Publications

We investigated the effect of CO₂ concentration and soilnutrient availability during growth on the subsequent decomposition andnitrogen (N) release from litter of four annual grasses that differ inresource requirements and native habitat. Vulpia microstachys isa native grass found on California serpentine soils, whereas Avenafatua, Bromus hordaceus, and Lolium multiflorum areintroduced grasses restricted to more fertile sandstone soils (Hobbs & Mooney 1991). Growth in elevated CO₂ altered litter C:N ratio,decomposition, and N release, but the direction and magnitude of thechanges differed among plant species and nutrient treatments. ElevatedCO₂ had relatively modest effects on C:N ratio of litter,increasing this ratio in Lolium roots (and shoots at high nutrients),but decreasing C:N ratio in Avena shoots. Growth of plants underelevated CO₂ decreased the decomposition rate of Vulpialitter, but increased decomposition of Avena litter from the high-nutrient treatment. The impact of elevated CO₂ on N loss fromlitter also differed among species, with Vulpia litter from high-CO₂ plants releasing N more slowly than ambient-CO₂litter, whereas growth under elevated CO₂ caused increased Nloss from Avena litter. CO₂ effects on N release in Lolium and Bromus depended on the nutrient regime in whichplants were grown. There was no overall relationship between litter C:Nratio and decomposition rate or N release across species and treatments.Based on our study and the literature, we conclude that the effects ofelevated CO₂ on decomposition and N release from litter arehighly species-specific. These results do not support the hypothesis thatCO₂ effects on litter quality consistently lead to decreasednutrient availability in nutrient-limited ecosystems exposed to elevatedCO₂.

Patterns of carbon source utilization, or community-level physiological profiles (CLPP), produced from direct incubation of environmental samples in BIOLOG microplates can consistently discriminate spatial and temporal gradients within microbial communities. While the resolving power of the assay appears significant, the basis for the differences in the patterns of sole carbon source utilization among communities remains unclear. Carbon source utilization as measured in this assay is a measure of functional potential, rather than <em class="EmphasisTypeItalic ">in situ</em> activity, since enrichment occurs over the course of incubation, which can range from 24 to 72 hours (or even longer) depending on inoculum density. The functional profile of a community could be an indicator of carbon source availability and concomitant selection for specific functional types of organisms. A more limited view of the profile is as a composite descriptor of the microbial community composition without any ecologically relevant functional information. We manipulated microbial community structure and function in laboratory microcosms to evaluate their influence on CLPP. The structure of rhizosphere communities was controlled by inoculating axenic plants (wheat and potato) with different mixed species (non-gnotobiotic) inocula. Inoculum source influenced CLPP more strongly than plant type, indicating that CLPP primarily reflected differences in microbial community structure than function. In order to more specifically examine the influence of microbial function on CLPP, specific carbon sources in the BIOLOG plates (asparagine and acetate) were added to a continuously stirred tank reactor (CSTR) containing a mixed community of microorganisms degrading plant material. Daily additions of these carbon sources at levels up to 50% of the total respired carbon in the bioreactor caused significant changes in overall CLPP, but caused no, or minor, increases in the specific response of these substrates in the plates. These studies indicate that the functional relevance of CLPP should be interpreted with caution.

The concentration of carbon dioxide (CO₂) in the Earth's atmosphere is rising rapidly^{<a href="http://www.nature.com/nature/journal/v388/n6642/full/388576a0.html#B1">1</a>}, with the potential to alter many ecosystem processes. Elevated CO₂ often stimulates photosynthesis^{<a href="http://www.nature.com/nature/journal/v388/n6642/full/388576a0.html#B2">2</a>}, creating the possibility that the terrestrial biosphere will sequester carbon in response to rising atmospheric CO₂ concentration, partly offsetting emissions from fossil-fuel combustion, cement manufacture, and deforestation^{<a href="http://www.nature.com/nature/journal/v388/n6642/full/388576a0.html#B3">3</a>},^{<a href="http://www.nature.com/nature/journal/v388/n6642/full/388576a0.html#B4">4</a>}. However, the responses of intact ecosystems to elevated CO₂ concentration, particularly the below-ground responses, are not well understood. Here we present an annual budget focusing on below-ground carbon cycling for two grassland ecosystems exposed to elevated CO₂ concentrations. Three years of experimental CO₂ doubling increased ecosystem carbon uptake, but greatly increased carbon partitioning to rapidly cycling carbon pools below ground. This provides an explanation for the imbalance observed in numerous CO₂ experiments, where the carbon increment from increased photosynthesis is greater than the increments in ecosystem carbon stocks. The shift in ecosystem carbon partitioning suggests that elevated CO₂ concentration causes a greater increase in carbon cycling than in carbon storage in grasslands.

<p id="p-1">David A. Wedin and David Tilman (Reports, 6 Dec.,<a href="http://science.sciencemag.org/lookup/volpage/275/1720">p 1720</a>) show that increased nitrogen inputs to terrestrial ecosystems might cause smaller increases in the capacity of those ecosystems to store carbon than expected. Their findings are important because nitrogen inputs have increased dramatically over the past decades through fertilizer production, cultivation of nitrogen-fixing legumes, and production of oxides of nitrogen associated with fossil-fuel burning (<a id="xref-ref-1-1" class="xref-bibr" href="http://science.sciencemag.org/content/275/5301/737.5#ref-1">1</a>). However, the simultaneous increase in atmospheric carbon dioxide (CO₂) concentrations caused by burning fossil fuels is likely to at least partially counteract the processes that limited carbon storage in Wedin and Tilman's experiment. CO₂ enrichment generally increases the amount of carbon fixed by plants per unit of nitrogen taken up from the soil, particularly in carbon-3 (C³) species (<a id="xref-ref-2-1" class="xref-bibr" href="http://science.sciencemag.org/content/275/5301/737.5#ref-2">2</a>) such as those that invaded their nitrogen-enriched plots. Compared with the C⁴ species that thrived before nitrogen was added, the invading C³ species have relatively lower C-to-N ratios, limiting the amount of carbon stored in response to nitrogen input. However, with elevated CO₂ tending to increase the C-to-N ratio of these C³ plants, N and CO₂ enrichment in concert would likely cause greater C storage than observed by Wedin and Tilman.</p> <p id="p-2">Rising atmospheric CO₂ may also increase N inputs to terrestrial ecosystems, amplifying the direct human impact on the N cycle. CO₂ enrichment often increases the growth of plants housing N-fixing bacteria in their roots, and this stimulation is relatively larger than non-N-fixing plants (<a id="xref-ref-3-1" class="xref-bibr" href="http://science.sciencemag.org/content/275/5301/737.5#ref-3">3</a>). Thus, in addition to the direct anthropogenic stimulation of N inputs to terrestrial ecosystems through agriculture and fossil-fuel burning (<a id="xref-ref-1-2" class="xref-bibr" href="http://science.sciencemag.org/content/275/5301/737.5#ref-1">1</a>), humans may indirectly increase N inputs to terrestrial ecosystms by increasing atmospheric CO2 concentrations. The interaction between CO2 and N enrichment, as well as shifts in plant species, will likely influence future C storage by the terrestrial biosphere</p>

We examined the effects of growth carbon dioxide (CO₂)concentration and soil nutrient availability on nitrogen (N)transformations and N trace gas fluxes in California grasslandmicrocosms during early-season wet-up, a time when rates of Ntransformation and N trace gas flux are high. After plant senescenceand summer drought, we simulated the first fall rains and examined Ncycling. Growth at elevated CO₂ increased root productionand root carbon:nitrogen ratio. Under nutrient enrichment, elevatedCO₂ increased microbial N immobilization during wet-up,leading to a 43% reduction in gross nitrification anda 55% reduction in NO emission from soil. ElevatedCO₂ increased microbial N immobilization at ambientnutrients, but did not alter nitrification or NO emission. ElevatedCO₂ did not alter soil emission of N₂O ateither nutrient level. Addition of NPK fertilizer (1:1:1) stimulatedN mineralization and nitrification, leading to increased N₂Oand NO emission from soil. The results of our study support a mechanisticmodel in which elevated CO₂ alters soil N cycling and NOemission: increased root production and increased C:N ratio in elevatedCO₂ stimulate N immobilization, thereby decreasingnitrification and associated NO emission when nutrients are abundant.This model is consistent with our basic understanding of how C availabilityinfluences soil N cycling and thus may apply to many terrestrial ecosystems.

Variation in competitive ability due to variation in soil characteristics is one possible mechanism allowing the local coexistence of plant species. We measured soil water, depth, and nitrogen pools and fluxes in distinct patches of three serpentine grassland species to determine whether soil heterogeneity existed and was correlated with plant species abundance. Through experimental manipulation of species’ abundances, we also examined the relative importance of inherent site characteristics vs. plant species’ effects in generating heterogeneity in the measured soil characteristics; and measured species’ competitive abilities in different patch types. The three common grassland annuals, <em>Calycadenia multiglandulosum, Plantago erecta,</em> and <em>Lasthenia californica,</em> were segregated with respect to the measured soil characteristics. Differences in soil water, soil depth, soil microbial nitrogen, and soil carbon to nitrogen ratio were due to inherent site characteristics, while differences in nitrate availability were strongly affected by the identity of the species currently growing in a soil patch. Furthermore, all species performed significantly better against one other species in the patch type where they are normally most abundant. These results demonstrate that species diversity within this grassland contributes to soil heterogeneity and suggest that soil heterogeneity could contribute to the coexistence of these species.