Harnessing livestock and microbes to improve rangeland productivity and soil health

Research Design and Set-Up

Our seed and soil-based treatments are designed to increase seed and developing seedling access to water and beneficial microbes, while also reducing weed cover and enhancing soil physical integrity (Fig. 3). Our treatments include: (1) seeding (applying a diverse seed mix of desirable rangeland plant species appropriate for the region) which adds plant propagules where the soil seedbank may be lacking; (2) live soil transfers (applying a “bulked” soil mix that has high concentrations of site-appropriate beneficial soil microbes) which provides developing seedlings access to beneficial soil microbes; (3) seedballs (small structures formed from seeds, clay, and soil mixtures) which provide seeds with increased protection, access to water through added clay content, and direct contact with microbes through the addition of live soils (when added); (4) soil pits (shallow depressions in the soil surface) which trap and retain water for both seed and developing seedlings. At each of the 10 RestoreNet sites, we will test seeding alone, seedballs alone, seeding + soil pits, and seedballs + soil pits. Each of these treatments will be tested with either live soil or sterile soil (Table 2). Treatments will occur within 2x2-m randomly assigned plots within the existing RestoreNet site. Each RestoreNet site will contain 10 soil and seeding treatment combinations with four replicates of each treatment both within the existing RestoreNet sites (40 treatment plots/site). Additionally, targeted grazing (high intensity, short duration cattle grazing after seeding but before seedling establishment) will be applied at 5 of the 10 selected RestoreNet Sites (Table 1). Targeted grazing will help reduce weed cover and provide microtopography for seedling establishment.

Scalability - Evidence suggests soil-based restoration strategies may benefit recovery of larger acreages even when deployed only in smaller patches. Deploying restoration treatments in small patches or “islands” within degraded ecosystems is a promising strategy for scalable dryland restoration (Hulvey et al. 2017; Shaw et al. 2020). When deployed in small patches within degraded landscapes, restoration treatments may create nucleated islands of resource availability that can support additional spread of plant (and microbial) propagules across the broader landscape (Hulvey et al. 2017). Moreover, the proposed soil-based treatments (i.e., live topsoil inoculation, seed balls, and soil pits) can all be implemented at large spatial scales. For example, managers in southwestern Colorado have developed custom-made land imprinters for deploying soil pits across large acreages. Similarly, live topsoil inoculation involves only minimal disturbance and bulked inoculum can be readily deployed using industrial equipment and seed balls can be successfully distributed at-scale in a variety of methods including by horseback and aerially via drones (Gornish et al. 2019).

Seeding treatments will involve application of seed mixes with species proven to be effective in our precipitation-limited project region using established RestoreNet protocols (Laushman et al. 2022) at recommended seeding rates. Seed mixes will be pre-measured and individually bagged per plot. Seeding will occur in plots on its own, or in combination with other treatments (Table 2). Live soil transfers will involve collecting topsoil from each site's paired reference site and “bulking” the topsoil in the NAU greenhouse to increase soil quantity and microbial concentrations without further disturbing the reference sites (Table 3); the resulting live soil mix will be sprinkled approximately 2cm deep across the treatment plots receiving live topsoil. Live topsoil includes beneficial microbes that may increase seedling emergence, survival, and biomass of desirable seeded species (Koziol et al. 2022; Muñoz-Rojas et al. 2016). Treatments receiving sterile soil will receive the same quantity of soil substrate, except completely sterilized (Table 2). Seedballs will be made following a protocol by Co-PI Gornish (Table 3) and applied in the treatment plots at a rate of 10 seedballs/m². Seedballs are designed to keep seeds protected until germinating rains occur, provide contact with live soil and microbes, and contain clay that creates a moist environment for seeds to germinate (Gornish et al. 2019). Seedballs are made with either live or sterile soil (Table 2). Soil pits and shallow, flat depressions in the soil (4 pits/plot) designed to trap and retain water (Rorive & Bainbridge, 1993); pits were found in the previous iteration of RestoreNet treatments to be highly successful for establishing seeded species (Havrilla et al. 2020). Targeted grazing. At 5 sites, we will employ targeted high-intensity, short duration “flash grazing” treatments (Lemus, 2011) early in the growing season. Targeted grazing stocking rates (AUMs) will be collaboratively determined with local producers using publicly available stocking rate calculators. During our biannual vegetation monitoring, we will also count the number of cattle hoofprints and dung at the site to further estimate utilization. Treatments will aim to maximize the reduction of weed cover, create hoof action and organic additions, and mitigate potential harmful effects on soil quality associated with longer-term intensive grazing (Lemus, 2011).

Data Collection & Monitoring

Objective 1: Plant Establishment and Vegetation Biomass: We will follow well-established RestoreNet protocols for vegetation monitoring (Laushman et al. 2022). All 2x2-m plots will be permanently marked with plot number and treatment type upon installation and monitored each spring and fall to estimate plant establishment and growth over 2 years. Monitoring will begin with the fall immediately after seeding, to capture initial plant germination. At each monitoring visit, we will estimate seedling density, cover, and height of each species inside plots. We will use seedling identification field guides we have previously developed for the RestoreNet sites to ensure accurate identification of seedlings. In the final year of the experiment, vegetation from each plot will be clipped at peak production to estimate total biomass for each treatment. Monitoring/sampling will be accomplished by the technician, CSU graduate student, with help from RestoreNet partners.

Objective 2: Soil Health: Site soil characteristics will be measured at each site prior to treatment installation (Summer 2023) to understand the baseline site conditions. Soil site characteristics include: soil texture, aggregate stability, bulk density, and chemical/nutrient characteristics of the disturbed sites as well as the paired ‘reference’ sites. These characteristics will be examined at the site level by collecting 3 replicate samples within each of the 10 RestoreNet sites and at each of their paired ‘reference’ sites (10 sites x 2 levels (disturbed vs. reference) x 3 reps = 60 samples). Most of these baseline measurements would not be expected to change due to treatments, but rather serve to help us characterize each site and reference site (from where we source the live topsoil transfers).

Soil health indicator measurements that we expect to change in a relatively short period of time in response to our treatments include: soil moisture (measured with a handheld probe), soil physical and biocrust cover, and aggregate stability (Table 4). These soil health indicators will be estimated at the same time as fall vegetation monitoring by the field technician and the NAU graduate student, beginning one year after treatment installation (Fall 2024 & 2025) within all plots. Biocrust cover will be visually estimated and a pocket penetrometer will be used to estimate the degree of soil compressive strength. Total soil health indicator measurement will be at the plot level: 1 composited sample/plot at each site (10 ungrazed sites x 5 treatments x 2 inoculation levels x 4 repetitions = 400 ungrazed samples + 5 paired grazed sites x 5 treatments x 2 inoculation levels x 4 repetitions = 200 grazed samples → 600 estimates taken annually.

Soil microbial communities will also be examined to further understand the effects of our treatments on soil health. One year post-treatment, we will measure the type and amount of plant-associated beneficial microbial organisms in the treatment soil and plant roots (Table 4). First, in Fall 2024, we will measure the abundance of beneficial fungi in the soil by sampling soil from each treatment plot at each site and estimating hyphal length density using standard microscopic techniques (Johnson et al. 1998). Second, also in Fall 2024, we will destructively harvest roots from select seeded plants for DNA sequencing of the root associated microbial communities. Due to the extensive resource allocation DNA sequencing requires, we have opted to compare 2 treatment levels that are expected to show the greatest effects of the treatments and the most variation from each other (live topsoil*seedball*pits vs. sterile soil*seed alone; 5 sites*2 treatments*4 reps = 40 total DNA sequencing samples). Soil sampling will be accomplished by the NAU Graduate Student and the field technician, under the supervision of Co-PI’s Gehring and Munson. Soil and root microbial analysis will be conducted at Gehring’s lab at NAU using methods employed in recent previous studies conducted in the Gehring lab (Cowan et al. 2022).

Data Analysis & Reporting

Objective 1: Plant Emergence, Establishment, and Biomass: This analysis will explore plant emergence and establishment using the metrics of density and cover by species. We will evaluate plant recruitment (seeded and unseeded) over time as a response to soil treatments and site environmental factors using linear mixed effects models that allow for flexibility in incorporating categorical, continuous, and random effect explanatory variables. Our analysis will parse apart whether biotic or abiotic factors were most limiting for plant recruitment to help us understand which soil-based treatment interventions may be most likely to succeed under different environmental conditions. Explanatory variables will include treatment type, site characteristics (e.g. soil texture, elevation), precipitation and soil moisture/temperature. We will also examine the degree of success of different seeded species. The results will be published in at least one peer-reviewed journal article, led by PI Havrilla and CSU Master’s student. The paper will also be distilled into an Extension-style write-up led by Extension specialist Gornish. Using the vegetation/forage production results (biomass), we will also perform a costeffectiveness analysis (e.g. Munson et al. 2020), which will relate treatment costs to associated outcomes. Cost-effectiveness results will be made into a factsheet for ranchers, as well as included in the extension write-up and BMP guide for managers. We will make the data publicly available and go over our findings with collaborating ranchers prior to publication.

Objective 2: Soil Health: The analyses of the second objective will focus on understanding how experimental treatments and livestock presence influence soil health indicators over time, as well as understanding the relationship between changes in soil health and changes in vegetation establishment, cover, and species. Soil moisture gathered from the continuous measurements from the soil moisture probes will be analyzed with respect to on-site rain gauges (already installed) and nearby weather stations. We will again use a mixed-effects modeling approach to examine the influence of site characteristics, soil treatment types, and grazing intensity in changes to soil health indicators. This analysis will be published in a peer-reviewed journal, led by the NAU grad student and Co-PIs Gehring and Munson. The paper will be distilled into an Extension-style write-up, led by Gornish. These results will also inform our Restoring Soil Health on Working Rangelands Best Management Practices (BMP) guide.