Final report for GW20-203
Project Information
Soil carbon (C) sequestration is a top priority for regenerative ranchers, researchers, and policy makers in California. Adaptive multi-paddock (AMP) grazing, characterized by short-duration, high intensity grazing events, is a promising grazing management strategy for increasing soil C sequestration. However, the potential of AMP grazing to sequester soil C is largely unquantified in the highly temporal and spatially variable dynamics of rangelands in the Western U.S. Rangeland systems are also socio-ecological system with complex interactions among social drivers and ecological outcomes, and very little is known about the decision making frameworks utilized by AMP ranchers regarding soil health. The goal of this project is to address these gaps with a unique, interdisciplinary approach. Soil C sequestration, water holding capacity, and other ecosystem functions (i.e. vegetation cover and biodiversity) will be measured on AMP and conventional (CONV) ranch pairs across CA. To further understand soil C stabilization and persistence (as a long term greenhouse gas mitigation strategy), a novel soil C fractionation method will be used. Semi-structured interviews with ranchers, technical assistance providers, support organizations and policy makers will also be used to address informational and network challenges facing ranchers to graze for soil C sequestration. This research will aid development of effective grazing strategies across CA rangelands to optimize soil health and rancher livelihoods, and will expand the monitoring and payment opportunities for soil C sequestration.
- Quantitative analysis of soil C sequestration and vegetation on AMP grazed ranches across CA. This Spring 2020, I will be conducting a comparative “fence-line” analysis of soil C and vegetation outcomes on AMP vs CONV grazed ranches across CA. To build on these data, this proposed project will allow me to conduct additional analyses that provide a more complete assessment of potential AMP benefits.
- Conduct interviews with AMP and CONV ranchers, technical assistance providers, and associated soil C advocates across CA. The goal of these interviews is to expand on the 11 AMP rancher interviews that I have already conducted. This is a necessary addition to understand the barriers to AMP grazing by identifying gaps in knowledge (e.g. with technical assistance providers) and identify opportunities to link grazing with soil C advocates.
- Host a producer workshop to disseminate information, research results, and facilitate knowledge exchange on grazing for soil health. I will host a producer workshop at Paicines Ranch with the support of the Central Coast Rangeland Coalition (see attached collaboration letter from CCRC Board Member Sheila Barry) and University of California Cooperative Extension (Collaborators Leslie Roche and Sheila Barry). The workshop will be structured to share research findings, perform an in-field soil health demonstration, and facilitate knowledge exchange among ranchers, technical assistance providers, soil C advocates and policy makers.
From August to December 2020, I will conduct qualitative data collection through interviews with AMP ranchers, technical assistance providers, organizational members such as NRCS and RCDs, with members of soil C advocacy organizations, and policy makers. I will also continue to process soil samples and will conduct additional lab analyses made possible through this project. In the Spring and summer 2021, I will work with a videographer to collect rancher profile footage. Fall and winter of 2021 will consist of data analysis and preparation for the producer workshop in Spring 2022. Scholarly and general audience publications will be completed by end of summer, 2022. Full timeline in Table 1.
Table 1. Timeline
Cooperators
Research
In the first year of the project, I identified seven paired ranches eligible for the study. I narrowed ranch pairs on the basis of grazing management (adjacent adaptive multi-paddock and low rotational grazing systems - chosen with previously developed criteria), land-use/inputs (no compost/fertilizer application or irrigation), and location. In Fall 2019, I conducted initial ranch visits and did "pasture walks" with each rancher to identify possible sampling locations. In concert with local extension agents, I obtained ranch shape files for each property and chose ultimate sampling locations based on topography, soil type, and slope aspect.
From Fall 2019 - Summer 2020, I conducted semi-structured interviews with ranchers following an email solicitation for project involvement to understand barriers and motivations for their grazing management related to soil health and soil C.
Five AMP/CONV ranch pairs were chosen based on these criteria. We were unable to collect enough soil samples at one site due to extremely high rock content (not reported). Sampling was conducted April-July 2020 on the four remaining paired sites in three distinct ecoregions of Northern California: Central Coast Rangeland (Site 1), Northern Coast Rangeland (Sites 3 and 4) and Sierra Nevada Foothills (Site 2)
Sampling began in April 2020 and ended in July, following COVID delays. At each ranch, I navigated to my pre-determined sampling location (described above), and constructed three transects based on hillslope catenal position (1 summit transect, 1 backslope transect, 1 footslope transect). Along each 50-m transect, I collected 25 30cm-deep cores (0-10cm and 10-30cm) and 5 1m-deep cores (0-10cm, 10-30cm, 30-50cm, 50-100cm). These sample sizes were chosen based on an a priori power analysis informed by previously analyzed CA rangeland soils (for spatial heterogeneity of soil C). This ultimately yielded 360 individual soil samples per ranch pair, and 1440 soil samples across all four ranch pairs (600 soil samples from each of the 0-10 and 10-30 depths, and 120 samples from each of the 30-50 and 50-100cm depths). Vegetation data was also collected with a combination of line-point intercept (to yield percent cover) and a nested quadrat method (to yield percent frequency of more rare species). To improve on historically inaccurate bulk density measures, I chose to use the equivalent soil mass method. Each soil sample within a depth was weighed in field, and a sub sample was taken and also weighed in field. This gives an "estimated" bulk density estimate for each sample (that will also be analyzed for C), and can be retroactively corrected with a reference soil mass. To measure soil water outcomes, 1 soil sample per transect was collected and analyzed using a series of lab instruments to calculate full water retention curves, including kSAT (saturated hydraulic conductivity), HYPROP (a soil moisture release curve instrument; Meter Group), and WP4C (a soil water potential instrument; Meter Group).
From July 2020 - February 2021, samples were air dried, sieved to 2mm, and subsamples ground on a ball-mill. Each sample was analyzed for TOC/N (total organic carbon and nitrogen) on an Elementar soliTOC EA (n=1440), completed in May 2021. After determining that the soliTOC was unreliable for N% analysis, only TOC was kept from these data. From February - June 2021, composited subsamples (n=240) were fractionated by size and density into MAOM (mineral associated organic matter), fPOM (free particulate organic matter), oPOM (occluded particulate organic matter), and DOM (dissolved organic matter) - to assess soil C persistence and formation in each fraction. Each fraction was subsequently ground and analyzed on an Elementar VarioEL for TC/TN% (total carbon/total nitrogen concentration). This fractionation method was chosen because it yields information on both soil C persistence (e.g. MAOM is a more persistent form of soil C and is thought to be more important for climate change mitigation), and how soil C is forming based on longevity of management (e.g. POM is a more "leading" indicator of soil C sequestration, while MAOM may take longer to form).
Over this time, all rancher interviews were transcribed verbatim.
July-December (2021): All composited samples were also analyzed for texture using a rapid pipette method (n=240). HYPROP samples (n=24) are currently being analyzed to build full water retention curves using a 3-step method: 1. kSAT 2. HYPROP and 3.WP4C. This method was chosen because it will yield both water holding capacity as well as water infiltration - both of which are important indicators of soil health with changing management. I finished analyzing all HYPROP samples as well as analyzing fractionated samples for TC/TN on the VarioEL.
All rancher interviews were iteratively coded into emergent themes and qualitatively analyzed.
2022-2023: I worked on statistical analyses, manuscript preparation and publications.
Current status (February 2024): I've completed statistical analyses on my soil C results and have drafted a manuscript. The manuscript was rejected after review from the first journal to which we submitted. It's currently under revision for resubmisson to another journal, planned for March 2024. I've published the rancher interview manuscript in Rangeland Ecology and Management in January 2024.
I have completed preliminary analysis of my soil C (including bulk and fractionated SOM (soil organic matter)) results using linear mixed models.
Bulk TC% varied from site to site. Site 3 (coastal range) had the highest average TC% (3.78%), and Site 1 (coastal range) had the lowest (1.65%). TC% declined with depth at all sites. Three of the four AMP grazed ranches contained significantly greater SOC (soil organic carbon) stocks in surface soils (0-10cm), and two contained greater SOC stocks down to 100cm. Bulk TC% trends generally followed SOC stocks, indicating that differences in SOC stocks between grazing management pairs were due to differences in SOC content and not changes in BD (bulk density).
The vast majority of these SOC differences between AMP and CONV ranches were in the MAOM fraction, which is the most persistent form of SOC, with implications for climate change mitigation. At three of the four sites, AMP ranches had higher MAOM-C stocks. However, the effect of AMP grazing on MAOM-C stocks was only statistically clear at sites 2 (valley range) and 4 (coastal range): AMP grazed ranches at both sites had significantly greater MAOM-C stocks in the 0-10 (site 2, p=3.52E-04; site 4, p=0.03) and 10-30cm (site 2, p=4.89E-05, site 4, p=1.62E-03) depths.
Vegetation data were analyzed using PERMANOVA and NMDS. Vegetation community by functional group varied widely by site from coastal to inland ranches. We did not detect any differences in whole plant community attributable to grazing management, though there were clear differences in percent cover of individual functional groups. At two of the four sites, the AMP grazed ranches were associated with greater perennial vs annual plant cover (sites 2 (valley range) and 3 (coastal range)), more live plant cover, and less bare ground. We did not observe differences in vegetation community composition between AMP and CONV grazing at site 1 (coastal range). At site 4 (coastal range), a large stand of perennial grasses at one transect drove higher percent cover of perennial plants at the CONV ranch compared to the AMP grazed ranch.
Results from the qualitative interview work have been published, and indicate that Holistic Management training, a decision making tool often associated with AMP grazing, works by effectively shifting ranchers' mental models. Under these new mental models, AMP ranchers are able to align their economic, ecological, and social goals in order to more "holistically" manage their ranches. They still face substantial barriers, such as lack of long-term land access. However, these novel results suggest that AMP grazing may help ranchers shift their mental models and adaptively manage their grazing in ways that could be important under uncertain conditions.
Research Outcomes
These results indicate that AMP-HM grazing in California may be a viable strategy for improving environmental outcomes in Mediterranean range systems (e.g., soil organic carbon sequestration and stabilization, perennialization, etc.). However, these results vary by site, which suggests that site specific characteristics such as climate, land-use history, soil texture and vegetation structure may mediate expected benefits from these grazing systems.
Education and Outreach
Participation Summary:
On April 21, 2022, I held a rancher workshop titled, "Adaptive rotational grazing and soil carbon sequestration," as a deliverable of this grant research. The event was held at Paicines Ranch in Paicines, CA, which is a philanthropic and research ranch on California's Central Coast. More than 150 people registered, and an estimated 130 attended, more than 70 of which were farmers or ranchers. I presented my research supported by this grant related to grazing management and soil C sequestration on CA rangelands. I also organized presentations by several guest speakers, including Jeanne Merrill of CalCAN and Dr. Chelsea Carey of Point Blue Conservation Science. I organized three field demonstrations after the presentations on AMP grazing at Paicines, soil sampling, and research and monitoring. Two handouts generated by the workshop and distributed to attendees are indicated in "curricula, factsheets or educational tools," and the two research presentations I gave that day are indicated under "webinars, talks and presentations." All of these materials were made available online after the completion of the workshop on the Central Coast Rangeland Coalition's website.