Crop production and environmental impacts under organic management with reduced tillage and diversified cropping

Abbreviations

During the past few decades, organic crop production has become increasingly widespread. The global increase in organic demand is attributed to factors such as increasing concerns about food safety and consumer health, environmental issues, rising input costs, and declining soil productivity and quality (Reganold and Wachter, 2016). Studies in the Canadian Prairies have reported that organic cropping systems have higher net returns and are more energy efficient than non–organic systems (Entz et al., 2005; Zentner et al., 2011a, b).

Crop yields are often reported to be lower with organic than conventional management (Benaragama et al., 2016a; Berner et al., 2008; Campiglia et al., 2015; Ponisio et al., 2015). However, a survey of organic farms in eastern regions of the Canadian Prairies revealed that grain and forage yields ranged from one-half to almost double conventional yields (Entz et al., 2001). Similarly, studies in the U.S. have reported no consistent differences in the yield of wheat among management systems (Cavigelli et al., 2008) and similar or greater grain productivity and superior quality of organic winter wheat compared with well-fertilized winter wheat grown in a no-till cropping system (Miller et al., 2008). These highly variable yield and quality differences reported between organic and non-organic systems appear to depend on the specific management practices employed and the site characteristics (Campiglia et al., 2015; Seufert et al., 2012).

Compared with conventional high-input management systems for annual crops, soil fertility is often considered to be lower under organic management. Deficiencies in available N, P, and S in organic cropping systems have been reported in Manitoba and Saskatchewan (Entz et al., 2005; Knight et al., 2010). However, in a U.S. study, N increased significantly after 22 years in organic animal and legume systems (Pimentel et al., 2005). Soil C has also been reported to be higher in organic than conventional plots (Gadermaier et al., 2011; Pimentel et al., 2005) attributable to the higher amounts of plant residues being returned to the soil.

Weed management is reported as one of the most difficult challenges under organic conditions. However, organic systems may be able to tolerate a greater abundance of weeds compared with conventional systems due to differing weed species compositions and differences in weed-crop competition influenced by fertility management (Ryan et al., 2009). Benaragama et al. (2016b) also showed that in the absence of weeds, the organic systems still produced 44% lower yields than the conventional systems, suggesting that the effect on yields could not be attributed to weed competition but was mostly attributable to lower soil productivity. Armengot et al. (2015) also reported that despite weed increases in the reduced-tillage treatments in an organic trial, yields of various crops were similar for reduced and conventional tillage.

In the Canadian Prairies, organic agriculture has traditionally relied on summer fallow and mechanical tillage to provide enhanced soil moisture and nutrients, mitigate the impact of weeds, and reduce crop diseases; however, in recent years, there has been a substantial increase in the use of legume green manure, diversified crop rotations, and reduced-tillage management. Conservation tillage has been shown to improve soil quality and lower the environmental impact of crop production. Under organic management, reduced tillage could help improve soil fertility and increase water retention, yields, and nutrient uptake (Krauss et al., 2010). Other benefits of reducing the frequency of tillage in organic systems include higher soil organic C and microbial biomass (Carr et al., 2013). However, adoption of reduced tillage by organic farmers has been slow due mostly to concerns about nutrient supply and weeds that may limit crop yields (Cooper et al., 2016; Peigne et al., 2007).

The inability of organic no-till cropping systems to control perennial weeds has been identified as their greatest shortcoming (Carr et al., 2013; Mirsky et al., 2012). According to a meta-analysis of organic systems by Cooper et al. (2016), weeds were about 50% higher when tillage intensity was reduced although this did not always impact yields. The lower yields with reduced-tillage organic systems have been correlated with higher weed pressure as well as lower N availability in spring and early summer due to retarded mineralization (Sans et al., 2011). However, several studies have found that weeds might be managed in organic no-tillage systems through suppression by surface crop residues and weed seed exposure to environmental extremes (Anderson, 2005; Halde et al., 2015; Vaisman et al., 2011).

Diverse crop rotations can reduce weeds by disrupting their life cycles. In a study in the Central Great Plains, Anderson (2010) reported lower weed density when there were more than two crop types being rotated. In a meta-analysis, Ponisio et al. (2015) found that yields under organic management were 19% lower than under conventional management, but through the use of multiple crop types and rotations, the yield gap could be reduced to 8 to 9%.

The objectives of this multi-year study conducted in the Brown soil zone of Western Canada were to determine if diversified crop rotations and reduced tillage under organic management can maintain soil fertility and quality at adequate levels, keep weed populations at low levels, and foster healthy plants for sustainable and profitable production of annual crops. Results from this study will also be of benefit to non-organic producers, who are increasingly interested in reducing their reliance on non-renewable energy inputs, lowering input costs, reducing greenhouse gas emissions, and finding alternate methods to control herbicide-resistant weeds.

Materials and Methods

This study was conducted near Swift Current, SK on an Orthic Brown Chernozem with silt-loam texture (Fernandez et al., 2016). The land had been organically managed under green manure (GM) for two years prior to the initiation of the study in 2010.

The experiment included two crop sequences × two tillage systems. Cropping sequences were simplified rotation (SR): GM–spring wheat; and diversified rotation (DR): GM–oilseed–pulse–spring wheat. In the DR rotation, the oilseed crop alternated between flax and mustard while the pulse crop alternated between field pea and lentil. All phases of the rotations were present each year. Tillage systems were high tillage (HT) and low tillage (LT), designed to reduce tillage frequency to only as needed to prepare the seedbed or control weeds. In the spring of all years, the HT plots were cultivated at least once with a cultivator, and/or tandem double disc, while the LT plots were either worked with a rotary hoe, a cultivator, or a rodweeder. In 2013 and subsequent years, tillage was increased in the LT, and fall tillage was added to all plots, due to an increase in perennial thistle populations. In all years, the forage pea was incorporated with two passes of a tandem double disk in the HT plots and by mowing in the LT plots. The tillage-rotation systems were arranged in a randomized complete block design with four replicates.

In the spring of each year, soil samples were collected and analyzed for NO₃–N and P. In the spring of 2013 and 2016, soil samples were also taken to determine soil organic C. In the fall of 2014, soil samples were collected for determination of dry aggregate distribution, which was determined by dry sieving. Wet soil aggregate stability was determined using the 0.83–2 mm (Fraction C) aggregates. From 2013 to 2015, at maturity, plants were cut at ground level from small areas on both sides of the middle rows to determine crop and weed biomass. Grain protein was determined by near-infrared scanning (Fernandez et al., 2010).

Results and Discussion

The total precipitation ranged from above average in the 2010 and 2011 crop years (>140% of long-term mean) to slightly below average in 2013 and 2015 (95 and 90%, respectively). In May–June of 2015, the area received less than 40% of normal precipitation; however, the July precipitation was 150% of normal. In contrast, in 2012, precipitation during May–June was 165% of normal, whereas precipitation during July was only 41% of the long-term mean. Except for 2015, June precipitation was favorable for crop growth (greater than 140% of the long-term mean). Average temperatures from May to July were lowest in 2010 (56.5°F) and highest in 2015 (59.9°F). Overall, crop growth stress was rated as high in 2015, average in 2013 and 2014, moderate in 2012, and low in 2010 and 2011.

Soil NO₃–N content tended to be lowest in the last few years of the trial (Table 1). In most years, there were significant differences in soil NO₃–N among the tillage-rotation systems. For all years combined, soil NO₃–N levels in the plots to be seeded to wheat were significantly higher under HT than LT for the SR only and were significantly higher in SR than DR under HT only. Whenever there was a significant difference between HT and LT in the other crops, the NO₃–N levels were higher under HT than LT (data not presented). Crop residue incorporation in the soil with HT likely increased microbial activity, which in turn, increased residue decomposition and N mineralization. Reduced tillage has been previously associated with lower soil NO₃–N levels (Burgess et al., 2014; Peigne et al., 2007; Vaisman et al., 2011).

There were few differences in soil PO₄–P in the 0- to 15-cm depth among individual years (data not presented). Overall, soil PO₄–P content was higher in the first three years than in the last three years of this trial. Contrasts among tillage-rotation combinations also revealed that there were significantly higher soil PO₄–P levels in plots following a pulse crop in LT-SR than in HT-SR. Higher soil-available PO₄–P under LT than HT agrees with the findings of Gadermaier et al. (2011).

Soil organic C in the 0- to 15-cm depth changed between the springs of 2013 and 2016 (data not presented). For all treatments combined, there was a 2.2% decrease under HT, mostly due to a decrease in the GM plots in the SR of 15.3%. Overall, for all treatments under LT, there was a mean increase of 2.1% between the two years. Increasing organic C in the LT over time was expected since conservation tillage enhances C sequestration by reducing the rate of soil organic matter decomposition (Berner et al., 2008; Emmerling, 2007).

In both 2013 and 2016, there were overall no significant differences in soil organic C among tillage-rotation systems. However, in 2013, there was significantly higher organic C in HT than LT following wheat in the SR (21.1%). For the treatments combined in 2013, organic C was significantly higher in the DR than the SR under LT by 14.7%. The greater diversity of crop residues and root exudates in the DR than in the SR treatment may result in greater C inputs (McDaniel et al., 2014). In 2016, soil organic C tended to be higher under LT; however, the difference was not significant.

Soil aggregate distribution differed for most fractions between LT and HT (data not presented). The fine fractions were present at lower levels, and the large fractions at higher levels, under LT than HT in the DR. Higher fine fractions of dry soil aggregates under HT indicate a higher potential for soil erosion under more intensive tillage management and agrees with the findings of Malhi et al. (2009). The wet aggregate stability was significantly higher in LT than HT in the DR. Previous studies have also shown that conservation tillage provides greater wet soil aggregate stability in the surface layer than conventional tillage (Emmerling, 2007; Peigne et al., 2007).

Higher-than-average moisture conditions in most years of this study resulted in high weed biomass (data not presented). Overall, this was highest in 2012 and 2014 and lowest in 2013. The weed biomass was significantly lower under HT than LT in 2012, 2014, and 2015. The opposite was true in 2011 in SR. Comparison between SR and DR showed no differences in weed biomass under LT while under HT, it was lower in SR than DR in 2012 and 2014.

Perennial thistles, first noted in 2012, became more pronounced in 2013 (data not presented). Overall, there was an increase of 12 times their density from 2013 to 2014. Combined thistle density was significantly higher for LT than HT. The increase in perennial thistles, especially under LT, agrees with Armengot et al. (2015) who reported they almost doubled over time under reduced tillage. In contrast to weed biomass, which varied with environmental conditions and competitiveness of the crops, perennial weeds were affected most by tillage.

Wheat yield decreased steadily over the years (Table 2). Wheat yield was lowest in 2015, the least favorable year due to a dry spring/early summer. In the first year, the organic wheat yield was higher than wheat yields in similar cropping sequences in a nearby conventional no-till study. Although yields were lower in the organic than in the conventional wheat in the following years, initially, the differences were not large even though the latter had fertilizer and herbicides applied. Similarly, comparison of yields with those of the same cultivar in commercial fields in the region revealed an initial higher yield of the organic wheat (131% in 2011) followed by lower yields that differed increasingly over time from the commercial wheat (88, 83, 65, and 61% in each subsequent year, respectively). Because of lower performance of the organic wheat in the last years, over the full period, yields were an average of 75% of the conventional no-till wheat and 85% of the commercial crops.

Although yields varied among years, wheat in HT-SR displayed the most stable yields (Table 2). There were significant differences among tillage-rotation systems in all years, except 2013. Over all years, yields were significantly higher in HT-SR than in LT-SR and HT-DR with the lowest yield in LT-DR. The latter represented only 59.1% of the yield in HT-SR. While the overall differences in weed biomass between SR and DR were not significant, yields were significantly higher under SR than DR, regardless of tillage, in most years and for all years combined. Similarly, regardless of crop rotation, wheat in HT had higher yields than in LT. Even though tillage was increased in LT starting in 2013, yields were consistently highest, and in most cases weed biomass was lowest, under HT. The lower yields under LT coincided with lower soil NO₃–N, and/or higher weed biomass.

Average grain protein in the organic trial was 14.2% (data not presented). In 2011, the highest mean protein coincided with the highest yield (Table 2) while 2014 had the lowest protein and the second lowest yield. In 2013, average protein was also lower than the five-year average, and yield was near average. In 2012 and 2015, protein was above average but yield was below average. There was no negative association between yield and protein among years. This might be explained by the release of mineralized N throughout the season from the previous GM or pulse crop. Low temperatures and water deficits after flowering have been reported to negatively affect protein in organic winter wheat (Casagrande et al., 2009); this might help explain the lower protein in 2014 (below-average temperature) and higher protein in 2015 (high July temperature and precipitation).

Comparison of grain protein in our trial to commercial wheat in the region showed that levels in organic wheat were higher by 9 to 26% in most years. However, in 2014, the organic wheat protein averaged 10% lower than in commercial fields. In comparison to the nearby conventional no-till study, protein in organic wheat was higher in the first three years and slightly lower (by 3%) over the full period. The observation that protein in organic wheat was higher than in the conventional zero-till wheat suggests that the former was more efficient at producing higher protein grain.

For individual years or years combined, there were no significant differences in protein among tillage-rotation systems or these were not consistent (data not presented). In 2013, for both rotations combined, average protein was significantly lower under LT than HT while in 2015, it was higher under LT than HT. Similar protein in LT and HT, despite lower spring soil NO₃–N under LT than HT, suggests that there might have been further N uptake by the crop later in the season.

According to linear regressions, precipitation in the growing season and previous 12 months explained 13 and 35%, respectively, of the yield variation while spring soil NO₃–N explained 22%. In contrast, weed biomass explained only 5% of the yield variability. Cavigelli et al. (2008) also reported that N availability explained more of the variation in yield than weed competition. Other factors also likely affected yields, including crop management, non-N factors, and crop diseases (Fernandez et al., 2014a, b).

Conclusions

Although intensive tillage (HT) and a two-year rotation of wheat with GM (SR) promoted higher N availability resulting in greater yields, a depletion of N over time was apparent. The decrease in yield, whether due to nutrient levels and/or weed infestations, suggests that growing wheat organically alternated with GM, although resulting in acceptable yields in the short term, might not be enough to maintain productivity in the longer term. More diversified rotations with GM every four years would be even more inefficient at achieving yields compared with conventional production. However, the lower-yielding LT and DR tended to have higher soil organic C in addition to fewer erodible particles, and more water-stable aggregates, increasing the soil's resistance to wind and water erosion and contributing to the environmental sustainability of organic production under reduced tillage. Thus, a different strategy for increasing soil NO₃–N would be needed without compromising soil quality.

This study was conducted in the Brown soil where precipitation is traditionally limited and variable but coincided with above-average precipitation. Low tillage did not appear to be viable under those conditions for more than a few years, after which intensive tillage would be needed for adequate perennial weed control. Under reduced tillage, the occasional use of more intensive and frequent tillage could help mitigate the potential adverse effects of perennial weeds before their levels become unmanageable, in addition to contributing to increased N mineralization. Therefore, flexibility remains necessary regarding reduced tillage in organic systems when perennial weeds need to be better controlled to minimize their effect on crop growth.