SUBSOIL CONSTRAINTS: IMPACTS OF CROP TYPE, VARIETY AND AGRONOMY ON CROP PERFORMANCE.

 

Background

Subsoil constraints to crop production can include one or more of; salinity, sodicity, acidity, nutrient deficiency, element toxicity, gravel layers, compaction and pathogenic biological activity. The main impact on crop yields is a reduction in the effective rooting depth so that plants have less access to stored soil water. Some species also suffer toxicity damage where concentrations of chloride and sodium in soluble salts are too high for normal plant growth.

 

NSWDPI has conducted paired good-site/bad-site field trials in the northwest NSW cropping region at Bellata, Garah, Spring Ridge and Narromine. In 2004, 2005, and 2006 we compared the yield performance of a range of winter crop species and varieties on soils with and without subsoil constraints. Sorghum varieties were compared in paired-site trials in 2001/2, 2003/4 and again in 2007/8. The most recent sorghum trial also included row configuration treatments. The main subsoil constraints at the trial sites were high salinity, high sodicity, and alkalinity.

 

Winter Crop Results

In four out of six paired-site trials, wheat yields at subsoil constrained (SSC) sites were significantly less than those grown at nearby unconstrained sites. Average yield decreases ranged from 12% (Bellata 2004) to 72% (Narromine 2006). At Spring Ridge 2006 and Narromine 2006 all varieties yielded more at the “good” site than the “bad” site, but at Bellata in both 2004 and 2005 the yields of Baxter, Lang and Strzelecki were not significantly different between paired sites. Both Sunvale and Sunstate had large yield declines at the “bad” site. Sunvale was the highest yielding variety in 2004 and equal highest with Sunstate in 2005.

 

Durum wheat yields were no different to bread wheat yields in 10 of the 13 trial sites, but yielded less at the other 3 trial sites; Bellata “bad” 2004, and Spring Ridge both sites in 2006. In contrast, durums at Spring Ridge in 2004 (a “good” site only) yielded higher than three of the bread wheat varieties. Durum yields declined by 17-36% in three of the six paired site comparisons, including Garah 2005 where there was no overall effect of trial site on bread wheat yields. Durum wheat declines due to subsoil constraints tended to be greater than those for bread wheats.

 

Averaged barley yields declined at five of the six paired-site comparisons by from 12% up to 60% (Narromine 2006). There were some significant barley variety differences at each site except Garah 2006. Binalong was not only highest or equal highest yielding variety at each trial site, it was also not significantly different between “good” and “bad” paired sites at any locations. Gairdner and Fitzroy were equal highest yielding at more than half of those trials where barley varieties produced any significant differences. Only at Garah 2005 did Fitzroy show a significant yield decline between paired sites. Yields of the other barley varieties declined at three or more paired site comparisons.

 

Triticale was only included in 2006 at two paired site comparisons, with no trial or variety effect at Garah 2006, but both trial effect (40% yield decline) and variety effect at Spring Ridge 2006 (Kosciuszko > Everest at the “good” site).

 

Chickpea yields significantly declined by from 28% to 41% at four of five paired site comparisons, but increased at the other location (Garah 2006) where starting soil water favoured the otherwise more constrained site.

 

Subsoil constraints led to significant fababean yield reductions of from 26% to 31% at two of four paired site comparisons, but no difference at the other two paired sites. Fiord outyielded Cairo at three of the four trials in 2005.

 

Fieldpea yields declined by from 35% to 57% at five of the six paired site comparisons, although declines at four of the paired sites were between 35 and 38% (57% decline was at Narromine 2006). Where there were significant differences in fieldpea yields between varieties, Yarrum in each case outyielded Boreen, both at “good” and “bad” trial sites.

 

Lentil yields declined by 48-84% in the 2005 paired trial sites, but not in the two paired site comparisons in 2006, despite some trial by variety interaction effects. CIPAL414, the experimental variety included due to its supposed greater tolerance to saline soils, outyielded Digger at the Spring Ridge “bad” trial site, and also the Bellata “bad” site in 2005 (when analysed separately from the “good” site).

 

In general, chickpea and fieldpea declines were similar, averaging 66% of “good” site yields, fababean declines were slightly less, averaging 76% of “good” site yields. Lentil yields were too variable to rate as better or worse than other legumes.

 

Subsoil constraints only influenced grain yields of canola at one paired site comparison (Spring Ridge 2006) out of five, where there was a 38% decline in yield.

At Spring Ridge in 2006, we found that Ripper canola outyielded Rivette at both trial sites, but it was the reverse at Bellata in 2005, and no different at any Garah trials. Mustard yields significantly declined by 38% at the Spring Ridge 2006 trials, and variety Micky declined by 22% at Bellata 2005. Individual trial site analyses showed Micky outyielded Kaye at three separate trial sites but was no different at the others.

 

Safflower yields declined by 25-39% at two of the four paired site comparisons, but were no different at the other two locations. Gila only outyielded variety 555 at the Spring Ridge “bad” site when analysed separately from the “good” site.

 

Linseed was even less responsive to subsoil constraints with the only significant yield decline (45%) found at the Spring Ridge 2006 paired site comparison, and no difference at three other paired sites.

 

While generally more variable in their response to growth on subsoil constrained soils, decrease in grain yields of oilseeds at the “bad” site compared to the “good” was not dissimilar to that of bread wheat or barley.

 

Seventy-five percent of the variation in grain yield of all crops grown in all trials could be explained by in-crop rainfall and soil water use, when the data was grouped by species.

 

Maximum rooting depths for cereals ranged from 50 cm to 154 cm; for legumes from 34 cm to 112 cm; and for oilseeds from 59 cm to 152 cm. Safflower had the deepest or equal deepest roots in seven of the eight trials where it was grown. One or more of the cereal species had equal deepest roots at eight of the thirteen trial sites. Either canola or mustard had equal deepest roots at six of eleven trial sites. Of six paired site comparisons, only at Garah in 2006 was there no significant effect of subsoil constraints on rooting depth of any species. Of the other paired sites, only bread wheat and linseed at Garah in 2005, and durum wheat and fieldpea at Narromine in 2006 showed no reduction in rooting depth between “good” and “bad” trial sites. All other species had significantly shallower roots when grown in constrained subsoils. Roots in constrained subsoils were, on average, 32% shallower than when grown in nearby unconstrained soils (range: 11-68%).

 

No crop type (cereals vs legumes vs oilseeds) was found to be affected more than any other on average across all the paired sites where differences were found. Maximum rooting depth was closely related to soil water used, with 78% of variance in soil water used by all crops at all sites being accounted for by rooting depth when grouped by site, year and species.

 

Sorghum Results

Early sorghum variety trials on paired-site were conducted on the Liverpool Plains. While there were some variety differences in grain yields between good and bad sites, these could not be related to either water use, rooting depth or chloride in the subsoil. Yield results from the Bellata 2007-8 trial are not finalised at the time of writing, but seem to indicate greatest yield decline (between good and poor subsoil sites) when grown as a solid plant system, with “wide row” yielding best when grown on constrained subsoils.

 

In the current sorghum varieties trial, all varieties declined in yield by 7-26% between the unconstrained and constrained trial areas. There was a distinct interaction between grain yield and yield decline. The higher the observed yield of a variety at the unconstrained site, the greater the yield decline when that variety was grown at the constrained site. This observation suggests that potential yield advantages of improved varieties seen at the unconstrained site were nullified by the subsoil constraining access to available soil water at the poor site.

 

Soil water use results from the constrained site of the Bellata 2007-8 trial were not all available at the time of writing.

 

Contact details

Graeme Schwenke

Ph: 02 6763 1137

Email: graeme.schwenke@dpi.nsw.gov.au

 

Tracey Farrell

Ph: 02 6799 1548

Email: tracey.farrell@dpi.nsw.gov.au