Evaluating Rotary Cores for Sampling Exchangeable Cations, Carbon, and Nitrogen in Rocky Soils
Carrie Rose Levine, MS 2011
Digging pits for quantitative soil sampling is labor-intensive and destructive, and as a result, few pits can be excavated in a given study. The use of rotary cores for soil sampling could improve estimates of soil nutrient contents due to the speed and efficiency of obtaining many measurements at depth and over a large area. We assessed whether this method was appropriate for estimating exchangeable cation concentrations in soils.
We hypothesized that the grinding of rock and soil that occurs during coring could potentially result in a bias towards overestimating concentrations of exchangeable cations in soils. We also hypothesized that the introduction of material from the upper soil profiles to the deeper samples during insertion and removal of the corer might bias exchangeable cation, carbon (C), and N estimates in rotary core samples, as the upper soil horizons have greater concentrations of cations, C, and N compared to deeper horizons.
We compared soil mass, rock mass, and exchangeable cation concentrations in samples from rotary cores and soil pits in four soil types at four sites in the US: NH (Spodosols), NY (Inceptisols), NV (Mollisols), and CA (Alfisols). Cores were sampled to depths of 90 cm (NH), 50 cm (NY), 52 cm (NV), and 60 cm (CA). The corer apparatus was a 9.5 cm internal diameter diamond tipped core bit mounted on a rotary motor. Cores were extracted in sequential segments that ranged in depth from 8-20 cm. Concentrations of sodium (Na), magnesium (Mg), potassium (K), and calcium (Ca) were measured using a neutral salt extraction.
At the NH and NY sites, we also used a modified coring method to quantify the amount of soil introduced to samples during the insertion and removal of the soil cores, in order to estimate the bias of exchangeable cation, C, and N contents introduced during the sampling procedure.
We analyzed soil mass, rock mass, and exchangeable cation concentrations for significant differences (p<0.05) between soil pits and total cores at all four sites using a mixed linear model. The model included repeat measures to account for depth increments of individual cores and an interaction between concentration or mass and depth.
Soil mass estimated by the soil cores was not systematically different from soil pits, but the four sites showed differed patterns. The soil mass of cores and pits NY and CA sites was general slightly higher than the pit soil mass. At the NV and NH sites, there did not seem to be a strong relationship between the mass estimates from pits and cores. At the NH site, the average absolute difference in soil mass between the pits and cores was high (36%), but the average difference in mass between the NH pits and cores was not significantly different from zero, indicating that though there was not a systematic bias, the cores and pits did differ in their estimates of soil mass.
Cores did not systematically over- or underestimate rock mass relative to the soil pits at the four sites. Similar to soil mass, the sites showed different patterns between pit and core rock mass estimates. At the NY site, estimates of rock mass from cores were similar to estimates from pits. At the NV site, there did not appear to be a consistent bias for cores to over- or underestimate mass, though the core estimates tended to vary from the pit estimates. At the CA and NH sites, the soil cores tended to underestimate rock volume relative to the pits. We found no correlations between rock volume and exchangeable cation concentrations.
The difference in exchangeable cation concentrations between pits and cores was often significant at a given site, but was not consistent among sites. Concentration of K in cores was significantly higher than in the pits at all four sites. Concentrations of Na were higher in cores at the NV, NY, and NH sites, though Na concentrations were higher in the soil pits than cores at the CA site. Concentrations of Ca were higher in cores at the CA site, but higher in the pits at the NV site. Concentrations of Mg were higher in CA cores than pits.
At the NY and NH sites, concentrations of exchangeable cations, C, and N introduced during the insertion and removal of the corer were generally similar to the concentrations found in the first 30 cm of mineral soil. The mass that the contamination samples contributed to the core mass was 9% of core mass at the NH sites and 2% of core mass at the NY site. When concentrations were multiplied to an areal basis, the contamination from the insertion and removal samples represented an additional 1500 ± 484 g C m-2 and 61 ± 20 g N m-2 at the NH sites, and 177 ± 34 g C m-2 and 16 ± 3 g N m-2 at the NY site.
Discussion and Conclusions
It appears that grinding of the soil coarse fraction during the use of a rotary core increased exchangeable cation concentrations of certain elements at all four study sites, but which cations showed this pattern varied among sites. We found that the rotary coring method is problematic when sampling for exchangeable cation concentrations. In the four soil types tested, we observed elevated cation concentrations relative to soil pits, leading to large biases when expressed on an areal basis. Though the cored samples exhibited higher exchangeable cation concentrations relative to quantitative soil pits at the four sites, this method provides a convenient way to sample these soil properties, particularly at depth. Additionally, the corer is better able to capture spatial heterogeneity relative to quantitative soil pits. In cases where soils will be measured repeatedly over time, the benefits of ease of sampling and capturing spatial heterogeneity may outweigh the possible biases introduced from grinding.
Funding support for this project was provided by the Northeastern States Research Cooperative (NSRC), a partnership of Northern Forest states (New Hampshire, Vermont, Maine, and New York), in coordination with the USDA Forest Service.