Nitrat‐Austräge auf intensiv und extensiv beweidetem Grünland,erfasst mittels Saugkerzen‐ und Nmin‐Beprobung II Variabilität der NO3‐ und NH4‐Werte und Aussagegenauigkeit der Messmethoden |
| |
| Authors: | Michael Anger |
| |
| Abstract: | Nitrate leaching from intensively and extensively grazed grassland measured with suction cup samplers and sampling of soil mineral‐N II Variability of NO3 and NH4 values and degree of accuracy of the measurement methods Data from a grazing experiment — comparison of mean values, see Anger et al. (2002) — were used to estimate within‐field variability to asses the accuracy of two frequently used methods of estimating NO3 leaching on pastures: (1) the ceramic suction cup sampling (with 34 cups ha—1 minimum, calculated climatic water balance, 4 leaching periods) and (2) using the soil mineral‐N method (vertical soil NO3 and NH4 content in 0—0.9 m (Nmin) measured at the beginning and end of two winters on a minimum of 10 different areas of 50 m2 each with a minimum of 7 different sample cores). These methods were used on two permanent pastures with high mean stocking density of cattle of 4.9 LU ha—1 on 1.3 ha with N‐fertilization of 250 kg N ha—1 (= intensive I]) and 2.9 LU without N fertilization on a 2.2 ha pasture (= extensive E]). The results show that NO3 leaching on pastures was largely due to few selectively extremely high NO3 amounts under a few excrement spots — mainly urine spots — which would not be sampled representatively with an acceptable effort in a conventional grazing experiment. In both grazing treatments, very large spatial variation occurred. This was greater between the different suction cups than between the compound mineral N samples of each area. Therefore, a marked skewness and kurtosis demonstrated a non‐normal distribution of samples from suction cups, while mineral N values did not show this effect consistently. Sampling selected mostly spots without noticeable influence of excrement, but a few samples with very high values identified evidently urine spots from summer or autumn grazing. The differences in mean coefficient of variation (CV) between the grazing treatments and estimation methods were mainly based on the stocking rate and the density of excrement spots. CV values were 131 % I] / 242 % E] for NO3 leaching measured with suction cup samplers and of 71 % I] / 116 % E] for soil NO3 values and 24 % I] / 34 % E] for soil NH4 values in 0—0.9 m according Nmin‐method. Results of the Nmin method are obviously inaccurate even with a sampling intensity much greater than 70 cores ha—1; and so making an estimation of NO3 leaching by this method is unsatisfactory for pastures. Compared to this, the results of suction cup sampling are more convincing; but even with a tolerated deviation of ± 20 % from the empirically estimated average and with a 95 %‐confidence interval, the calculated mean minimum number of samples in our experiment should be increased to 146 and 265 suction cups ha—1 for the intensively and extensively grazed treatments, respectively. This requirement would be prohibitive for many field experiments. |
| |
| Keywords: | grazing nitrate leaching variability suction cups soil mineral‐N stocking density |
|
|
