003). Plant growth and competition at el-evated CO2: on winners, losers and functional groups. New Phy-tol 157(2):175–198

Potvin C, Tousignant D (1996) Evolutionary consequences of simu-lated global change: Genetic adaptation or adaptive phenotypicplasticity. Oecologia 108(4):683–693

Potvin C, Vasseur L (1997) Long-term CO2 enrichment of a pasture commu-nity: Species diversity and dominance pattern. Ecology 78:666–667

Ramseir D, Connolly J, Bazzaz FA (2005) Carbon dioxide regime,species identity and influence of species initial abundance asdeterminants of change in stand biomass composition in five-species communitites; an investigation using a simplex designand RGRD analysis. Journal of Ecology 93:502–511

Reich PB, Knops J, Tilman D, Craine J, Ellsworth D, Tjoelker M, Lee T,Naeem S, Wedin D, Bahauddin D, Hendrey G, Jose G, Wrage K, GothJ, Bengston W (2001a) Plant diversity enhances ecosystem responsesto elevated CO2 and nitrogen deposition. Nature 410:809–812

Reich PB, Tilman D, Craine J, Ellsworth D, Tjoelker M, Knops J, WedinD, Naeem S, Bahauddin D, Goth J, Bengston W, Lee T (2001b) Dospecies and functional groups differ in acquisition and use of C, Nand water under varying atmospheric CO2 and N deposition re-gimes? A field test with 16 grassland species. New Phyt 150:435–448

Reich PB, Tilman D, Naeem S, Ellsworth D, Knops J, Craine J, WedinD, Trost J (2004) Species and functional diversity independentlyinfluence biomass accumulation and its response to CO2 and N.Proc Nat Acad Sci USA 101 (27) 10101–10106

Reynolds JF, Bugmann H, Pitelka LF (2001) How much physiologyis needed in forest gap models for stimulating long-term veg-etation response to global changeÉ Challenges, limitations, andpotentials. Climatic Change 51:541–557

Roumet C, Laurent G, Canivenc G, Roy J (2002) Genotypic variationin the response of two perennial grass species to elevated car-bon dioxide. Oecologia 133(3):342–348

Sala OE, Lauenroth WK, McNaughton SJ, Rusch G, Zhang X (1996)Biodiversity and ecosystem function in grasslands. FunctionalRoles of Biodiversity: a Global Perspective. In: Mooney HA,Cushman JH, Medina E, Sala OE, Schulze E-D (eds) pp 129–149,John Wiley and Sons, London

Sala OE, Chapin FS, III, Armesto JJ, Berlow E, Bloomfield J, Dirzo R,Huber-Sanwald E, Huenneke LF, Jackson RB, Kinzig A, LeemansR, Lodge DM, Mooney HA, Oesterheld M, Poff NL, Sykes MT,Walker BH, Walker M, Wall DH (2000) Global Biodiversity Sce-narios for the Year 2100. Science 287:1770–1774

Schimel D S (1995) Terrestrial Ecosystems and the Carbon-Cycle.Glob Ch Biol 1(1):77–91

Schmid B, Birrer A, Lavigne C (1996) Genetic variation in the re-sponse of plant populations to elevated CO2 in a nutrient-poor,calcareous grassland. In: Körner C, Bazzaz FA, (eds) Carbon di-oxide, populations and communities. Physiological Ecology Se-ries. San Diego, CA, USA: Academic Press, pp 31–50

Smith SD, Huxman TE,Zitzer SF, Charlet TN, Housman DC, ColemanJS, Fenstermaker LK, Seemann JR, Nowak RS (2000) ElevatedCO2 increase productivity and invasive species success in an aridecosystem. Nature 408:79–82

Spehn EM, Hector A, Joshi J, Scherer-Lorenzen M, Schmid B, Bazeley-White E, Beirkuhnlein C, Caldeira MC, Diemers M, Dimitrako-poulos PG, Finn JA, Freitas H, Giller PS, Good J, Harris R, HogbergP, Huss-Danell K, Jumpponen A, Koricheva J, Leadley PW, LoreauM, Minns A, Mulder CPH, O’Donovan, Ottawy SJ, Palmborg C,Pereira JS, Pfisterer AB, Prinz A, Read DJ, Schulze ED, Siamantziou-ras A-S D, Terry AC, Troumbis AY, Woodward EI, Yachl SY, LawtonJH (2005) Ecosystem effects of biodiversity manipulations ineuropean grasslands. Ecological Monographs 7591) 37–63

Spinnler D, Egli P, Körner C (2003) Provenance effects and allom-etry in beech and spruce under elevated CO2 and nitrogen ontwo different forest soils. Basic Appl Ecol 4(5):467–478

Stewart J, Potvin C (1996) Effects of elevated CO2 on an artificialgrassland community: Competition, invasion and neighborhoodgrowth. Funct Ecol 10:157–166

Stiling P, Moon D, Hymus G, Drake B (2004) Differential effects ofelevated CO2 on acorn density, weight, germination, and preda-tion among three oak species in a scrub-oak forest. Glob ChBiol 10(2):228–232

Stocker R, Körner C, Schmid B, Niklaus PA, Leadley PW (1999) Afield study of the effects of elevated CO2 and plant species di-versity on ecosystem-level gas exchange in a planted calcareousgrassland. Glob Ch Biol 5(1):95–105

Stocklin J, Körner C (1999) Interactive effects of elevated CO2, Pavailability and legume presence on calcareous grassland: re-sults of a glasshouse experiment. Funct Ecol 13(2):200–209

Teyssonneyre F, Picon-Cochard C, Soussana JF (2002) How can wepredict the effects of elevated CO2 on the balance between pe-rennial C3 grass species competing for light? New Phytol 154(1):53–64

Thomas CD, Cameron A, Green RE, Bakkenes M, Beaumont LJ,Collingham YC, Barend F, Erasmus N, Ferreira de Siqueira M,Grainger A, Hannah L, Hughes L, Huntley B, van Jaarsveld AS,Midgley GF, Miles L, Ortega-Huerta MA, Peterson AT, PhillipsOL, Williams SE (2004a) Extinction risk from climate change.Nature 427:145–148

Thomas JA, Telfer MG, Roy DB, Preston CD, Greenwood JJD, AsherJ, Fox R, Clarke RT, Lawton JH (2004b) Comparative losses ofBritish butterflies, birds, and plants and the global extinctioncrisis. Science 303:1879–1881

Thürig B, Körner C, Stocklin J (2003) Seed production and seedquality in a calcareous grassland in elevated CO2. Glob Ch Biol9(6):873–884

Tilman D, Downing JA (1994) Biodiversity and stability in grass-lands. Nature 367:363–365

Tilman D, Knops J, Wedin D (1997) The influence of functional di-versity and composition on ecosystem processes. Science277:1300–1302

Tilman D, Reich PB, Knops J, Wedin D, Mielke T, Lehman C (2001)Diversity and productivity in the long-term grassland experi-ment. Science 294:843–845

Van Voris P, O’Neill RV, Emanuel WR, Shugart Jr HH (1980) Func-tional complexity and ecosystem stability. Ecology 61:1352–1360

Vasseur L, Potvin C (1997) Natural pasture community response toenriched CO2 atmosphere. Plant Ecol 135:31–41

Volk M, Niklaus PA, Körner C (2000) Soil moisture effects determineCO2 responses of grassland species. Oecologia 125(3):380–388

Walker BH (2001) Tropical savanna. In: Chapin III FS, Sala OE,Huber-Sannwald E (eds) Global Biodiversity in a Changing En-vironment: Scenarios for the 21st Century. Springer-Verlag, NewYork, pp 139–156

Wand SJE, Midgley GF, Jones MH, Curtis PS (1999) Responses ofwild C4 and C3 (Poaceae) species to elevated atmospheric CO2concentration: a meta-analytic test of current theories and per-ceptions. Glob Ch Biol 5:723–741

Wang ZM, Lechowicz MJ, Potvin C (1994) Early selection of blackspruce seedlings and global change: Which genotypes shouldwe favor? Ecol Appl 4:604–616

Ward JK, Kelly JK (2004) Scaling up evolutionary responses to el-evated CO2: lessons from Arabidopsis. Ecol Lett 7(5):427–440

Ward JK, Antonovics J, Thomas RB, Strain BR (2000) Is atmo-spheric CO2 a selective agent on model C3 annuals? Oecologia123(3):330–341

Woodward FI (2002) Potential impacts of global elevated CO2concentrations on plants. Curr Opin Plant Biol 5:207–211

Zavaleta ES, Shaw MR, Chiariello NR, Mooney HA, Field CB(2003a) Additive effects of simulated climate changes, elevatedCO2, and nitrogen deposition on grassland diversity. PNAS100(13):7650–7654

Zavaleta ES, Shaw MR, Chiariello NR, Thomas BD, Cleland EE, FieldCB, Mooney HA (2003b) Grassland responses to three years ofelevated temperature, CO2, precipitation, and N deposition. EcolMonog 73(4):585–604