Fig. 1.

Geographic locations of the sites examined in this study.

 


Fig. 2.

Examples of contrasting weather conditions among sites 30 d before and after sidedressing (SD): (a) low cumulative precipitation (PPT)–low precipitation evenness (Shannon’s diversity index, SDI) before SD and low PPT–low SDI after SD; (b) low PPT–high SDI before SD and high PPT–low SDI after SD; (c) high PPT–high SDI before SD and high PPT–high SDI after SD; (d) high PPT–high SDI before SD and low PPT–high SDI after SD; CHU is corn heat units, Tmax and Tmin are maximum and minimum temperatures, respectively.

 


Fig. 3.

Effects of weather parameters before and after sidedressing (SD): (a) corn heat units (CHU), (b) cumulative precipitation (PPT), (c) precipitation evenness determined as Shannon’s diversity index (SDI), and (d) abundant and well-distributed rainfall (AWDR = PPT × SDI). Selected values for nb and na leading to a higher contrast between the two classes (high and low) are indicated by black circles.

 


Fig. 4.

Distribution of studies in the space according to corn heat units (CHU) and abundant and well-distributed rainfall (AWDR). Labels indicate the identification number of each study as given in Table 1.

 


Fig. 5.

Yields (mean of replicates) for each study by in-season N rate (ISNR) or 178 kg ha−1 applied at sowing with no N fertilizer at sidedressing (N-rich rate). Studies are grouped by soil texture (fine or medium) and weather parameters (low or high abundant and well-distributed rainfall [AWDR] and corn heat units [CHU]).

 


Fig. 6a.

Tree diagrams for each in-season N rate (ISNR) or 178 kg ha−1 applied at sowing with no N fertilizer at sidedressing (N-rich rate) with the effect size (Y ±SEY) for all studies grouped in fine- and medium-textured soil classes combined with low and high corn heat unit (CHU) classes. The standard error SEY is the square root of the variance Vy. Dashed lines indicate the weighted mean of Yi for a given study i for each texture–CHU subgroup.

 


Fig. 6b.

Tree diagrams for each in-season N rate (ISNR) or 178 kg ha−1 applied at sowing with no N fertilizer at sidedressing (N-rich rate) with the effect size (Y ± SEY) for all studies grouped in fine- and medium-textured soil classes combined with low and high abundant and well-distributed rainfall (AWDR) classes. The standard error SEY is the root square of the variance Vy. Dashed lines indicate the weighted mean of Yi for a given study i for each texture–AWDR subgroup.

 


Fig. 7.

Weighted means of the response ratio (RR) for subgroups of fine- and medium-textured soil classes. Error bars represent standard errors of RR in each subgroup by N rate (N rate = starter + in-season N rate or 178 kg ha−1 applied at sowing with no N fertilizer at sidedressing [N-rich rate]).

 


Fig. 8.

Weighted means of the response ratio (RR) for subgroups of high and low (a) corn heat units (CHU), (b) cumulative precipitation (PPT), (c) precipitation evenness as Shannon’s diversity index (SDI), and (d) abundant and well-distributed rainfall (AWDR = PPT × SDI). Error bars represent standard errors of RR in each subgroup at each N rate (N rate = starter + in-season N rate or 178 kg ha−1 applied at sowing with no N fertilizer at sidedressing [N-rich rate]).

 


Fig. 9.

Weighted means of the response ratio (RR) for subgroups combining (a,b) medium or (c,d) fine soil texture classes with low or high (a,c) corn heat units and (b,d) abundant and well-distributed rainfall (AWDR). Error bars represent standard errors of RR in each subgroup at each N rate (N rate = starter + in-season N rate or 178 kg ha−1 applied at sowing with no N fertilizer at sidedressing [N-rich rate]).

 


Fig. 10.

Weighted averages of the response ratio (RR) for retained subgroups: (a) all retained subgroups, and (b) details in the zone showing lower RR subgroups. Error bars represent standard errors in each subgroup and N rate (N rate = starter + in-season N rate or 178 kg ha−1 applied at sowing with no N fertilizer at sidedressing [N-rich rate]).