This paper reviews options for the design and analysis of on-farm experiments. It covers both older approaches that have been popular since the Green Revolution, and more recent developments made possible by the availability of online monitoring systems as used in precision farming. The roles of randomisation as well as of geostatistical methods of analysis for these kinds of experiments are critically discussed. Two case studies are provided for illustration.

Long-term effects of elevated atmospheric CO 2 on the ammonia-oxidizing and denitrifying bacteria in a grassland soil were investigated to test whether a shift in abundance of these N-cycling microorganisms was responsible for enhanced N 2 O emissions under elevated atmospheric CO 2 . Soil samples (7.5 cm increments to 45 cm depth) were collected in 2008 from the University of Giessen Free Air Carbon dioxide Enrichment (GiFACE), a permanent grassland exposed to moderately elevated atmospheric CO 2 (+20%) since 1998. GiFACE plots lay on a soil moisture gradient because of gradually changing depth to the underlying water table and labeled as the DRY block (furthest from water table), MED block (intermediate to water table), and WET block (nearest to water table). Mean N 2 O emissions measured since 1998 have been significantly higher under elevated CO 2 . This study sought to identify microbial and biochemical parameters that might explain higher N 2 O emissions under elevated CO 2 . Soil biochemical parameters [extractable organic carbon (EOC), dissolved organic nitrogen (DON), NH 4 + , NO 3 À ], and abundances of genes encoding the key enzymes involved in ammonia oxidation (amoA) and denitrification (nirK, nirS, nosZ) depended more on soil depth and block (underlying soil moisture gradient) than on elevated CO 2 . Ammonia oxidation and denitrification gene abundances, relative abundances (ratios) of nirS to nirK, of nosZ to both nirS and to nirK, and of the measured soil biochemical properties DON and NO 3 À tended to be lower in elevated CO 2 plots as compared with ambient plots in the MED and WET blocks while the DRY block exhibited an opposite trend. High N 2 O emissions under elevated CO 2 in the MED and WET blocks correlated with lower nosZ to nirK ratios, suggesting that increased N 2 O emissions under elevated CO 2 might be caused by a higher proportion of N 2 O-producing rather than N 2 O consuming (N 2 producing) denitrifiers.

No abstract

The objectives of this study were to compare measurements of retention and prececal (pc) digestibility in evaluating mineral phosphorus (P) sources in 3- and 5-wk-old broilers. A corn-soybean meal-based basal diet was used (0.35% P on DM basis). Anhydrous monosodium phosphate (MSP(a)) or anhydrous dibasic calcium phosphate (DCP(a)) was supplemented to increment the P concentration by 0.08%, 0.16%, and 0.24%. Titanium dioxide was used as the indigestible marker. Two retention trials with excreta collection from d 16 to 20 and d 30 to 34 were conducted (n = 8 birds per diet). Another 8 pens of 10 birds from the same hatch were allocated to each diet on d 11 or 25 each to measure pc digestibility in both age periods. After 10 d of feeding, these birds were euthanized and the content of a defined section of the terminal ileum was obtained. Percentage P retention and pc digestibility for MSP(a) and DCP(a) were calculated by linear regression analysis. In 3-wk-old broilers, P retention for MSP(a) was 70% and significantly higher (P < 0.001) than for DCP(a) (29%). Values determined for pc digestibility at the same age were very similar (67% for MSP(a) and 30% for DCP(a); P < 0.001). In 5-wk-old broilers, P retention was 63% (MSP(a)) and 29% (DCP(a); P < 0.001) and pc digestibility was 54% (MSP(a)) and 25% (DCP(a); P = 0.002). We concluded that both retention and pc digestibility can be used for evaluating mineral P sources in broilers based on a regression approach. In 3-wk-old broilers, results obtained with both approaches were the same. In 5-wk-old broilers, the ranking of the 2 P sources was also the same for both approaches. Values did not differ significantly between the 2 age periods, but further studies on the relevance of broilers' age in P evaluation are suggested.

Disease incidence and severity are often assessed by either an ordinal rating scale, e.g., with scores from 1 to 9, or a percentage rating scale. This paper compares three different rating scales regarding accuracy, precision, and time needed for scoring. Pictograms of mildew diseased cereal leaves were generated following a right skewed beta-distribution. Persons with different rating experience were asked to rate the leaves on three different scales: two different percentage scales [1%-steps (P1) and 5%-steps (P5)] and an ordinal 9-point rating scale (R9) where thresholds followed a logarithmic pattern with respect to the underlying percentage scale. A transformed value of the estimated disease severity as well as the transformed time needed to estimate per leaf was documented and evaluated using mixed models. In most cases both percent ratings performed better than the ordinal rating scale. For the time needed per leaf by the untrained group, method R9 was better. With the trained group P5 performed better than both other methods. The raters mostly preferred R9, especially when untrained. Nevertheless, the results suggest that P5 can be recommended in terms of accuracy.