Biointensive no-till farming in California: farmer-driven research and education on soil health, water efficiency and economic resiliency

Our on-farm trials and field experiments revealed rapid shifts in soil health indices and ecological functioning under no-till management. Some of the most striking features of these results have been in soil structure and hydrology, fungal community structure, and soil carbon cycling. 

Soil structure and hydraulic properties changed by farming system, with tillage negatively affecting gravimetric water content at all depths (ANOVA, F=16.5, p=0.00015), and cover cropping increasing water content by as much as 30%, especially at deeper soil depths (at 50 cm; ANOVA, F=5.6, p=0.021) (Fig. 2). Soil structure, here expressed as pore size distribution derived from a soil water retention curve, changed significantly with farming system in the second but not the first year (Fig. 2).

This could be explained by a shift in pore architecture, a unique and powerful finding that could drive many of the ecological shifts demonstrated in this paper. Figure 2 illustrates the shift in pore size distribution between 2018 (top) and 2019 (bottom). In the first year of implementing no-till management, while there were some minor changes in the total amount of water held in the soil profile, the distribution of pore classes did not change significantly. In 2019, just two years after the implementation of no-till management, the data is quite different. Minimizing disturbance and introducing winter cover crops shifted the soil matrix towards a relatively larger population of smaller pores, more effective at retaining plant-available moisture during severe conditions. Meanwhile, conservation tillage combined with continuous crop production began to shift pores in the opposite direction, opening up larger pores that freely drain or evapotranspire more easily. This physical shift in soil structure was realized in a relatively short time frame, with significant implications for water conservation, reduced irrigation, and more. Pore size mean, median, and mode were all demonstrably larger in tilled systems, illustrating the physical creation of soil cavities by the tractor during tillage. 

[Figure 2. Soil hydrological properties, demonstrating a) soil moisture content at four depths (5 cm, 15 cm, 30 cm, and 50 cm), showing higher quantities of water stored under no-till management and b) pore size distributions evaluated in 2018 and 2019, showing a rapid shift in pore architecture between the first and second year of the study. The bottom-right quadrant illustrates a shift towards small pores in low disturbance, cover cropped systems and demonstrates a physical mechanism for observed changes in soil properties.]

Fungal species richness within the soil was clearly influenced by management; both observed (F=4.9, P=0.03) and Chao1 (F=3.5, P=0.0740) richness indices were estimated as higher in continuous crop production systems, as illustrated in Figure 6. Fungal diversity was also found to be higher in the soils of high-disturbance systems, as estimated by the Shannon index. Shannon’s index accounts for species evenness, or the degree of spread in population count across the total list of species within a particular population. Keeping all else the same, higher degrees of species evenness within the soil fungal populations led to a higher Shannon index, likely driven by the homogenization of the soil profile by tillage. This demonstrates a complex interaction between disturbance, cropping system, and soil fungal species richness and diversity, with continuous plant cover emerging a key driver of fungal species richness regardless of differences in disturbance. And despite management driving fungal community structure in the soil, fungal communities within crop root systems demonstrated less pronounced response to management, with inconclusive patterns in indices of species richness and diversity. This suggests a selection bias on the part of plant roots, especially during periods of severe stress such as a drought. 

[Figure 3. Relative abundances of a) arbuscular mycorrhizal fungi (AMF) at the species level, and b) plant pathogens in the soil fungal community at the genus level, both divided by soil management system. Changes in the AMF communities between no-till and tilled systems are stark, with a shift from Claroideoglomus to higher relative abundance of Rhizophagus and Funneliformus species. There is also a significant difference across cropping systems even within no-till plots, demonstrating the influence of crop cover and crop residue recycling methods on mycorrhizal partners. These changes in the AMF ASVs found through molecular methods corroborates the low colonization counts under the no-till, cover cropped treatment. Meanwhile, soil management also had a measurable influence on fungal plant pathogens living in the soil, specifically an enrichment in the genus Cylindrocarpon, a large family of pathogens that cause root rot. Higher rates of fungal root damage were observed during the implementation of the project; these data confirm observations at the study site of plant pathogen pressure with conversion to no-till.

The relative abundance of AMF, plant pathogenic, and soil saprotrophic ASVs were all influenced by soil management system, with distinct assemblages of these fungal trophic groups across each treatment. Changes in the AMF communities between no-till and tilled systems shifted from Claroideoglomus to higher relative abundance of Rhizophagus and Funneliformus species. There is also a significant difference across cropping systems even within no-till plots, demonstrating the influence of crop cover and crop residue recycling methods on mycorrhizal partners. These changes in the AMF ASVs found through molecular methods corroborates the low colonization counts under the no-till, cover cropped treatment. Meanwhile, soil management also had a measurable influence on fungal plant pathogens living in the soil, specifically an enrichment in the genus Cylindrocarpon, a large family of pathogens that cause root rot. Higher rates of fungal root damage were observed during the implementation of the project; these data confirm observations at the study site of plant pathogen pressure with conversion to no-till.

This data is contextualized by arbuscular mycorrhizal fungi (AMF) colonization, which was most influenced by cropping system (F=7.6, P=0.01767 ), with colonization rates in the continuous, no-till farming system double that of the other systems (Figure 3). This, coupled with the changing AMF communities between farming systems, strongly suggests an ecological shift in the fungal symbionts recruited by the crops while under stress.

Soil organic carbon (SOC) increased by 207% (at 5 cm) and 47% (at 15 cm) after two years of implementing the hand-scale no-till system, with tillage systems (ANOVA, F=18.5, p<0.0001) driving changes in soil carbon stocks (Fig. 4). Carbon stock increases were more modest at depth, but were still increased by hand-scale no-till systems, including increases from 6.5 Mg C/ha with tillage to over 8 Mg/ha under no-till at 30-50 cm. Size-based fractionation of the soil revealed higher carbon stocks in all size fractions (2 mm-250 um, 250 um-53 um, 53-20 um, <20 um) when soils were managed for minimal disturbance. Higher mass percent of carbon in mineral associated fine fractions (<20 um), along with lower rates of carbon cycling enzymes in those same fractions, is possible evidence for improved soil C storage by mineral association under no-till management. While compost application can account for relative changes in soil carbon quantities by management system, it does not necessarily explain a shift in the division of soil carbon across particle size within those systems. This phenomenon, coupled with changing potential enzyme activity within classes of soil pores and changing pore architecture (reported below), suggests a shift in carbon cycling that results from minimizing soil disturbance.

Ammonium and nitrate were twice as high under no-till systems with continuous production, emphasizing the control that continuous crop cover and compost application have on plant-available N supplies. POXC, a form of labile carbon that readily reacts with oxidizing agents in the soil, also had the highest concentration in no-till systems, with an average 55% increase in permanganate oxidizable carbon from baseline in no-till plots. Meanwhile, tilled systems were depleted in labile carbon stocks from baseline by an average of 145%.

[Figure 4. Total organic carbon (TOC), mineral-associated organic matter (MAOM), and particulate organic matter (POM). Total organic carbon was 2-3 times higher under more intensive farming systems. Of significant interest are the modest increases at 50 cm depth, from about 3.25% to almost 4.5% under no-till. MAOM is a measure of fine particulate organic matter (<53 um) that is likely to be bound to mineral surfaces, decreasing the likelihood of its degradation by microorganisms. POM includes larger pieces of organic matter (>53 um) and include plant residue, partially decomposed material, and more chemically recalcitrant particles.] 

Comparing tomato dry farming between tillage systems

Tillage did not significantly affect marketable tomato yields or quality, nor significantly affect plant physiology including plant water use efficiency (intrinsic [WUEi] and long-term [δ13C]), gas exchange, stem water potential (Ψstem) and leaf relative water content (RWC). Early Girl produced significantly greater number of fruits and fewer non-marketable fruits (<6%) than Momotaro and Zapotec. AM fungi root colonization was very low for both 76R and rmc but differed significantly depending on genotype and tillage treatment. 76R produced slightly greater marketable fruit, but physiological responses did not differ significantly between genotypes.

This study provides field-based evidence that genotype has a greater impact on dry farmed tomato yield and plant attributes than site management. After five years of agroecological management and soil health improvements, tillage did not significantly influence dry farmed tomato marketable fruit yields. Genotype had a greater influence than tillage on tomato fruit quality and yields for the indeterminate tomatoes. While the grower standard, EG, had greater fruit yields than M and Z, fruit quality as represented by fruit Brix and moisture did not differ significantly between the three. Overall, Early Girl was a superior genotype, with future research needed to investigate the dry farming potential of other hybrid and heirloom genotypes. To optimise dry farming productivity as a meaningful option for food production in drier climates, strategies that increase and retain soil moisture should be combined with selection of drought tolerant, water efficient vegetable varieties.

[Figure 5. Mean fruit yield per plant for each harvest of the indeterminate tomato genotypes Early Girl (EG), Momotaro (M), and Zapotec (Z), grown with either biointensive no-till (NT) or tilled (T) soil regimes under field conditions. The first harvest occurred at 80 DAP on 29 July 2022. Shown are observed means ± se (n = 4). Means, error bars, and lines have been slightly dodged horizontally to avoid overlapping and for better visualization.]