Soil Carbon Pools in Dryland Agroecosystems as Affected by Several Years of Drought.

No-till and increased cropping intensity (CI) can increase yield and soil organic C (SOC) in the US Great Plains compared with traditional wheat ( L.)-fallow management. However, gains in SOC and other C pools may not be permanent. Increasing frequency of drought may reduce C inputs and potentially reverse gains accrued during wetter periods. This study examined the effect of drought on the persistence of SOC with two objectives: (i) to determine soil C pools (0-20 cm) after 24 yr in no-till as influenced by potential evapotranspiration (PET), landscape position (slope), and CI; and (ii) to compare the size of the C pools after the first 12 yr (wet) versus the subsequent 12 yr, notable for frequent droughts. Rotations were wheat-corn ( L.)-fallow (WCF), continuous cropping (CC), and a grass Conservation Reserve Program mixture planted across slopes at three sites in Colorado with similar precipitation but increasing PET. After 24 yr, water-soluble organic C increased with CI from WCF to CC to grass with 250, 340, and 440 kg C ha, respectively. Soil microbial biomass C also increased with CI-1500, 1660, and 2135 kg C ha for WCF, CC, and grass, respectively. The particulate organic matter C pool had a three-way interaction with PET, slope, and CI. Overall, between Years 12 and 24, SOC increased in grass by 16.9%, with a rate of 425 kg C ha yr sequestration compared with 10.5 and 1.4% for the WCF and CC systems, respectively.

Effects of manipulated herbivore inputs on nutrient flux and decomposition in a tropical rainforest in Puerto Rico.

Forest canopy herbivores are known to increase rates of nutrient fluxes to the forest floor in a number of temperate and boreal forests, but few studies have measured effects of herbivore-enhanced nutrient fluxes in tropical forests. We simulated herbivore-induced fluxes in a tropical rainforest in Puerto Rico by augmenting greenfall (fresh foliage fragments), frassfall (insect feces), and throughfall (precipitation enriched with foliar leachates) in replicated experimental plots on the forest floor. Background rates of greenfall and frassfall were measured monthly using litterfall collectors and augmented by adding 10× greenfall or 10× frassfall to designated plots. Throughfall fluxes of NH(4), NO(3) and PO(4) (but not water) were doubled in treatment plots, based on published rates of fluxes of these nutrients in throughfall. Control plots received only background flux rates for these compounds but the same minimum amount of distilled water. We evaluated treatment effects as changes in flux rates for NO(3), NH(4) and PO(4), measured as decomposition rate of leaf litter in litterbags and as adsorption in ion-exchange resin bags at the litter-soil interface. Frass addition significantly increased NO(3) and NH(4) fluxes, and frass and throughfall additions significantly reduced decay rate, compared to controls. Reduced decay rate suggests that nitrogen flux was sufficient to inhibit microbial decomposition activity. Our treatments represented fluxes expected from low-moderate herbivore outbreaks and demonstrated that herbivores, at these outbreak levels, increase ecosystem-level N and P fluxes by >30% in this tropical rainforest.

The influence of a neotropical herbivore (Lamponius portoricensis) on nutrient cycling and soil processes.

The role of phytophagous insects in ecosystem nutrient cycling remains poorly understood. By altering the flow of litterfall nutrients from the canopy to the forest floor, herbivores may influence key ecosystem processes. We manipulated levels of herbivory in a lower montane tropical rainforest of Puerto Rico using the common herbivore, Lamponius portoricensis (Phasmatidea), on a prevalent understory plant, Piper glabrescens (Piperaceae), and measured the effects on nutrient input to the forest floor and on rates of litter decomposition. Four treatment levels of herbivory generated a full range of leaf area removal, from plants experiencing no herbivory to plants that were completely defoliated (>4,000 cm(2) leaf area removed during the 76-day study duration). A significant (P<0.05) positive regression was found between all measures of herbivory (total leaf area removed, greenfall production, and frass-related inputs) and the concentration of NO (3) (-) in ion exchange resin bags located in the litter layer. No significant relationship was found between any of the herbivory components and resin bag concentrations of NH (4) (+) or PO (4) (-) . Rates of litter decay were significantly affected by frass-related herbivore inputs. A marginally significant negative relationship was also found between the litter mass remaining at 47 days and total leaf area removed. This study demonstrated a modest, but direct relationship between herbivory and both litter decomposition and NO (3) (-) transfer to the forest floor. These results suggest that insect herbivores can influence forest floor nutrient dynamics and thus merit further consideration in discussions on ecosystem nutrient dynamics.