Use of natural abundance of δ15N as indicator of long-term N management on grassland farms

Abstract

Productive agroecosystems usually rely on the addition of Nitrogen (N) generally as chemical synthetic N fertilizers and animal excreta, but inefficient N management nitrogen losses contribute to environmental pollution, global warming, and a decline in biodiversity. This thesis was designed to test the hypothesis that the natural abundance of δ¹⁵N signatures of various N pools may be used to reconstruct the efficiency of long-term N management of grassland farms retrospectively. Two types of long-term fertilizer experiments in a low mountain range pasture ecosystem at the former Grassland Research Station of the University Bonn, Germany, were considered to examine the changes in natural N isotope composition of various N pools. In a first experiment, calcium ammonium nitrate (δ¹⁵N = -1.0 ‰) was applied at a rate of 0 to 240 kg N / ha^-1a^-1 combined with a two and four cut management throughout 22 years. In a second study, a field lysimeter experiment, calcium ammonium nitrate and cattle slurry (δ¹⁵N = 8.9 ‰) were applied. N was applied at rates of 0 to 480 kg N / ha a for a time period of 22 years. Sampling different plant species and soils or particulate organic matter (>250 µm, 63-250 µm) then allowed to relate excess N fertilization and resulting N losses to the cumulative changes of δ¹⁵N signatures. Based on these N isotope data the study was extended by a survey on nine common grassland farms, located in North Rhine-Westphalia and Rhineland-Palatinate in Germany and once in Styra, Austria. These farms comprised different N management, and could be grouped into low and high N input categories thus served as an independent set of test sites to verify the use of δ¹⁵N signature as an indicator of practically relevant N management on grassland farms. Samples from these farms consisted of soil, plant biomass, hay,silage, milk, hair, urine, faeces, slurry, and manure. The results showed that the mean δ¹⁵N values of plant biomass (-1.3 to 0.95 ‰) were approximately 2-5 delta units lower compared with those of the topsoils (3.1 to 3.7 ‰), suggesting that light N was lost during soil organic matter formation. The variations in plant δ¹⁵N values can be attributed to differences in plant growth, resource acquisition strategies, or δ¹⁵N values of the soil N pools for which they compete. If higher N fertilizer rates were applied (up to 480 kg N / ha), the mean δ¹⁵N increased from -1.2 ‰ to 4.8 ‰ for plants, and from 1.8 ‰ to 6.0 ‰ for the topsoils, respectively. The largest enrichment of¹⁵N was found in samples that were manured with organic fertilizer. As total soil N contents did not change significantly during the fertilization experiment, and since N losses via leaching were negligibly small, it was concluded that beside plant uptake and removal of N with biomass, most N was lost from the field plots through the gaseous phase. Particularly, N volatilization processes may account for a significant ¹⁵N isotope enrichment. In soils, this enrichment of δ¹⁵N values was preserved, suggesting that the respective N was incorporated into soil organic matter pools with slow N turnover. The screening of different farm types confirmed a significant variation in the δ¹⁵N values of harvested plant materials (-2.18 ‰ to 6.79 ‰) and the topsoils (1.47 ‰ to 7.91 ‰). Relative to farms under low N input management, the ¹⁵N of samples taken from high N input managed grasslands were elevated by 2.8 delta units. Moreover, mean δ¹⁵N values of plant biomass, soil and hair were closely correlated with stocking rate and milk production. High N input farms with a stocking rate larger than 1.4 livestock units/ha were characterized by higher amounts of organic N return per area and hence, higher losses of N from the slurry or farmyard manure, being reflected in elevated δ¹⁵N values of the studied N pools. In turn, lower δ¹⁵N values of topsoils and plants were related to low emission fertilizer application techniques and larger distances between stable and field. The results consistently show that high N input management on grassland farms systematically influences the δ¹⁵N values of soils and plant biomass. Hence, the δ¹⁵N signatures of different farm N pools, notably plant biomass, soil and hair could be identified as useful indicators for the efficiency of N management, i.e. the δ¹⁵N tracing technique represents an attractive approach to reconstruct the retrospective intensity and efficiency of fertilizer management in agricultural practice.