Larry G. Bundy

Soil carbon pools and fluxes in long-term corn belt agroecosystems

The dynamics of soil organic carbon (SOC) play an important role in long-term ecosystem productivity and the global C cycle. We used extended laboratory incubation and acid hydrolysis to analytically determine SOC pool sizes and fluxes in US Corn Belt soils derived from both forest and prairie vegetation. Measurement of the natural abundance of 13 C made it possible to follow the influence of continuous corn on SOC accumulation. The active pools (Ca) comprised 3 to 8% of the SOC with an average field mean residence time (MRT) of 100 d. The slow pools (Cs) comprised 50% of SOC in the surface and up to 65% in subsoils. They had field MRTs from 12‐28 y for C4-C and 40‐80 y for C3-derived C depending on soil type and location. Notill management increased the MRT of the C3-C by 10‐15 y above conventional tillage. The resistant pool (Cr) decreased from an average of 50% at the surface to 30% at depth. The isotopic composition of SOC mineralized during the early stages of incubation reflected accumulations of labile C from the incorporation of corn residues. The CO2 released later reflected 13 C characteristic of the Cs pool. The 13 C of the CO2 did not equal that of the whole soil until after 1000 d of incubation. The SOC dynamics determined by acid hydrolysis, incubation and 13 C content were dependent on soil heritage (prairie vs. forest), texture, cultivation and parent material, depositional characteristics. Two independent methods of determining C3 pool sizes derived from C3 vegetation gave highly correlated values. # 2000 Elsevier Science Ltd. All rights reserved.

Measurements and Modeling of Carbon and Nitrogen Cycling in Agroecosystems of Southern Wisconsin: Potential for SOC Sequestration during the Next 50 Years

Landmanagement practices such as no-tillage agriculture and tallgrass prairie restoration have been proposed as a possible means to sequester atmospheric carbon, helping to refurbish soil fertility and replenish organic matter lost as a result of previous agricultural management practices. However, the relationship between land-use changes and ecosystem structure and functioning is not yet understood. We studied soil and vegetation properties over a 4-year period (1995–98), and assembled measurements of microbial biomass, soil organic carbon (SOC) and nitrogen (N), N-mineralization, soil surface carbon dioxide (CO2) flux, and leached C and N in managed (maize; Zea mays L.) and natural (prairie) ecosystems near the University of Wisconsin Agricultural Research Station at Arlington. Field data show that different management practices (tillage and fertilization) and ecosystem type (prairie vs maize) have a profound influence on biogeochemistry and water budgets between sites. These measurements were used in conjunction with a dynamic terrestrial ecosystem model, called IBIS (the Integrated Biosphere Simulator), to examine the long-term effects of land-use changes on biogeochemical cycling. Field data and modeling suggest that agricultural land management near Arlington between 1860 and 1950 caused SOC to be depleted by as much as 63% (native SOC approximately 25.1 kg C m−2). Reductions in N-mineralization and microbial biomass were also observed. Although IBIS simulations depict SOC recovery in no-tillage maize since the 1950s and also in the Arlington prairie since its restoration was initiated in 1976, field data suggest otherwise for the prairie. This restoration appears to have done little to increase SOC over the past 24 years. Measurements show that this prairie contained between 28% and 42% less SOC (in the top 1 m) than the no-tillage maize plots and 40%–47% less than simulated potential SOC for the site in 1999. Because IBIS simulates competition between C3 and C4 grass species, we hypothesized that current restored prairies, which include many forbs not characterized by the model, could be less capable of sequestering C than agricultural land planted entirely in monocultural grass in this region. Model output and field measurements show a potential 0.4 kg C m−2 y−1 difference in prairie net primary production (NPP). This study indicates that high-productivity C4 grasslands (NPP = 0.63 kg C m−2 y−1) and high-yield maize agroecosystems (10 Mg ha−1) have the potential to sequester C at a rate of 74.5 g C m−2 y−1 and 86.3 g C m−2 y−1, respectively, during the next 50 years across southern Wisconsin.

CARBON BUDGETS FOR A PRAIRIE AND AGROECOSYSTEMS: EFFECTS OF LAND USE AND INTERANNUAL VARIABILITY

Midwestern grasslands have undergone dramatic changes in land use and management practices, but the effects of these changes on terrestrial carbon budgets are poorly understood. This study compared, for five years, the effects of land-use type on components of the carbon (C) budget (above- and belowground net primary production [NPP], C leaching, soil surface CO2 flux, vegetation and soil C contents, and C export from burning and grain removal) of a restored tallgrass prairie and maize agroecosystems on a silt loam soil. Interannual variation of the C budget was addressed by correlating annual fluctuations of environmental variables and soil properties with C-budget components. The C losses we estimated, in order of increasing magnitude, were C leaching, grain C removal from the maize agroecosystems, and soil surface CO2 flux. NPP was significantly greater for N-fertilized maize (10.4 Mg C·ha−1·yr−1) than unfertilized maize agroecosystems (6.2 Mg C·ha−1·yr−1), and both were significantly greater than rest...

Methodological limitations and N-budget differences among a restored tallgrass prairie and maize agroecosystems

Abstract Interpretation of elemental balances requires careful assessment of component terms and their errors, especially for the major terms of the nitrogen (N) budget which has implications for environmental health. This study reports results from independent field measurements of major annual N-budget components, including atmospheric deposition, fertilizer added, net mineralization, residue returned, soil storage changes of inorganic N, leaching, and plant uptake. Measurements were made in a restored tallgrass prairie and optimally and deficiently N-fertilized, no-tillage and chisel-plowed maize (Zea mays L.) agroecosystems on Plano silt loam soil (fine-silty, mixed, superactive, mesic Typic Argiudoll (USDA); Haplic Phaeozem (approximate FAO)) in Wisconsin between 1995 and 1999. Denitrification and N losses due to runoff were assumed negligible and bulk density was assumed uniform with depth and across ecosystems. Annual inorganic N leaching was negligible in the restored prairie, but represented 3–57% of the amount of fertilizer-N applied in the optimally N-fertilized agroecosystems. On an annual basis, closure of the inorganic-N budget yielded cumulative errors that were often undesirably large; indicating methodological problems with quantifying ecosystem N cycling in situ. Increased spatial sampling is required to reduce individual measurement errors of two components with large uncertainties; namely net N-mineralization and soil inorganic N changes. Profile-scaled net N-mineralization generally did not balance with the residue N input from the previous year, but the imbalance agreed with the N-budget imbalance. Both results suggest that the prairie is accumulating N slowly, the deficiently N-fertilized maize plots are losing N more rapidly, and the optimally N-fertilized maize plots have too large an uncertainty to be interpreted confidently. Nitrogen-use efficiency, defined on a N-uptake basis, did not differ among the prairie and deficiently N-fertilized maize for 3 out of 5 years, but the prairie was significantly more efficient than the optimally N-fertilized maize.

Effects of management practices on annual net N-mineralization in a restored prairie and maize agroecosystems

Nitrogen (N) mineralization is a spatially variable and difficult component of the N cycle to quantify accurately under field conditions. Net N-mineralization was compared by direct measurement, indirect estimate, and laboratory incubation for a restored tallgrass prairie and for deficiently and optimally N-fertilized, no-tillage (NT) and chisel-plowed (CP) maize (Zea mays L.) agroecosystems on Plano silt loam soil (fine-silty, mixed, superactive, mesic Typic Argiudoll) in Wisconsin, USA. Four years of in-situ field measurements using an incubated-soil-core/ion-exchange-resin-bag technique showed that land use significantly affected net N-mineralization. Net N-mineralization was consistently smaller in the restored prairie than in the maize agroecosystems and typically larger in the CP than in the NT maize agroecosystems. Three independent methods for indirectly estimating annual net N-mineralization (i.e., N budget residual, deficiently N-fertilized plant N uptake, and profile-scaled in-situ field measurements) were relatively consistent at capturing land-use and tillage effects on net N-mineralization. Laboratory incubation and periodic leaching of Fall-sampled soils demonstrated that both mineralized N and labile C were co-limiting factors influencing N-mineralization in agricultural soils and generally supported field measurements by showing a significant difference in net N-mineralization with and without added fertilizer-N.

Phosphorus source effects on soil test phosphorus and forms of phosphorus in soil

Phosphorus (P) is often supplied to field crops in organic forms such as manures or biosolids, but P availability and appropriate application rates may differ between sources. An incubation study was conducted using a Ringwood silt loam soil and seven P sources. The P sources were low, medium, and high P manure (feces) from a dairy feeding study, whole manure, fiber manure from a liquid–solid separator, biosolids from a municipal sewage treatment facility, and inorganic P applied as calcium phosphate (CaHPO4). Phosphorus sources were applied at rates of 0, 101, 202, and 404 kg total P ha−1 and incubated at 25°C. Five soil samplings were taken at 16-week intervals and analyzed for deionized water extractable P, Mehlich 3, Bray–Kurtz P1, ammonium oxalate extractable P, P saturation, bioavailable P, and anion exchange membrane extractable P. In general, the low P and fiber manures supplied the least available P, CaHPO4 the most, and medium, high, whole, and biosolids contributed intermediate amounts of P as determined by the soil P tests. The 101 and 202 kg total P ha−1 rates did not differ from each other, but were significantly lower than the 404 kg total P ha−1 rate. Except for fiber at 101 kg total P ha−1, all treatments significantly increased soil test P compared to the control. The amount of P available did not change over time except at the 404 kg total P ha−1 rate where available P usually increased with time. Correlations among soil P tests indicated significant positive relationships for each test. Bray–Kurtz P1 was most highly correlated to Mehlich 3 (r2=0.92) and least correlated to anion exchange extractable P (r2=0.13). Bioavailable P and deionized water extractable P were similarly correlated to Bray–Kurtz P1 with relatively high r2 values of 0.83 and 0.84, respectively. Ammonium oxalate extractable P and P saturation had lower r2 values (0.78 and 0.74, respectively), but were still positively correlated to Bray–Kurtz P1 and could have useful predictive value. These results indicate P availability in soil varies with the type and composition of the P source. Strong correlations among agronomic and environmental soil P tests suggest that routinely used agronomic tests or the simple DI water extractable P test could be used in place of more time consuming and expensive environmental tests to assess the P status of soils and to determine risks of various fields to release P in runoff.

Phosphorus Availability to Wheat from Manures, Biosolids, and an Inorganic Fertilizer

A greenhouse study was conducted to assess phosphorus (P) availability (as measured by plant uptake) from various P sources using Kaskaskia winter wheat (Triticum aestivum L.). The phosphorus sources included feces with low, medium, and high P concentrations from a dairy feeding study, whole (unfractionated) manure, fiber manure from a liquid–solid separator, biosolids from a municipal waste treatment facility, and calcium phosphate (CaHPO4). All P sources were applied at 0, 101, and 202 kg total P ha−1 to a sand medium. A randomized complete block design with four replicates was used. Top growth was harvested three times and roots were collected at the final harvest. An identical set of treatments was applied to a growing medium of sand plus 5% Ringwood silt loam soil. Additionally, a sub-experiment was conducted using Ringwood silt loam soil as the growing medium with whole manure, biosolids, and CaHPO4 as the P sources. Wheat total P uptake and total dry matter yield were determined for each treatment. In the sand medium, whole and low P manures produced the lowest P availability; medium P, high P, and fiber manures had intermediate P availability; and biosolids and CaHPO4 had the highest P availability. In contrast, the sand+soil experiment showed that CaHPO4, biosolids, and fiber manure produced the lowest P availability, while the other P sources (whole, low P, medium P, and high P manures) were significantly higher and similar to each other. In the soil sub-experiment, biosolids and whole manure produced significantly higher total P uptake and total dry matter yield than CaHPO4. In these studies P availability from sources with high soluble P contents was reduced in soil or sand+soil growing media possibly by reaction with Fe, Al, Mg, or Ca in the soil or adsorption of P to clay and mineral surfaces. This observation emphasizes the importance of growing media characteristics in greenhouse experiments to assess P availability. These results suggest that soluble P sorption by soil, available P contributions from soil, and mineralization of organic P from the P source treatments masked initial differences in P availability among these P sources.

Measuring Water-Extractable Phosphorus in Manures to Predict Phosphorus Concentrations in Runoff

Water-extractable phosphorus (WEP) in manures can influence the risk of phosphorus (P) losses in runoff when manures are land applied. We evaluated several manure handling and extraction variables to develop an extraction procedure for WEP that will minimize pre-analysis manure-sample-handling effects on WEP measurements. We also related manure WEP determinations to runoff dissolved reactive phosphorus (DRP) concentrations found in previously conducted field simulated rainfall experiments using the same manures to evaluate WEP as a predictor of P runoff losses. Dairy and poultry manure WEP concentrations increased with manure-to-water extraction ratio and shaking time. Relative to fresh manures, drying and grinding dairy manures before analysis usually decreased WEP concentrations, while WEP in poultry manures was often increased. Pre-analysis handling effects on WEP were minimized at the 1:1000 extraction ratio with a 1-h shaking time. Relationships between manure WEP and runoff DRP concentrations were strongly influenced by season of year and WEP extraction procedure. The best prediction of DRP concentration in spring runoff experiments was with manure WEP concentration at the 1:1000 extraction ratio. With fall runoff studies, DRP concentrations were best predicted with WEP application rate rather than concentration. These seasonal differences can be explained by the greater percentage of rainfall that ran off in the fall compared to the spring. For all studies, runoff DRP concentrations were strongly related (r2 = 0.82) to the ratio of runoff to rainfall volumes, confirming that models need to take runoff hydrology into account as well as manure WEP in P-loss risk assessments.

Scale-of-measurement effects on phosphorus in runoff from cropland

Phosphorus (P) risk loss assessment tools such as P indices are usually developed from small-plot scale data showing the relationships between various site and management variables and runoff P losses. Little information is available on how small-plot runoff composition compares with field-scale measurements. This study was conducted to compare runoff volume and composition measurements at the field scale with those obtained from natural runoff at the small-plot (1 m2 [10.8 ft2]) scale. Sediment, dissolved P, and total P in natural runoff from small plots located in two fields (7.2 and 12 ha [17.8 and 29.6 ac]) were compared with similar measurements from the fields over an 18-month period. Runoff from small-plot rainfall simulations in both fields was also analyzed for P and sediment. The fields, cropped with either corn (Zea mays L.) or alfalfa (Medicago sativa L) interseeded with bromegrass (Bromus inermis), were located on a Tama silt loam soil (fine-silty, mixed, superactive, mesic Typic Argiudoll) in southwest Wisconsin (42°42′ N, 90°22′W). Statistical analysis using repeated measures showed no significant differences between the two scales of measurements for dissolved P concentrations in runoff. Total P concentrations in small-plot runoff were greater than those in field runoff. Runoff volume and dissolved P concentrations were greater in winter than in summer, but summer runoff had higher sediment concentrations. Small plots had greater cumulative runoff volumes per unit area in both seasons compared to the fields. The dissolved P concentration relationships between the two scales of measurement for individual runoff events were very good in the corn field (r2 = 0.90) but not in the alfalfa field (r2 = 0.09). Sediment P enrichment ratios varied by crop and were similar in the small-plot and field runoff. Cumulative runoff-dissolved P concentrations were strongly related (r2 = 0.95) to average soil-test P at both scales. The agreement of small-plot and field runoff-dissolved P concentrations during the 18-month measurement period supports use of small-plot data in P loss risk assessment tools.

Soil Organic Matter Dynamics in the North American Corn Belt: The Arlington Plots

The plots are located at the University of Wisconsin Arlington Agricultural Research Station in south central Wisconsin, USA, approximately 25 km north of Madison (43°18′N; 89°21′W). The climate is cool temperate and arable crops form the present ecosystem. Monthly mean temperatures are -9.1 and 21.8°C for January and July, respectively. Mean annual precipitation totals 79.1 cm. The climate of this area normally provides 47.2 cm of precipitation and 2,570 growing degree days (50°F base) from May 1 to September 30, which closely matches the normal frost-free period of 172 days (data for 1961–1990). Meteorological data from the site (1958–1991) includes daily rain and air temperature. Daily records containing solar radiation, wind speed, relative humidity, dew point and potential ET data are provided from Madison.