Plant nitrogen concentration and isotopic composition in residential lawns across seven US cities.

Human drivers are often proposed to be stronger than biophysical drivers in influencing ecosystem structure and function in highly urbanized areas. In residential land cover, private yards are influenced by individual homeowner preferences and actions while also experiencing large-scale human and biophysical drivers. We studied plant nitrogen (%N) and N stable isotopic composition (δ(15)N) in residential yards and paired native ecosystems in seven cities across the US that span major ecological biomes and climatic regions: Baltimore, Boston, Los Angeles, Miami, Minneapolis-St. Paul, Phoenix, and Salt Lake City. We found that residential lawns in three cities had enriched plant δ(15)N (P < 0.03) and in six cities higher plant N (%) relative to the associated native ecosystems (P < 0.05). Plant δ(15)N was progressively depleted across a gradient of urban density classes in Baltimore and Boston (P < 0.05). Lawn fertilization was associated with depleted plant δ(15)N in Boston and Los Angeles (P < 0.05), and organic fertilizer additions were associated with enriched plant δ(15)N in Los Angeles and Salt Lake City (P < 0.04). Plant δ(15)N was significantly enriched as a function of housing age in Baltimore (r (2) = 0.27, P < 0.02), Boston (r (2) = 0.27, P < 0.01), and Los Angeles (r (2) = 0.34, P < 0.01). These patterns in plant δ(15)N and plant N (%) across these cities suggests that N sources to lawns, as well as greater rates of N cycling combined with subsequent N losses, may be important drivers of plant N dynamics in lawn ecosystems at the national scale.

Vehicle emissions and fertilizer impact the leaf chemistry of urban trees in Salt Lake Valley, UT.

The urban nitrogen (N) and carbon (C) cycles are substantially influenced by human activity. Alterations to these cycles include increased inputs from fossil fuel combustion and fertilizer use. The leaf chemistry of urban trees can be used to distinguish between these different N and C sources. Here, we evaluated relationships between urban vegetation and different N and C sources in street and residential trees in the Salt Lake Valley, Utah. We tested three hypotheses: 1) unfertilized street trees on high traffic density roads will have higher leaf %N, more enriched δ15N and more depleted δ13C than unfertilized street trees on low traffic density roads; 2) trees in high income residential neighborhoods will have higher leaf %N, more depleted δ15N and more enriched δ13C than trees in lower income neighborhoods; and 3) unfertilized street trees will have lower leaf %N, more enriched δ15N and more depleted δ13C than fertilized residential trees. Leaf δ15N was more enriched near high traffic density roads for one study species. However, street tree δ15N and δ13C were largely influenced by vehicle emissions from primary and secondary roads within 1000 m radius rather than the immediately adjacent road. Leaf δ13C was correlated with neighborhood income, although this relationship may be the result of variations in irrigation practices rather than variations in C sources. Finally, unfertilized trees in downtown Salt Lake had lower leaf %N, more enriched δ15N and more depleted δ13C than fertilized trees. These results highlight that urban trees can serve as biomonitors of the environment. Moreover, they emphasize that roads can have large spatial footprints and that the leaf chemistry of urban vegetation may be influenced by the spatial patterns in roads and road densities at the landscape scale.

Gender-specific patterns of aboveground allocation, canopy conductance and water use in a dominant riparian tree species: Acer negundo.

Acer negundo Sarg. (box elder) is a dioecious tree species that dominates riparian systems at mid elevations throughout the southwest and Intermountain West of the United States. Previous studies have shown that female A. negundo trees occur at higher frequencies along stream margins, whereas males occur at higher frequencies in drier microsites. To better understand the adaptive significance of sex ratio biases and their impact on the ecohydrology of riparian ecosystems, we examined whole-plant water relations and hydraulic properties of mature male and female A. negundo trees occurring within 1 m of a perennial stream channel. We hypothesized that (1) females would have significantly greater canopy water fluxes than males (particularly during periods of seed production: May-June), and (2) xylem in females is more hydraulically efficient but more vulnerable to cavitation than xylem in males. Mean sap flux density (J(s)) during the early growing season (May and June) was 43% higher in female trees than in male trees (n = 6 and 7 trees respectively, P < 0.0001). Mean J(s) in July and August remained 17% higher in females than in males (P = 0.0009). Mean canopy stomatal conductance per unit leaf area (g(s,leaf)) in May and June was on average 140% higher in females than in males (P < 0.0001). Mean g(s,leaf) in July and August remained 69% higher in female trees than in male trees (P < 0.0001). Canopy stomatal conductance scaled to basal area was 90 and 31% higher in females relative to males during May-June and July-August, respectively (P < 0.0001 during both periods). Conversely, there were no apparent differences in either branch hydraulic conductance or branch xylem cavitation vulnerability between genders. These results improve our capacity to describe the adaptive forces that shape the spatial distribution of male and female trees in dioecious species, and their consequences for ecohydrological processes in riparian ecosystems.

Sensitivity of mean canopy stomatal conductance to vapor pressure deficit in a flooded Taxodium distichum L. forest: hydraulic and non-hydraulic effects.

We measured the xylem sap flux in 64-year-old Taxodium distichum (L.) Richard trees growing in a flooded forest using Granier-type sensors to estimate mean canopy stomatal conductance of the stand (G S). Temporal variations in G S were investigated in relation to variation in vapor pressure deficit (D), photosynthetic photon flux density (Q o), and the transpiration rate per unit of leaf area (E L), the latter variable serving as a proxy for plant water potential. We found that G S was only weakly related to Q o below 500 µmol m-2 s-1 (r 2=0.29), but unrelated to Q o above this value. Above Q o=500 µmol m-2 s-1 and D=0.6 kPa, G S decreased linearly with increasing E L with a poor fit (r 2=0.31), and linearly with lnD with a much better fit (r 2=0.81). The decrease of G S with lnD was at a rate predicted based on a simple hydraulic model in which stomata regulate the minimum leaf water potential. Based on the hydraulic model, stomatal sensitivity to D is proportional to stomatal conductance at low D. A hurricane caused an ~41% reduction in leaf area. This resulted in a 28% increase in G S at D=1 kPa (G Sref), indicating only partial compensation. As predicted, the increase in G Sref after the hurricane was accompanied by a similar increase in stomatal sensitivity to D (29%). At night, G Sref was ~20% of the daytime value under non-limiting light (Q o>500 µmol m-2 s-1). However, stomatal sensitivity to D decreased only to ~46% (both reductions referenced to pre-hurricane daytime values), thus having more than twice the sensitivity expected based on hydraulic considerations alone. Therefore, non-hydraulic processes must cause heightened nighttime stomatal sensitivity to D.

Inferring biogenic and anthropogenic carbon dioxide sources across an urban to rural gradient.

We continuously monitored CO(2) concentrations at three locations along an urban-to-rural gradient in the Salt Lake Valley, Utah from 2004 to 2006. The results showed a range of CO(2) concentrations from daily averages exceeding 500 p.p.m. at the city center to much lower concentrations in a non-urbanized, rural region of the valley. The highest values were measured in the wintertime and under stable atmospheric conditions. At all three sites, we utilized weekly measurements of the C and O isotope composition of CO(2) for a 1-year period to evaluate the CO(2) sources underlying spatial and temporal variability in CO(2) concentrations. The results of an inverse analysis of CO(2) sources and the O isotope composition of ecosystem respiration (delta(18) O (R)) showed large contributions (>50%) of natural gas combustion to atmospheric CO(2) in the wintertime, particularly at the city center, and large contributions (>60%) of biogenic respiration to atmospheric CO(2) during the growing season, particularly at the rural site. delta(18) O (R) was most enriched at the rural site and more isotopically depleted at the urban sites due to the effects of irrigation on ecosystem water pools at the urban sites. The results also suggested differences in the role of leaf versus soil respiration between the two urban sites, with seasonal variation in the contribution of leaf respiration at a residential site and relatively constant contributions of leaf respiration at the city center. These results illustrate that spatial and temporal patterns of urban CO(2) concentrations and isotopic composition can be used to infer patterns of energy use by urban residents as well as plant and soil processes in urban areas.

Canopy conductance of Pinus taeda, Liquidambar styraciflua and Quercus phellos under varying atmospheric and soil water conditions.

Sap flow, and atmospheric and soil water data were collected in closed-top chambers under conditions of high soil water potential for saplings of Liquidambar styraciflua L., Quercus phellos L. and Pinus taeda L., three co-occurring species in the southeastern USA. Responses of canopy stomatal conductance (g(t)) to water stress induced by high atmospheric water vapor demand or transpiration rate were evaluated at two temporal scales. On a diurnal scale, the ratio of canopy stomatal conductance to maximum conductance (g(t)/g(t,max)) was related to vapor pressure deficit (D), and transpiration rate per unit leaf area (E(l)). High D or E(l) caused large reductions in g(t)/g(t,max) in L. styraciflua and P. taeda. The response of g(t)/g(t,max) to E(l) was light dependent in L. styraciflua, with higher g(t)/g(t,max) on sunny days than on cloudy days. In both L. styraciflua and Q. phellos, g(t)/g(t,max) decreased linearly with increasing D (indicative of a feed-forward mechanism of stomatal control), whereas g(t)/g(t,max) of P. taeda declined linearly with increasing E(l) (indicative of a feedback mechanism of stomatal control). Longer-term responses to depletion of soil water were observed as reductions in mean midday g(t)/g(t,max), but the reductions did not differ significantly between species. Thus, species that employ contrasting methods of stomatal control may show similar responses to soil water depletion in the long term.

Sap-flux-scaled transpiration responses to light, vapor pressure deficit, and leaf area reduction in a flooded Taxodium distichum forest.

We used 20-mm-long, Granier-type sensors to quantify the effects of tree size, azimuth and radial position in the xylem on the spatial variability in xylem sap flux in 64-year-old trees of Taxodium distichum L. Rich. growing in a flooded forest. This information was used to scale flux to the stand level to investigate variations in half-hourly and daily (24-hour) sums of sap flow, transpiration per unit of leaf area, and stand transpiration in relation to vapor pressure deficit (D) and photosynthetically active radiation (Q(o)). Measurements of xylem sap flux density (J(s)) indicated that: (1) J(s) in small diameter trees was 0.70 of that in medium and large diameter trees, but the relationship between stem diameter as a continuous variable and J(s) was not significant; (2) J(s) at 20-40 mm depth in the xylem was 0.40 of that at 0-20 mm depth; and (3) J(s) on the north side of trees was 0.64 of that in directions 120 degrees from the north. Daily transpiration was linearly related to daily daytime mean D, and reached a modest value of 1.3 mm day(-1), reflecting the low leaf area index (LAI = 2.2) of the stand. Because there was no soil water limitation, half-hourly water uptake was nearly linearly related to D at D < 0.6 kPa during both night and day, increasing to saturation during daytime at higher values of D. The positive effect of Q(o) on J(s) was significant, but relatively minor. Thus, a second-order polynomial with D explained 94% of the variation in J(s) and transpiration. An approximately 40% reduction in LAI by a hurricane resulted in decreases of about 18% in J(s) and stand transpiration, indicating partial stomatal compensation.

Water relations in grassland and desert ecosystems exposed to elevated atmospheric CO2.

Atmospheric CO2 enrichment may stimulate plant growth directly through (1) enhanced photosynthesis or indirectly, through (2) reduced plant water consumption and hence slower soil moisture depletion, or the combination of both. Herein we describe gas exchange, plant biomass and species responses of five native or semi-native temperate and Mediterranean grasslands and three semi-arid systems to CO2 enrichment, with an emphasis on water relations. Increasing CO2 led to decreased leaf conductance for water vapor, improved plant water status, altered seasonal evapotranspiration dynamics, and in most cases, periodic increases in soil water content. The extent, timing and duration of these responses varied among ecosystems, species and years. Across the grasslands of the Kansas tallgrass prairie, Colorado shortgrass steppe and Swiss calcareous grassland, increases in aboveground biomass from CO2 enrichment were relatively greater in dry years. In contrast, CO2-induced aboveground biomass increases in the Texas C3/C4 grassland and the New Zealand pasture seemed little or only marginally influenced by yearly variation in soil water, while plant growth in the Mojave Desert was stimulated by CO2 in a relatively wet year. Mediterranean grasslands sometimes failed to respond to CO2-related increased late-season water, whereas semiarid Negev grassland assemblages profited. Vegetative and reproductive responses to CO2 were highly varied among species and ecosystems, and did not generally follow any predictable pattern in regard to functional groups. Results suggest that the indirect effects of CO2 on plant and soil water relations may contribute substantially to experimentally induced CO2-effects, and also reflect local humidity conditions. For landscape scale predictions, this analysis calls for a clear distinction between biomass responses due to direct CO2 effects on photosynthesis and those indirect CO2 effects via soil moisture as documented here.