Sustainable Soil Resource Management and Produce Marketing on Resource-limited Urban Farms

Objective 1: Vacant Land Identification and Suitability Assessment

Work on Objective 1 progressed well across the three reporting cycles, beginning with the identification of church-owned properties (2022–2023) and culminating in a comprehensive, GIS- and machine-learning–based suitability assessment incorporating all vacant lots in the study area (2024–2025).

During the first reporting cycle, 145 churches were identified within the target zip codes of southeast Fort Worth, of which 76 churches (52%) had at least one acre of potentially leasable land suitable for urban crop production. Ground-truthing in the 2023–2024 period refined this estimate further: based on criteria such as land condition, availability of basic infrastructure (parking, fencing, storage), and absence of conflicting land uses, 50 church properties remained viable. These properties were distributed across the study area as follows: 76119 (n = 12), 76105 (n = 22), 76104 (n = 14), and 76111 (n = 2). Collectively, these church properties represent approximately 50 acres of potentially productive urban cropland—highlighting the importance of faith-based institutions as partners in community food initiatives.

Building on these findings, the project expanded the search to include non-church vacant parcels across southeast Fort Worth by combining GIS layers, parcel attribute data from the Tarrant Appraisal District, and high-resolution WorldView-2 multispectral satellite imagery. A supervised machine learning classification (Support Vector Machine, SVM) identified an additional 248 properties unused or underutilized that may support urban agriculture. These properties were subsequently incorporated into a broader suitability assessment to quantify potential cropland availability across the entire study area.

Using six criteria - area, zoning, slope, flood risk, soil characteristics, and road proximity - the analysis classified parcels into high, moderate, and low suitability for urban agriculture. Across the region, 16 km² (3,953.69 acres) of vacant land were identified as potentially viable (Figure 3A). Although only 1.14% (1.8 acres) met all optimal criteria and were classified as highly suitable, the majority (96.6%, or 3,802 acres) were moderately suitable, meaning they could be converted into productive land with modest investment in site preparation and infrastructure. Parcels classified as low suitability represented 2.26% of the total. Recognizing that very small parcels are often impractical for production agriculture, suitability was reassessed using a minimum lot size threshold of 445 m² (0.11 acre) a common benchmark in the urban agriculture literature (ref). When applying this threshold, 15.5 km² (3,835 acres) of land remained suitable for farming, offering a more realistic estimate of land that could support food production within the targeted neighborhoods (Figure 3B).

Taken together, these results provide a robust, data-driven foundation for future urban agriculture planning in southeast Fort Worth. They also demonstrate that, although a limited number of parcels meet ideal conditions, there exists a substantial area of moderately suitable land where targeted investment in infrastructure, soil improvement, and site consolidation could enable meaningful food production efforts. An interactive version of the suitability map is publicly accessible (with a free account sign up) here. Signup for the requisite free ArcGIS account can be complete here.

Objective 2: Evaluating food-waste compost with a legume cover crop as a low-cost, sustainable soil amendment for supporting tomato growth

During the 2023–2024 reporting cycle, project activities under Objective 2 shifted toward strengthening on-farm data quality due to the discontinuation of two partner sites. These sites were removed because of agronomic failures, inconsistent research compliance, and, in one case, a credible safety concern reported by project personnel. All discontinued partners were compensated, and the research design was consolidated to a single functioning site. Two new on-farm research plots were established in December 2023 to ensure adequate replication and continuity into the 2024 growing season.

Despite early setbacks, a full dataset was successfully generated at the remaining site, including tomato yield, monthly soil chemistry (pH, EC, %C, %H, %N, %S, C:N, C:S), and monthly drone-based NDVI to assess canopy vigor and spatial variability. The dataset provides early evidence regarding the benefits and limitations of food-waste compost and legume cover crops in small-scale urban farming systems. The data set used for soil chemical assessment and ANOVA/Tukey HSD analysis are provided here: Soil_Data_ANOVA_report_2024 TukeyHSD_Treatment_Plots1to3-report 

1. Soil pH and Electrical Conductivity

Across the study, soil pH remained slightly above (0.16 - 0.21 pH units) the optimal range for Celebrity tomatoes (6.2 - 6.8), but consistent treatment differences emerged. Compost-only (COMP) treatments frequently produced small but statistically significant increases in pH (Figure 4). For example, in Plot 1, COMP had significantly higher pH than CC (p = 0.0028), CC+C (p = 0.0062), and NT (p = 0.0016). Similar patterns were observed in Plot 3, where COMP again had higher pH relative to CC (p = 0.0016). These shifts are consistent with food-waste compost containing base-forming cations that impart a mild liming effect.

Electrical conductivity (EC) did not differ significantly among treatments in any plot (all Tukey p-adj > 0.05), indicating that compost additions did not cause salt accumulation during the study period, a key concern for urban growers using food-waste-derived materials.

2. Soil Carbon and Nitrogen Dynamics

Although treatment effects on soil chemistry were generally modest, several statistically significant differences emerged in organic carbon and nitrogen. In the double-dig Plot 3 where soil structure was improved through deep loosening, cover crop alone (CC) produced significantly higher %C than CC+C (p = 0.0245) and COMP (p = 0.0004). This suggests that in physically improved soils, legume cover crops may contribute more to short-term surface soil carbon than compost additions.

Soil nitrogen (%N) differences were concentrated in Plot 1, where COMP had significantly higher %N than CC (p = 0.0247), CC+C (p = 0.0013), and NT (p < 0.0001). This pattern aligns with the rapid mineralization of food-waste compost and demonstrates the potential of compost to supply short-term soil fertility benefits.

No significant differences were detected in %H, %S, C:N, or C:S in any plot, consistent with early-stage soil amendment research where multi-year observations are typically required to detect strong treatment effects.

3. NDVI-Based Crop Health Assessment

Monthly NDVI data from drone imagery (Figure 5 and 6) provided a spatially explicit assessment of crop vigor across treatments. Across all plots, NDVI increased from March through May-June and declined by July due to heat, water stress, and tomato phenology (Figure 3). However, treatment-specific patterns were evident.

Conventional Tillage (Plot 1)

• CC+C sustained the highest NDVI values through early- and mid-season, reflecting strong vegetative vigor.

• CC exhibited consistently higher NDVI than NT early in the season.

• COMP and NT declined gradually but remained within the “moderately healthy” NDVI range.

Conventional Tillage (Plot 2)

• COMP exhibited the highest midseason NDVI, indicating improved soil fertility and water retention.

• CC and CC+C supported above-average early-season NDVI compared to NT.

  

Double Dig (Plot 3)

• All treatments displayed higher early-season NDVI than the conventionally tilled plots, reflecting the advantages of deep soil loosening.

• COMP and CC+C maintained the highest NDVI levels in June, suggesting strong synergy between improved rooting conditions and organic inputs.

Soil-Plant Correlations

Correlation analyses across all plots showed:

1. Negative relationships between pH and NDVI, indicating reduced canopy vigor as soil pH increased beyond optimal tomato ranges.

2. Positive associations between soil carbon and NDVI, especially in the double-dig plot where CC increased %C.

These findings highlight the importance of balancing soil pH, organic carbon inputs, and physical soil structure to support crop vigor in urban soils.

Tomato Yield (2023 Season)

While treatment differences in fruit count and fruit weight were not statistically significant in Tukey post-hoc tests (all p-adj > 0.05), several practical patterns emerged:

1. Total yield ranged from 70,000 to 76,050 lb/ac, with the highest yield under COMP.

2. Fruit per plant increased from NT (41) → COMP (45) → CC (46) → CC+C (48), suggesting cover crops influence fruit number.

3. Average fruit weight was highest in NT and COMP (≈6 oz) and lowest in CC and CC+C (5.2–5.4 oz).

Together, these early patterns suggest that compost tends to improve fruit size and total yield, whereas cover crops increase fruit number. Combining both practices may balance these effects, although additional years of data are needed.

Across soil chemistry, NDVI, and yield data, the earliest and most consistent impacts were observed in:

1. pH and nitrogen (driven by compost inputs),

2. soil carbon accumulation (driven by cover crops, especially in deep-loosened soils),

3. canopy vigor (enhanced by combined compost + cover crop treatments),

4. root-zone improvement (enhanced by deep-digging).

These results indicate that food-waste compost and legume cover crops are accessible, low-cost strategies that can improve soil health and crop vigor in resource-limited urban farming environments. Importantly, the data also show that soil physical improvements (deep digging) amplify the benefits of biological amendments.

Objective 3: Evaluating the Feasibility of a 75:25 For-Profit : Non-Profit Sales Model

The soil amendment research under Objective 2 directly informs the project’s broader goal of strengthening sustainable urban agriculture systems in Southeast Fort Worth. While Objective 2 focused on biological and soil management strategies that increase crop productivity and ecological function, Objective 3 shifts the focus to the economic and social sustainability of small-scale urban farming, including how farmers interact with markets, consumers, and community needs. By combining soil-health research (Objective 2) with market and community analysis (Objective 3), the project provide insights into both the production and distribution challenges that urban farmers face - increasing the probability that technical improvements in soil management translate into meaningful economic and social benefits.

The goal of Objective 3 was to evaluate whether a 75:25 revenue model, where a portion of sales supports subsidized or low-cost produce for underserved communities, could function sustainably within the context of small and urban farms in Fort Worth. Using a mixed-methods approach that incorporated farmer interviews, customer surveys, regional farmers’-market assessments, and community conversations, the project examined feasibility from three essential perspectives: the community, the farmer, and the marketplace.

1. Community Perspectives on Food Access and Local Markets

Roundtable discussions and community conversations revealed significant structural barriers affecting food purchasing behavior in East and Southeast Fort Worth. Residents consistently identified lack of full-service grocery stores, limited transportation, and the high prevalence of fast-food and convenience stores as key obstacles to healthy eating. While some participants used food pantries or church-based programs, stigma, pride, and inconsistent quality limited widespread utilization.

Awareness and perception of farmers’ markets were low. Many residents perceived markets as expensive, culturally misaligned, or physically inaccessible, and most were unaware of the Cowtown Farmers Market. Although some residents expressed interest in purchasing directly from farmers, they also highlighted practical barriers such as hours of operation, limited product knowledge, and a lack of inclusive or welcoming market environments.

These findings suggest that for a subsidized 75:25 pricing model to be effective, farmers’ markets or farm stands must be situated closer to underserved neighborhoods, designed with cultural relevance in mind, and paired with active education and outreach strategies.

2. Farmer Needs, Constraints, and Opportunities

Farmer interviews highlighted the diversity of farm types operating in Fort Worth (from sole proprietorships to nonprofits and emerging social enterprises). Across these models, farmers reported several persistent challenges affecting profitability and their ability to implement subsidized pricing. These include that:

1. Profit margins remain thin, particularly for startup farms facing land preparation, infrastructure, certification, and labor costs.
2. Most farmers rely on supplemental income, grants, or value-added products to remain operational.
3. Market participation requires significant time and resources, and prices often need to match grocery store prices to retain customers.
4. Infrastructure gaps (particularly cold storage, wash/pack areas, and processing spaces) limit scalability.
5. Labor shortages and limited local training resources remain major barriers.
6. Climate stress and extreme weather events impact yields, consistency, and profitability.

While many farmers embrace sustainable practices (no-till, organic methods, regenerative agriculture), these practices require time and labor that can be difficult to sustain without adequate support. In this context, most farmers expressed limited capacity to adopt a 75:25 pricing model on their own, especially without external investment or a strong organizational mission centered on food justice.

3. Market Environment and Customer Insights

Customer survey findings from Cowtown Farmers Market reveal strong consumer motivations around quality, freshness, supporting local farmers, and market ambiance, with average importance ratings of 4.7–4.9 out of 5. Customers primarily purchase produce, baked goods, eggs, and dairy, and many attend weekly or twice monthly, indicating stable demand. However, customers who frequent Cowtown differ demographically and socioeconomically from residents in underserved East and Southeast Fort Worth communities. Most Cowtown shoppers are women, middle-aged, and from smaller households, with many paying by card or cash rather than SNAP/WIC. These characteristics signal that existing regional farmers’ markets may not currently reach the populations most in need of subsidized produce.

The assessment of 39 farmers’ markets across Dallas–Fort Worth further highlights variability in pricing, vendor composition, cultural relevance, payment options, and accessibility. While some markets offer SNAP/WIC and Double Up Food Bucks, many operate in higher-income neighborhoods or during hours misaligned with working families. For farmers evaluating market entry, the compiled data serve as a valuable decision-support tool.

4. Feasibility of the 75:25 Revenue Model

Across all data streams, the project found limited evidence that a 75:25 model would be broadly feasible for most farms. Profit margins, production variability, and operational expenses pose challenges to providing subsidized produce while maintaining farm viability. Furthermore, farmers differ widely in their missions, capacity, business structures, and community engagement goals. However, the model appears most feasible for farms operating under a nonprofit structure or a hybrid social-enterprise approach, particularly when aligned with an explicit mission to address food insecurity and when supported by external grants, donations, or volunteer networks. Opal’s Farm represents a strong proof-of-concept scenario:

1. It is embedded within a nonprofit organization (Unity Unlimited).
2. It operates with a clear food-justice mission.
3. It benefits from city land-use agreements, community partnerships, and strong branding.
4. Its tiered pricing model with donations (10%), discounted neighborhood sales (30%), and market/restaurant sales (60%) demonstrates how a mission-driven farm can balance operations with community impact.

Such a model, however, depends heavily on mission alignment, external support, and unique partnerships that are not replicable across all farms.

5. Summary of Key Findings

Overall, the evaluation of the 75:25 for-profit : non-profit sales model revealed that while community demand for fresh, locally grown produce is strong, significant barriers (including limited grocery access, transportation challenges, and low awareness or mistrust of farmers’ markets) continue to shape purchasing behavior in underserved neighborhoods. Farmers in the region operate under a wide range of business structures, but most face thin profit margins, high startup and input costs, and ongoing labor shortages, making subsidized pricing difficult to sustain without external support. Likewise, although existing farmers’ markets attract committed customers, their locations, hours of operation, and inconsistent acceptance of SNAP/WIC benefits limit their reach among lower-income households. As a result, a universal 75:25 pricing model is not feasible for the majority of small or emerging farms. The model appears most viable within nonprofit or mission-driven farm structures that can leverage external funding, volunteers, and community partnerships to offset operational costs. Opal’s Farm provides a strong example of how tiered pricing can succeed when embedded within a food-justice mission and supported by institutional relationships. Looking forward, promising opportunities include the development of culturally tailored farmers’ markets, expanded SNAP/WIC matching, improved business and branding support for growers, and targeted investments in farm infrastructure and farmer training to strengthen the long-term feasibility of socially inclusive pricing models.