Local and regional food systems, soil phosphorus, and resilience in a northeastern regional farmers’ market network

Final report for GNE15-105

Project Type: Graduate Student
Funds awarded in 2015: $14,796.00
Projected End Date: 12/31/2017
Grant Recipient: Penn State
Region: Northeast
State: Pennsylvania
Graduate Student:
Faculty Advisor:
Expand All

Project Information


The purpose of this dissertation project is to understand how farmers participating in local/regional food systems (LRFS) manage soil fertility, the many factors that influence those decisions, and soil fertility resilience in a LRFS in the New York City region including parts of the mid-Atlantic and lower New England. These goals were accomplished through a mixed-methods case study of the New York City Greenmarket network of farmer’s markets that includes farmer interviews, soil analysis, and nutrient supply mapping. This project  provides a better understanding of the many social and ecological factors that influence soil management decisions, and an assessment of the resilience and sustainability of nutrient supply and soil fertility in this LRFS. These findings have been made available to farmers through direct feedback, and will be used for potential future workshops.

Interviews were conducted with all participating farmers (n = 22) during the winter of 2016. Bulk soil was collected from each participating farm field (n = 226) twice during the 2016 growing season for laboratory analysis. Soils collected in early spring were assessed for basic chemical fertility measures and soil organic matter. Soils collected mid-growing season were assessed for active carbon, soil respiration, and acid phosphatase enzyme activity related to the hydrolysis of organic phosphorus for plant and microbial use. Information gathered during interviews was assessed to understand soil fertility management on participating farms, and was used to trace phosphorus supply networks in this LRFS.

This research found, similar to national-level surveys (Low et al. 2015), that participating farmers utilize a wide variety of fertility practices, ranging from synthetic chemicals to manure and cover crops, but a majority also incorporate some form of biomass-based fertilizer. This project grouped participating farmers into three management categories: biomass, synthetic, and hybrid. Biomass farmers use only biomass-based fertilizers (manure, compost, cover crops), synthetic farmers use only processed chemical fertilizers, and hybrid farmers use some combination of the two. Soil analysis found that farms incorporating more biomass-based fertilizers had greater potential to support sustainable phosphorus cycling without the importation of external nutrients. However, regardless of this potential, all but one farm imported phosphorus each year, and 85% of sampled fields had excessive levels of soil test phosphorus. Additionally, all farms relied on phosphorus derived from non-renewable mineral sources.

Interviews indicate that soil fertility (and many other aspects of farm management) are most significantly influenced by formative experiences in farming, and additionally by access to resources (capital, raw materials, etc.). Farmers reported discussing farming practices with customers in the farmers’ markets on a regular basis, but this does not seem to directly influence management decisions in a significant way. There are significant levels of trust and care that evolve through farmers’ market interactions, however, and these relations (and relationships) do seem to influence farmer attitudes toward environmental farm management over relatively long tenures of market participation (10-15+ years). In numerous cases, these changing attitudes lead to shifts in management that are generally considered more sustainable.

When read together, these findings indicate that the Greenmarkets have high potential soil fertility resilience, yet this potential is not currently realized. Efforts must be made to build regional systems of grain and manure production to create sustainable nutrient cycles, and to assist farmers in transitioning to fertility practices that carefully monitor soil and utilize regional biomass for fertility.

Soil fertility analysis and interpretation were delivered to all participating farmers, and several farmers either made shifts in practice, or expressed interest in doing so. Efforts are currently underway to develop training modules through sustainable agriculture organizations including the Pennsylvania Association for Sustainable Agriculture. This work has the potential to assist farmers and food system planners to develop sustainable nutrient systems in the Northeast, and provides a model for future work that combines social and ecological methods to address pressing challenges for sustainable agriculture.

Low, Sarah A., Aaron Adalja, Elizabeth Beaulieu, Nigel Key, Steve Martinez, Alex Melton, Agnes Perez, et al. 2015. Trends in U.S. Local and Regional Food Systems. AP068. United States Department of Agriculture, Economic Research Service.


In recent decades, LRFS have emerged as a prominent strategy for achieving ecological sustainability in food systems. The last thirty years have seen the phenomenal increase in LRFS, which include things like farmers' markets, community supported agriculture, and many others (Low et al. 2015). In some parts of the northeastern United States, farms that participate in this sector are the only category to buck the trend of agricultural abandonment, making them critical to meeting future food needs, and to managing the flow of nutrients through the region (Glynwood 2010). Yet very little is actually known about the farm management practices used on these farms or the extent to which they achieve the promise of sustainability. The most recent USDA report on local and regional food systems indicates that farmers in these food systems are just as likely to use conventional nutrient inputs, and in similar application rates, as farmers that sell through commodity markets (Low et al. 2015). This is important data, but it is derived from large-scale quantitative surveys and fails to capture crucial factors in soil management, namely the factors that lead individual farmers to select certain management strategies over others (Prokopy 2011). Early returns on a farm management survey in the farmers' market system that serves as the case study for this research indicate soil management practices that mirror those found by the USDA. This research build on the findings of the USDA report by investing in a specific case study, interviewing farmers and studying their soil to better understand the complex web of factors that influence soil management and soil outcomes in LRFS.

Glynwood. 2010. The State of Agriculture in the Hudson River Valley. Cold Spring, NY: Glynwood Center. http://www.glynwood.org/publications-multimedia/state-of-ag/.

Low, Sarah A., Aaron Adalja, Elizabeth Beaulieu, Nigel Key, Steve Martinez, Alex Melton, Agnes Perez, et al. 2015. Trends in U.S. Local and Regional Food Systems. AP068. United States Department of Agriculture, Economic Research Service.

Prokopy, Linda S. 2011. “Agricultural Human Dimensions Research: The Role of Qualitative Research Methods.” Journal of Soil and Water Conservation 66 (1): 9A – 12A.

Project Objectives:

The research objectives for this project are as follows:

  1. Assess and describe soil fertility management on farms that participate in the New York City Greenmarket farmers’ markets, which includes:
    1. Documenting all farm management practices each farmer attributes to soil fertility.
    2. Documenting and tracing all input sources and application frequency.
    3. Measuring soil fertility (Mehlich 3 P, K, Ca, Mg, Zn, Cu, S, pH, cation exchange capacity, soil organic matter, active carbon), microbial activity (acid phosphatase and soil respiration).
  2. Explore and document the many factors that influence soil fertility management decisions on these farms, especially the role of participation in the Greenmarket farmers’ markets.
  3. Assess and document soil fertility resilience in the Greenmarket food system.

Objective 1:

This objective provides an empirical assessment of soil fertility management and soil fertility outcomes on vegetable producing farms participating in the Greenmarket case study. This type of analysis is essential for better understanding social-ecological interactions in LRFS and identifying points of intervention for building more sustainable food systems. Data for this objective was obtained through farmer interviews and soil analysis.

Objective 2:

This objective provides insight into the complexity of farmer decision-making relating to soil management, and pays particular attention to how participating in an LRFS does (or does not) influence farm management. This objective is critical to the study as it directly seeks to understand if participating in this rapidly growing market sector is a factor in environmental management on participating farms. Data in support of this objective was collected through farmer interviews.

Objective 3:

This objective synthesizes findings from objectives one and two and uses resilience thinking to provide an assessment of the social-ecological sustainability in the Greenmarket case study.



Click linked name(s) to expand/collapse or show everyone's info
  • Dr. Karl Zimmerer


Materials and methods:

This research utilized qualitative, quantitative, and geovisual methods of data collection and analysis. This approach was designed to provide a holistic picture of soil fertility management in this LRFS, which is critical to sustainability.

Participants were recruited directly from the Greenmarkets during the summer and fall of 2015. Farms were sought to represent both the geographic and demographic variation among Greenmarket farmers as well as the diversity or management practices. The total number of participants varied over the course of recruitment and farmer interviews, with the final number of farms totaling 23 and 28 farmers (four farms are co-managed).

The median cultivated acreage for participants is 18, with the largest two farms growing more than 240 acres of vegetables and the smallest 2 cultivating an acre or less.  Participating farms range in age from two years of operation to an intergenerational farm in operation since 1840 (median 15 years). FM tenure also covers a broad range from farms in their first year of participation to a farm with 37 years in the FM network (median 12 years). Four of the participating farms are certified organic, with two others transitioning to certification. An additional 3 farms participate in the Certified Naturally Grown program, which has restrictions similar to the USDA organic rules. Five farms in this study mainly use what are considered conventional farming practices, and the remaining farms fall along a spectrum of practices described below. Participants cluster in three distinct regions: the lower Delaware River region (­n = 6), the mid-Hudson Valley region (n = 9), and the Catskill Mountains region (n=5) (Figure 1). The variation found among case study participants is representative of the variation found in the total population of vegetable producers in this FM network, and more importantly, they capture variation in variables that past studies have found to be indicators of environmental farm management (Selfa, Jussaume, and Winter 2008).

Figure 1: Map of study region. The Greenmarket draws from farms within the region within the large ellipse. Participating farms are drawn from the inset sub-regions. The northernmost sub-region is the Catskill Mountains, followed by the Mid-Hudson, and the Lower Delaware region to the south

In-depth, semi-structured interviews were used to gather data on how and why participating farmers make management decisions, their knowledge and practices related to soil fertility, and their experiences farming and participating in the Greenmarket farmers' markets. Interviews were transcribed and qualitatively analyzed using an inductive coding method adapted from grounded theory methodology (Charmaz 2000). This approach allows themes to emerge from the data and provided key insights into why participating farmers make soil management decisions, and how their decision making process intersects with their experience selling through the Greenmarkets.

Interviews were also coded using a list of keywords that provided a means for extracting quantitative and demographic data about each farm from the interview transcript. These data points, such as size of farm, years in farming, years in Greenmarket, land tenure, fertilizer types, fertilizer sources, etc. were compiled in a database along with the results of soil analysis to provide social and demographic data for statistical analysis.

All qualitative analysis was completed in the RQDA qualitative data analysis package (Huang 2014), which is an open-source software package that runs through the R statistical software platform.

Soil was collected from each field on all participating farms twice during the 2016 growing season. Twelve to fifteen cores were extracted from each field using a 2 cm diameter soil probe from the upper 20 cm of the soil profile. Individual cores were aggregated to provide a representative sample for each filed. Soil samples collected in the spring of 2016 were air dried and sent to the Penn State Agricultural Analytical Services Lab for basic fertility (Melich 3 extractable P, K, Mg, Ca, S, Cu, Fe, pH, and cation exchange capacity (CEC)) and soil organic matter (SOM) based on weight loss on ignition. This analysis provides baseline fertility for each field.

Soils collected mid-growing season 2016 where divided after aggregation, with one portion air dried for respiration and active carbon assays (Hurisso et al. 2016), and the other frozen to capture field conditions for enzyme assay (DeForest 2009). This analysis is based on soil health assessments, which approach soil fertility and soil management from an ecosystem perspective that incorporates physical and biological aspects of the soil in addition to chemical fertility (Moebius-Clune et al. 2016; Drinkwater and Snapp 2007).

Active carbon was assessed using the permanganate oxidizable carbon (POXC) method following the procedure developed by Weil et al. (2003). Briefly, 2.5 g of air dried soil is shaken for two minutes with 20 ml of 0.02 KMnO4, allowed to settle for ten additional minutes, and then a 0.5 ml aliquot of solution is diluted to 50 ml with dionized water and absorbance is measured at 550 nm on a hand-held spectrophotometer and plotted standard curve. This method has been found to be an excellent proxy for carbon storage and availability of food sources for soil microbe communities which are essential to nutrient cycling and soil stability (Hurisso et al. 2016; Weil et al. 2003).

Respiration was assessed by the generation of CO2 following the rewetting of dried soil using a method adapted from Franzlubbers et al. (2000). Briefly, 10 g of dried soil is placed in a 50 ml plastic beaker, wetted to 50% water-filled pore space, and incubated in a sealed 1 pint canning jar for 24 hrs at 25° C ± 1° C. At the end of the incubation period, a 1 ml air sample is drawn via syringe from the head space in the jar and assessed on a LiCOR LI-7000 infrared gas analyzer. Total CO2 generated is assessed as the difference between sample and an empty chamber (ambient CO2) and plotted on a standard curve. This method has proven to be an excellent proxy for microbial biomass and vigor, and short-term nutrient cycling (Franzlubbers et al. 2000; Hurisso et al. 2016).

To assess the potential for in situ provision of P to crops through the hydrolysis of organic phosphate compounds, each field was assessed for acid phosphatase enzymes, which hydrolyze monoester phosphates to release inorganic orthophosphate which is available to plants and microbes. Phosphate monoesters are generally the largest fraction of the labile organic P pool in soils, and have been found to be significantly important to plant growth (Turner et al. 2003; Dodor and Tabatabai 2003). While some studies assess soils for both acid and alkaline phosphatases, which have optimal performance in acid and alkaline pH ranges, respectively, nearly all fields in this study had a soil pH range within the optimal range of acid phosphatase (pH 6.5), thus only acid phosphatase was assessed and used as a proxy for in situ P cycling.

Acid phosphatase was assessed using the standard method method of Tabatabai (1982) as adapted by Kremer (1994) for high through-put analysis on a 96-well microplate spectrophotometer. Briefly, samples of frozen, field-moist soil was thawed at 4° C over night and 0.5 g samples where measured into 25 mL screw cap centrifuge tubes and brought to room temperature of approximately 25° C. Samples were then incubated at 37° C for one hour in a solution of toluene and modified universal buffer (pH 6.5) with a p-Nitrophenyl phosphate substrate in the same buffer solution. After incubation, the reaction was ended by the addition of 0.5 M NaOH and 0.5 CaCl2, and the sample was filtered though folded no. 2 quantitative filter paper. 250 μL samples were then added to the wells of a 96-well culture plate and absorbance was measured at 405 nm  on a microplate reader. Acid phosphatase activity was measured as as μg of p-nitrophenol released per hour.

Together, this soil health approach provides a comprehensive assessment of how farm management is able to provide basic agronomic fertility and build soil conditions that are capable of reducing reliance on external inputs and provide P fertility through in situ nutrient cycling. Comparisons between management practices were assessed using analysis of variance and mixed-effects modeling using the R and Minitab 18 statistical packages.

Sources phosphorus fertilizers were obtained from farmer interviews and traced back through the supply chain to the furthest extent possible for each source. This was done through iterative rounds of contacting suppliers of feed and fertilizer used by participating farmers and tracing their phosphorus sourcing back through the supply chain to the original source. In the case of mineral sources, this approach provided direct links back to mined phosphate rock sources. In the case of organic sources of P, this approach linked either to a marine source, or a conventional feed grain source, after which supply was assumed to link to mineral sources based on national level statistics (MacDonald et al. 2009). Phosphorus supply networks were compiled in a geographic information system (GIS) database and mapped using ESRI ArcMap 10.5 to visually assess nutrient supply and compare management types.

This mixed methods approach provides a means of holistically assessing resilience and sustainability in this LRFS case study. Reading these different data sets against each other produces a more nuanced conceptualization of resilience and sustainability by associating the results of quantitative soil assessments with not only types of practices, but farm experiences that inform their decision-making process. This methodology provided a detailed picture of the factors that enable and constrain sustainability in this LRFS. Furthermore, the tracking and mapping of phosphorus supply provides an explicitly spatial component to the project that places these findings squarely within debates about food system localization as a sustainability strategy.


Charmaz, K. 2000. “Grounded Theory: Objectivist and Constructivist Methods.” In The Handbook of Qualitative Research, edited by N. K. Denzin and Y. S. Lincoln, 2nd ed., 509–35. Thousand Oaks, CA: Sage.

DeForest, J. L. 2009. The influence of time, storage temperature, and substrate age on potential soil enzyme activity in acidic forest soils using MUB-linked substrates and L-DOPA. Soil Biology and Biochemistry, 41(6), 1180-1186.
Dodor, D. E., & Tabatabai, M. A. 2003. Effect of cropping systems on phosphatases in soils. Journal of Plant Nutrition and Soil Science, 166(1), 7-13.

Drinkwater, Laurie E., and S. S. Snapp. 2007. “Nutrients in Agroecosystems: Rethinking the Management Paradigm.” Advances in Agronomy, 92: 163–86.

Franzluebbers, A. J., Haney, R. L., Honeycutt, C. W., Schomberg, H. H., & Hons, F. M. 2000. Flush of carbon dioxide following rewetting of dried soil relates to active organic pools. Soil Science Society of America Journal, 64(2), 613-623.
Huang, R. 2014. RQDA: R-based qualitative data analysis. R package version 0.2–7.
Hurisso, T. T., Culman, S. W., Horwath, W. R., Wade, J., Cass, D., Beniston, J. W., ... & Lucas, S. T. 2016. Comparison of permanganate-oxidizable carbon and mineralizable carbon for assessment of organic matter stabilization and mineralization. Soil Science Society of America Journal, 80(5), 1352-1364.
Kremer, R. J. 1994. Determination of soil phosphatase activity using a microplate method 1. Communications in Soil Science & Plant Analysis, 25(3-4), 319-325.

MacDonald, James, Marc Ribaudo, Michael Livingston, Jayson Beckman, and Wen Huang. 2009. Manure Use for Fertilizer and for Energy. Report to Congress no. AP-037. USDA Economic Research Service. https://www.ers.usda.gov/publications/pub-details/?pubid=42740.

Moebius-Clune, B.N., D.J. Moebius-Clune, B.K. Gugino, O.J. Idowu, R.R. Schindelback, H.M. Ristow, H.M. van Es, et al. 2016. Comprehensive Assessment of Soil Health: The Cornell

Selfa, Theresa, Raymond A. Jussaume, and Michael Winter. 2008. “Envisioning Agricultural Sustainability from Field to Plate: Comparing Producer and Consumer Attitudes and Practices toward ‘environmentally Friendly’food and Farming in Washington State, USA.” Journal of Rural Studies 24 (3): 262–76.

Tabatabai, M.A. 1982. Soil enzymes. pp. 903-947 IN: A. L. Page et al.
(Eds.) Methods of Soil Analysis, Part 2: Chemical and Microbiological Properties. Agronomy Monograph No. 9 (2nd edition). American Society of Agronomy, Madison, WI.

Turner, B. L., Cade-Menun, B. J., & Westermann, D. T. 2003. Organic phosphorus composition and potential bioavailability in semi-arid arable soils of the western United States. Soil Science Society of America Journal, 67(4), 1168-1179.
Weil, R. R., Islam, K. R., Stine, M. A., Gruver, J. B., & Samson-Liebig, S. E. 2003. Estimating active carbon for soil quality assessment: a simplified method for laboratory and field use. American Journal of Alternative Agriculture, 18(1), 3-17.
Research results and discussion:

Objectives 1:

a) Interviews indicated that there is a significant degree of variation in soil fertility management among participants, but participating farms could be grouped into three general soil fertility paradigms, as defined below:

  • Biomass Fertility (n = 8) – The application of organic matter to soil to meet plant nutrient needs. Inputs compost (farm-made and purchased), manure (raw, composted, and pelletized), soy meal, fish and kelp emulsions, humates, and cover crops grown as fertility amendments (green manure).
  • Synthetic Fertility  (n = 3) – The application of chemical compounds derived from industrial processes to produce highly soluble and concentrated nutrient fertilizers.
  • Hybrid Fertility  (n = 8) – The use of soil amendments that correspond to both general categories to meet plant nutrient needs.

Initial classification included a fourth management group, rock mineral fertility, which included farmers that used crushed rock mineral inputs that are used by many certified organic and organic-minded farmers. While these types of inputs are commonly used by farmers in this group, most of these farmers also use cover crops and compost (though not necessarily every year), thus there were very few farms and farm fields that fell squarely in this group. Given these circumstances, the rock fertility group was combined with the hybrid fertility group.

There are also three participating farms on the Black Dirt soils of the Walkill River Valley of Upstate New York that do not fall into any management group. These farms, as is typical of vegetable producers in the Black Dirt region, do not actively manage soil fertility, and rely instead on the in situ fertility that is charactaristic of drained histosols. These farms were not considered a separate management paradigm since their practices are a result of unique soil resources and therefore not applicable to farms outside the region. Rather these farms are used a type of idea case for comparison to other management types.

The presence of a significant number of participants that use hybrid fertility management strategies suggests that these categories might be better thought of as a spectrum of approaches. Most of the hybrid fertility participants use a combination of rock minerals and biomass-based amendments, which is in keeping with certified organic practices (four participant are certified organic and three are certified naturally grown, which has similar restrictions). However, two participants in the hybrid category also use synthetic fertility amendments, which makes it difficult to provide a more precise definition to the category.

b) Nutrient network mapping, seen in the maps below, produce complex webs of phosphorus supply. While there is some variation in local suppliers among the three sub regions, there is also significant overlap. Figures 2 and 3 depicts this first link in the phosphorus network that flows from local feed and input distributors to each sub-region. 

Figure 2: Primary phosphorus network with links from sub-regions to regional grain and input distributors.
Figure 3: Expanded view of primary network showing link to producer in Florida.










As these maps indicate, with the exception of one supplier in norther Florida, the majority of farmers source their phosphorus nutrients from distributors in the northeastern U.S. The majority of the overlap in networks occurs in south-central Pennsylvania, where several important distributors of organic-approved soil amendments are based.

The primary distributors of grain and soil amendments depicted in figures 2 and 3 are, with few exceptions, not the locations of origin for the the phosphorus nutrients used by participants. Each sub-region has farms that uses municipal leaves as a key component of compost amendments, and these inputs supply phosphorus extracted by trees from local soils. Additionally, one farm in the Lower Delaware sub-region uses composted manure from their small herd of dairy cattle that are entirely grass-fed. Beyond these exceptions, all other sources of phosphorus used by participating farmers eventually link back to mined mineral phosphate extracted in Florida, North Carolina, and Morocco, as depicted in figure 4. In the case of synthetic and rock mineral suppliers, this connection was directly documented through network tracing. Biomass-based inputs were only traceable to either the animals that produced manure, or the feed that was fed to such animals. Since tracing beyond that point was not possible, subsequent links are implied. However, all such sources found in this study document links to conventionally grown grain, which national statistics indicate are predominantly grown with synthetic phosphorus fertilizers. Thus the phosphorus contained in these biomass-based inputs originated from the same mined mineral sources (Macdonald et al. 2009).

Figure 4: Complete phosphorus network, including implied links from biomass-based inputs to mineral sources.

This line of analysis provides two important findings for the sustainability of local food system initiatives such as the Greenmarkets. First, while the farms are located within a clearly defined region, the network of nutrient supply, at least for phosphorus, is certainly not a local network. In terms of conventional fertilizers this is not surprising as there are no mineral phosphate deposits in the northeast. However, this analysis also indicates that even fertility based on locally produced biomass is largely reliant on the same sources of phosphorus, albeit through longer supply chains. Furthermore, all farms, with the exception of the three growing on Black Dirt and one farm in the Catskills, import phosphorus from these sources on an annual basis, and their fertility management does not reflect in situ soil phosphorus resources. These findings indicate that building truly sustainable and agroecological soil fertility requires not only switching from synthetic to boimass-based fertility, but changing the paradigms under which these fertilizers are produced and applied.

c. The results of soil analysis indicate that there are significant differences between the three management groups, and the Black Dirt group differs from all three management groups (Figure 5).

Figure 5: Box plots of management groups and Black Dirt group for primary assays.

For the SOM, POX C, CO2 Burst, and Ptase assays, higher numbers are better as they indicate soil building, microbial community size and vigor, and in situ nutrient cycling. For each assay, the management groups broke down as Biomass>Hybrid>Synthetic, with the differences between each group statistically significant at the α<0.05 level.

The high level of organic matter in the Black Dirt soils contained significant tannins that dyed the Ptase extractant even after filtering, and interfered with colorimetric analysis. Therefore the Black Dirt soils are not included in the phosphatase enzyme comparison. In all other assays, however, the Black Dirt soils were significantly higher than all three management groups. The differentiation between Black Dirt and other management groups was most pronounced for SOM and POX C, with the difference in CO2 Burst statistically significant, but less pronounced. There is also more significant variation in the CO2 Burst assay, with numerous fields in both the biomass and hybrid groups scoring at least as high as the Black Dirt fields. Hurisso et al. 2016 argue that this indicates the dominance of C mineralization rather than C storage in non-Black Dirt fields, which is likely due in par to much lower levels of organic matter.

Mixed effects modeling was also conducted to account for clustering issues related to the significant variation among farms. While this approach did reveal significant clustering among farms, management type remained a significant predictor at the α<0.05 level for each assay. These models explained at least 85% of the variation for each assay. These models also revealed that Ptase more significantly correlated with CO2 Burst than POX C, which supports the relationship between CO2 Burst and short-term nutrient mineralization put forth by Hurisso et al. 2016.

These models also indicated close relationships between SOM, organic C, and enzyme activity that was difficult to tease apart with this analysis. Future research utilizing structural equation models may better reveal the nuances of these relationships.

Mehilch 3 extractable phosphorus was also compared across groups with the following relationships: Black Dirt≥Synthetic>Hybrid≥Biomass. In this case, Black dirt M3P levels where slightly higher on average than Synthetic fields, but the difference was not statistically significant. The same relationship holds between Hybrid and Biomass fields. Both Black Dirt and Synthetic fields are significantly higher than Hybrid and Biomass fields at the α<0.05 level. Unlike the other assays discussed here, higher levels for Mehlich 3 P indicate excessive levels of soil phosphorus that offer no agronomic benefit beyond a certain threshold, and threaten connected ecosystems with nutrient loading and eutrophication. While Hybrid and Biomass fields did have lower M3P levels, all groups had mean extractable P levels of at least 134 ppm, with 85% of all fields containing excessive levels of M3P. For reference, in most northeastern soils, M3P levels above 50 ppm are considered excessive. So while management practices that rely more significantly on biomass do have lower soil test P levels, none of the management groups (as they are employed by participants in this study) seem able to manage soil phosphorus bellow levels that threaten other critical ecosystem services.

While very few of the fields in the non-Black Dirt farms have assay scores that attain the levels of this ideal case, these results clearly indicate that farm management that relies more heavily on biomass-based inputs have greater potential for in situ nutrient supply that can reduce reliance on external inputs and nutrient loading in ground water and connected ecosystems. However, this analysis also suggests that this potential is not being fully realized by participating farms.

Objective 2:

Interview analysis indicates that the primary influence on soil fertility management strategy is a participant’s formative experiences as farmers. Nearly all participating farmers manage soil fertility (and many other environmental aspects of farming) in a manner that is consistent with the way they learned to farm – whether that be through apprenticeship, experiences as in an intergenerational family farm, or through reading and self-education. Geographic location is also an important factor that not only controls soil type (and thus native fertility), but also access to resources – especially raw materials for biomass amendments. Geographic location is particularly important when considered in combination with the total cultivated acres. Participating farms cultivating more tan 100 acres of vegetables are more likely to purchase fertility - either synthetic or biomass - due to cost considerations and the large volume of biomass fertilizers needed for that size operation.

Interview analysis reveals a complex relationship between farmer’s market participation and soil fertility management. Direct interactions with consumers in the context of the market do not seem to have a direct influence on soil management, however this is likely due in part to the fact that very few consumers ask questions about soil management. There is frequent producer-consumer dialogue about other environmental aspects of farm management, such as the use of pesticides, and interview analysis suggests that over time, farmers do begin to adopt different approaches to pest and disease management that are more environmentally friendly. In terms of soil, the farms that use synthetic fertility are beginning to change, or rethink, their approaches to soil fertility, and this does seem to be indirectly related to participating in the Greenmarket farmer’s markets. However, it is also very common for producer-consumer interactions to reinforce current practice rather than induce changes. This is because farmers are often able to effectively explain their practices and assuage the concerns of potential customers. And the establishment of long-term relationships with repeat customers only seems to bolster this effect. In other words, there is currently no economic pressure in the Greenmarkets for farmers to change practices. And while the relations of care and trust that are central to many farmers' markerts (Hinrichs 2000) are important, farmers are able to establish and maintain those relations without changing practices.

It is is interesting to note, though, that over relatively long periods of participation, some participating farmers did change management practices in ways that might be considered more environmentally benign and sustainable. One farmer adopted integrated pest management (IPM) rather than using chemicals, another farmer shifted soil management from synthetic to compost based fertility, and two others are considering altering soil management to practices that build organic matter. In each of these cases, the farmer has at least 18 years of Greenmarket tenure, and changes did not manifest until at least ten years of participation had passed. Interviews present these cases as a result of personal transformation on the part of the participating farmer, thus drawing a direct connection to the farmers' market participation is tenuous. However, the language used by farmers to describe these changes frequently express care and concern for customers, and for their land. This suggests that the relations of care and trust that are important to the functioning of farmers' markets may lead to changes in farming practices over time, though the exact out comes of these changes are heterogeneous.

Objective 3:

Resilience is defined by ecologists as the ability of a system to maintain certain ecological processes or structures in the face of specific disturbances. While this framework has been widely applied, it has not been widely adopted by those working on food and agricultural systems. In this case study, we develop the concept of agrobiogeochemical resilience, which we define as the ability of a farm system to maintain their fertility needs through in situ nutrient cycling, while reducing (or eliminating) dependence on non-renewable nutrient sources and the transfer of nutrients to connected ecosystems. This study considers agrobiogeochemical resilience in regard to phosphorus, and focuses particular attention on the relationship between participation in local food systems and resilience.

As the results from objective 1 indicate, there is greater potential for in situ phosphorus provision with biomass-based fertility, yet farms utilizing these management strategies continue to import phosphorus to their fields on an annual basis, even though biological and chemical soil test results indicate that additional phosphorus application is not necessary in nearly every case.

This suggests two issues for soil management. First, the differences among these management groups is largely one of input substitution rather than a reliance on artificial versus ecological processes for fertility. Secondly, most farms using biomass-based fertility in this study are using annual compost amendments to supply all fertility rather than using annual cover crop mixtures to supply nitrogen and maintain other nutrient levels.

The compost-only approach has two consequences that can make it difficult to reduce excessive phosphorus levels. Firstly, because the nitrogen:phosphorus ratio in compost and manure is generally 4:1, management that seeks to meet nitrogen demand with compost generally over applies phosphorus by a factor of four (crop demand is typically 16:1). Secondly, since compost is a broad spectrum fertilizer, it is impossible to calibrate fertility inputs to accommodate in situ resources. This may explain why many participating farmers using biomass based fertility continue to import phosphorus when soil resources are sufficient to meet plant demand.

These results suggest that building agrobiogeochemical resilience in this local food system should focus on building capacities and infrastructure for the use of ecological processes for soil fertility that can assist farmers in moving away from input substitution.

The results of nutrient network mapping also indicate that all farms in this study, regardless of management type, are significantly dependent on the same non-renewable phosphorus resources. Biomass-based fertility recycles nutrients, and thus these links back to mined phosphate are longer, yet this type of recycling does not address the underlying problem of soil fertility paradigms that rely on non-renewable resources. These results indicate that building agrobiogeochemical resilience in this (and likely other) local food systems requires regional restructuring to create regional systems of grain and manure that can close fertility cycles and create regenerative recycling. This is a daunting structural challenge, but this analysis clearly indicates that this type of paradigm shift is necessary to reduce reliance on non-renewable fertility resources.

Finally, the results of Mehlich 3 extractable phosphorus indicate that much work remains to be done to reduce the potential for nutrient loading from agriculture on these participating farms. The potential fixes outlined above will address this challenge, but this effort will likely also require investment in regular soil testing on regional farms. Most of the farms participating in this study did not regularly test soil, particularly farms cultivating less than 100 acres. The primary barrier is time for collection, though financial considerations are also important in many cases. It can take many years of concerted effort to draw down excessive phosphorus levels, and a long term commitment to monitoring is essential.

Based on this analysis, the agrobiogeochemical resilience of this local food system initiative is relatively low, yet this study does offer a silver lining. Many of the farms participating in this study use at least some biomass for fertility, and results suggest that this approach has great potential (with modifications) to increase agrobiogeochemical resilience. While there is certainly some selection bias in the relatively high number of farmers using biomass-based fertility (both in terms of study participation and farmers' market participation), this study suggests that long-term farmers' market participation can nudge farmers toward soil fertility practices that build soil organic matter. This suggests that with directed effort, local food systems like the Greenmarket farmers' markets can build agrobiogeochemical resilience.

Achieving these goals will require coordination across sectors, but these efforts are attainable in the long term, and study participants indicated that there is interest in meeting these goals. Food system localization is not a panacea, but it does show some potential for building resilient and sustainable food systems.

Hinrichs, C.Clare. 2000. “Embeddedness and Local Food Systems: Notes on Two Types of Direct Agricultural Market.” Journal of Rural Studies 16 (3): 295–303.

Hurisso, T. T., Culman, S. W., Horwath, W. R., Wade, J., Cass, D., Beniston, J. W., ... & Lucas, S. T. 2016. Comparison of permanganate-oxidizable carbon and mineralizable carbon for assessment of organic matter stabilization and mineralization. Soil Science Society of America Journal, 80(5), 1352-1364.

MacDonald, James, Marc Ribaudo, Michael Livingston, Jayson Beckman, and Wen Huang. 2009. “Manure Use for Fertilizer and for Energy.” Report to Congress AP-037. USDA Economic Research Service. https://www.ers.usda.gov/publications/pub-details/?pubid=42740.

Research conclusions:

As indicated in the discussion of the results in the previous section, phosphorus use in this study is not currently sustainable, but there are clear (though challenging) pathways toward agrobiogeochemical resilience.

The primary barriers to phosphorus resilience are management paradigms that focus on input substitution, and fertility infrastructure that relies on either the direct use of mined phosphates, or the recycling of mined phosphates from conventional animal feeding operations. While these are significant challenges, there is some indication that local food systems such as the farmers' markets in this case study can facilitate a transition to more resilience soil fertility management.

Firstly, the farmers' markets in this case study support a significant number of farms that recycle biomass for at least some of their fertility needs, and there is also evidence to suggest that participation in these farmers' markets can, over longer time scales, lead to changes in fertility management toward the use of biomass. The majority of the raw biomass inputs used in this study are conventionally produced manure, so the recycling is merely an additional link in a phosphorus supply that is non-renewable. However, the significant use of recycled biomass suggests that if a new infrastructure could be created that facilitated truly local cycles of biomass production, a local food system initiative such as the Greenmarket farmers' markets could achieve a high degree of agrobiogeochemical resilience. These are long-term goals that have not yet been put into practice.

The issue of soil testing and excessive phosphorus levels is an on-going problem in most food producing landscapes in the U.S., and the work of this study has begun to impact practices on participating farms. Soil analysis and consultation have been delivered to all participating farms, and numerous participants have altered practices as a result. As mentioned above, drawing down phosphorus levels is a long-term processes, but this study has aided that work on participating farms.

Participation Summary
28 Farmers participating in research

Education & Outreach Activities and Participation Summary

23 Consultations
7 Webinars / talks / presentations

Participation Summary:

28 Farmers participated
Education/outreach description:

Soil analysis for both chemical and biological assessments was provided to each participating farm, along with comments and recommendations for soil management. Further consultation was offered to each farm, with many participants continuing to discuss soil fertility issues.

The following presentations and talks were given based on this research:

2017    Panel on Agroecology and Environmental Futures, contributing panelist. Nature-Society Geography Workshop, Clark University Graduate School of Geography, September 15-16, 2017, Worcester, MA.

2017    “A soil health approach to phosphorus cycling and agrobiogeochemical resilience in a Mid-Atlantic local food system initiative.” Paper presented at the Ecological Society of America Annual Meeting, August 6-11, 2017, Portland, OR.

2017    “Towards a Framework for Agrobiogeochemical Resilience: Phosphorus, Soil Health, and Sustainability in a Local/Regional Food System Initiative.” Paper preseted at the Annual Meeting of the American Association of Geographers, April 5-9, 2017, Boston, MA.

2017    “What’s the Market got to do with it? Social-Ecological Embeddedness and Environmental Decision-Making in a Northeastern Farmers’ Market Network.” Guest lecture for Geography 108: Introduction to Human Geography, West Virginia University, March 30, 2017.

2016    “Mixing Methods in Political Ecology: Combining Qualitative Data with Biophysical Science.” Paper presented at the Annual Meeting of the American Association of Geographers, March 29-April 2, San Francisco, CA.

2015    “Graduate Showcase: perspectives from the recipients of the Center for Landscape Dynamics 2015 Grad Award.” Panel presentation for Penn State Earth and Environmental Systems Institute EarthTalks Seminar Series, September 28, 2015.

2015    “Food and Phosphorus in the Northeast.” Panel presentation on the Landscapes and Livelihoods of the Northeast, Nature-Society Workshop, October 2-3, 2015, Cornell University, Ithaca, NY.

Additionally, conference panel sessions were organized for the 2017 and 2018 Annual Meeting of the American Association of Geographers on the themes of resilience in agri-food systems (2017), and farmer decision-making (2018).

Three journal manuscripts are currently in progress:

Hedberg, Russell C., and Karl S. Zimmerer. What’s the market got to do with it? Social-ecological embeddedness and environmental practices in a local food system initiative. Renewable Agriculture and Food Systems

Hedberg, Russell C. Coming out of the foodshed: Phosphorus cycles and the many scales of local food. Annals of the American Association of Geogrpahers.

Hedberg, Russell C., Erica Smithwick, and Charles White. Agrobiogeochemical resilience: Phosphorus, soil health, and sustainability in a local food system initiative. Agriculture, Ecosystems & Environment.

Additionally, discussions have begun with staff at the Pennsylvania Association for Sustainable Agriculture to develop training modules for soil health and resilience, and efforts are currently underway to work with the staff of GrowNYC (administrators of the Greenmarkets) to increase the impact of this research and build on its findings.

Project Outcomes

10 Farmers reporting change in knowledge, attitudes, skills and/or awareness
4 Farmers changed or adopted a practice
1 Grant applied for that built upon this project
1 Grant received that built upon this project
$8,187.00 Dollar amount of grant received that built upon this project
25 New working collaborations
Project outcomes:

This project delivered soil fertility analysis to all participating farmers, and introduced participants to soil health assessments that most had not been able to access prior to the study. Based on participant responses, this information has influenced numerous farmers, both in terms of awareness to soil sustainability challenges, and in several cases, led to farmers making adjustments to soil fertility management to improve fertility and phosphorus management.

This project has the potential to contribute to future sustainability providing foundational research to support the building of regional systems of fertility based on feed grain and manure production. Furthermore, this research has provided a means for me to connect with regional food system organizations to develop training programs, consumer education, and to build momentum toward developing the regional systems of fertility outlined above.

Knowledge Gained:

During the course of this project my knowledge and skills in soil analysis increased significantly. The research for this project made clear to me that while individual farm management is an important component of building more sustainable food systems at the local and regional level, the primary challenges to sustainability are structural problems that limit the ability of farmers to truly break from the paradigms of industrial agriculture. Furthermore, awareness of these problems, particularly as it relates to soil fertility, seem low - even among farmers, which was somewhat surprising.

I have secured a university faculty position in sustainability, and will focus on undergraduate teaching, and research in the Northeastern U.S. My future work will continue to examine soil health approaches to phosphorus management, and building sustainable nutrient supply networks in the the region. I will also develop research focusing on consumer and farmer education, focusing on farmers' markets and food hubs as loci of education and knowledge dissemination.

Any opinions, findings, conclusions, or recommendations expressed in this publication are those of the author(s) and do not necessarily reflect the view of the U.S. Department of Agriculture or SARE.