Phenology and climate elements

Figure 1 presents the dynamics of climate elements, it is noted that during the vegetative and reproductive stage, Tmax ranged from 22 °C to 25 °C and Tmin between 7 °C - 12 °C and 1.5 °C -10 °C, respectively. ^{Morrison and Stewart (2002)} mention that under controlled greenhouse conditions, at temperatures greater than 27 °C flower sterility increases, while lower temperature adversely affects RG due to a reduction in the number and size of seed per silique (^{Ortegón et al., 2006b}). Seasonal rainfall (PP) and evaporation (EV) was 326 mm and 408 mm, respectively.

Figure 1. Dynamics of the weekly average maximum (Tmax) and minimum (Tmin) temperature, weekly amount of rainfall (PP, mm) and evaporation (EV, mm) during canola crop development. Montecillo, State of Mexico. Summer 2011. 

Seasonal ETc was 163 mm of which 69 mm occurred during the vegetative stage and 94 mm during reproductive stage. This indicates that evaporative demand was higher than the available water. As shown in Figure 2, the crop was under severe water deficit (higher SDH), particularly during the reproductive stage of siliques filling (period from MS80 to MT), limiting a higher expression of canola tield, as has been reported by ^{Olsson (1960)} and ^{Seyedmohammadi et al. (2011)}.

Figure 2. Dynamics of severe water deficit (SDH= ETc / PP) during canola crop development. Montecillo, State of Mexico. Summer 2011. 

The time to occurrence of phenological stages, accumulated heat (UC, °C d^-1), and seasonal crop evapotranspiration (ETc, mm) was similar between treatments. Crop emergence (E) occurred 8 days after sowing (dds), beginning of stem growth (CT) at 30 dds, flowering initiation (IF) at 47 dds, end of flowering ( FF) at 67 dds, 50% to mature siliques (MS50) at 84 dds, 80% of mature siliques at 102 dds and full maturity of siliques (MT) or physiological crop maturity (FM) at 115 dds. At each stage, UC were 107 at E, 338 at CT, 517 at IF, 727 at FF, 905 at MS50, 1094 at MS80 and 1231 °C d^-1 at MF.

Biomass and harvest index

For total biomass (BT) and harvest index (CI), ANOVA showed significant changes by effect of G, N, DH and interaction G*N*DH (Table 1). The highest BT production corresponded to CII with 469 gm^-2, which outperformed Westar and CI by 29% and 43%, respectively. Regarding N, canola with N showed higher BT (508 g m^-2) and exceeded by 92% the control (without N). Similar trends were observed by shortening DH and consequent increase in DP, as in this way the soil is covered from early growth stages, with consequent higher intercepted radiation and increased BT production. This has also been reported for sunflower by ^{Escalante (1995)} and in beans by ^{Escalante et al. (2011)}. Moreover, in relation to the interaction G*N*DH, the highest BT (920 gm^-2) was achieved with the combination CIIN + DC, followed by WN + DC and CIN + DC with 713 gm^-2 and 518 gm^-2, respectively (Table 2). This indicates that by sowing CII with DC and N was more efficient in the use of space, water, nutrients and sunlight.

As for IC, genotype CI was more efficient in distributing larger amount of MS towards the grain and showed the highest IC (24%) which was higher by 20% and 33% to IC of CII and West, respectively (Table 1). However, as will be seen later, this was not reflected in higher RG. Moreover, N influenced canola to show higher IC than control. Similar response has been reported in beans (^{Escalante and Rodriguez, 2010}) and broad bean (^{Escalante and Rodríguez, 2011}). Such behavior is probably due to the formation of more demand organs (siliques and grain), which manage the distribution of MS towards grain.

Also, shortening DH led canola to show higher IC regarding planting under wide DH. Regarding the interaction, treatment CIIN0DC showed the highest IC with 30%, followed by WN0DC with IC of 24% and CIN + DC with 22% and CIIN + DC with 21%. The lowest IC (15% -18%) corresponded to treatment CIN + DA, WN + DA, CINODA, overall G and N that combined with DA (Table 2).

Grain yield, oil yield and its components

RG and NT showed significant differences due to G, N, DH and interaction N *DH, G*N*DH (Table 1). CII genotype had the highest RG with 107 g m^-2 and exceeded by 46% and 72% to West and CI, respectively. With N, canola had the highest RG, possibly due to increased demand generated by the larger number of grains and fruits (^{Escalante and Rodríguez, 2010}). Canola planted at DC with consequent increase in DP, showed higher RG. These results contrast with those reported by ^{Díaz et al. (2011)}, who report that RG from cv Monty, decreased as DP was increased from 3 to 10 plants m^-2, due in first instance to different planting pattern that was only 0.80 m between rows and genetic material.

Note that the effect of the interaction G*N*DH was determined mostly by the interaction N*DH withn each genotype. Thus, the highest RG and RA corresponded to the combination CIIN+DC with 197 gm^-2 and 92 gm^-2, which was higher by 45 and 44%; and 74 and 73% to WN+DC and CIN+DC, respectively. The lowest values (<40 g m^-2) corresponded to plantings in DA and combinations CIN+DA, CIIN0DA, WN0DA, WN+DA and CIN0DA (Table 3). Such increases in RG and RA were associated with increases in BT, NG, NF and NT. Grain size (PMG) and CA were not affected by the interaction (Table 3).

Relations between yield and its components

Table 4 shows that RG and RA showed high relationship with NG, NF, NT and BT. In contrast, the relationship between RG with grain size (represented here by PMG) and CA was low and not significant, this indicates that both RG components exhibit stability to changes in FN and DH. Thus, to achieve increases in RG of canola, must seek plants with larger canopy (BT), higher NT, NF and NG. Note that by applying the Stepwise procedure, the variable that best explained the changes in RG and RA was NF. Thus, the model was RG= -0.49 + 0.098 NF; with R²= 0.99**; and the equation RA= 0.25 + 0.046 with R²= 0.99 NF** was the best model to estimate RG.

Water use efficiency (EUA)

Water use efficiency (EUA) for RG (EUARG), RA (EUARA) and BT (EUAB), ANOVA showed significant differences between genotypes (G), N levels, DH and by the interaction between these factors (Table 5). Table 6 shows that CII had a EUA for RG, RA and BT, higher than West and CI showing similar efficiency. Also, with N and planting at DC, EUA was higher. Such increased EUA may be related to the size of the canopy, also induced by nitrogen fertilization and by planting short distances, as was also demonstrated by ^{Escalante (1995)} in sunflower. As for the interaction G*N*DH; the treatment with highest EUA for RG, RA and BT was CIIN + DC, followed by WN+DC and CIN + DC. The lowest values corresponded to WN0DA, WN+DA and CIN0DA (Figure 3). It should be noted that overall the highest EUA was found in genotypes with N and DC (Table 6).

Figure 3. Water use efficiency for grain yield (EUARG, g m-2 mm-1) of canola genotypes in funcetion of nitrogen and planting distance between rows. Montecillo, State of Mexico. Summer 2011. 

Finally, for the study region, the highest RG and RA from CII was associated with a greater number of grains, siliques, stems and greater water use efficiency and radiation generated by a larger canopy, as evidenced by the highest BT. Also, with N and planting at DC achieved higher EUA due to the increase in the size of the canopy, as indicated by the highest BT, which could lead to increased soil cover from early growth stages and as consequence higher intercepted radiation as demonstrated in sunflower (^{Escalante, 2001}) and beans (^{Escalante and Rodríguez, 2011}).