Muller further notes,“if GTM is to serve as a way of knowing, then the knowledge that it produces should be placed in relation to other knowledge”, planting the GTM-variant taken on by HCI and CSCW researchers squarely in the Straussian strain. Indeed, in this study, I use the Corbin and Strauss methodology that encourages searching for patterns in the data, constant comparison, and theoretical sensitivity. I also heed the practical advice Charmaz provides in “Constructing Grounded Theory” .In total, there were 16 participating farms , with 19 participants. An overview of each of the participating farms is avaialble in Table 4.1. At three farms, two people were interviewed. Both participants and farms are referred to by pseudonyms, to assist in tracking participant-farm relationships, the first letters of the farm and participant pseudonyms are the same: e.g., Carol and Charles, Campbell Farms. Artifacts have been scrubbed of identifying information to protect privacy, ensure anonymity, and reduce bias. Each of the three farming regions, — the North Coast and Mountain region, the Central Valley, and the Central Coast and Southern region — encompass diverse climates, geographies, cultures, products, and agricultural systems. Coverage of all regions would not have necessarily resulted in coverage of the full diversity of farm types, as farms vary widely within regions, resulting in many further subdivisions. Therefore, while we did not manage to recruit any farms from the North Coast and Mountain region, the farms that did participate still represent a broad range of agriculture in California. The participating farms were from 10 California counties, grow racks as far north as Butte County and as far south as San Diego County, and farming landscapes ranged from coastal farms in Ventura and San Diego Counties, to hilltop farms in San Luis Obispo and Yolo Counties.
Of our participating farmers, 6 were first generation farmers and the other 13 came from families that had been involved in farming for at least two generations. Participant ages ranged from 30 to over 75. Stewardship of land and water and the relationship of the farm with local wildlife was a prominent theme that came up during the interviews. Farmers described the interplay between their farms and the environment, taking delight in their relationship with nature. This delight is certainly not something quantifiable, or adequately expressible through environmental certifications and analysis tools alone. Farmer Carol, of Campbell Farms, said: “We’re stewards of the land. I love the wildlife; I would never want to harm it with the way I’m farming. […] A year ago, there were this little birds called killdeer, they nest on the ground. One of them had a nest, so we went around it. And we let her have her babies you know? […] We’re trying to preserve and promote the wildlife in our farm and a lot of times the animals are attracted to the farm because it’s a safe haven. There’s water and there’s food.” In the 2010 report, “Toward Sustainable Agricultural Systems in the 21st Century”, the National Research Council defines four goals that a sustainable agricultural system should meet: the ability to satisfy human food needs; sustained economic viability; enhanced quality of life for farmers, wkers, and society; and enhanced environmental quality. However, this report also concludes that “no simple typology or set of categories can capture the complexity of the farming practices and systems used on diverse U.S. farm systems” , and that the lack of such a typology complicates our understanding of what it means for a farm to be sustainable. Indeed, agricultural systems are not binary: it’s not that they’re either conventional or sustainable, but there exists a “farming systems continuum”. Based on our findings, we posit that a typology of farms is not necessary, and that instead, the ability to express the spectrum of sustainability may suffice.
Majority of the farms were located at a single site. Both Cambell Farms and Lowry Fields consisted of two sites. Different products were grown across each of the sites, where the day-to-day operations of each site were managed relatively independently, with the main farmer splitting time and coordinating between sites. At Iyer Organic Orchards, the same orchard style was simply replicated at each site. Marsden Organic Farms, while spread across several sites, integrated certain aspects of farm management across the sites, but subdivided field-level operations within sites. While the farm was distributed across space, the farmers and collaboratively farmed the land across all farm sites, in effect treating it as they would have a single site. The outlier on this distribution is Jordan Hives. This beekeeping operation is distributed over the greatest number of locations as the beekeeper, Jack, places groups of hives on other farmers’ properties to pollinate their orchards and fields. The primary location, Jack’s home, contains a honey processing facility, a queen breeding and hive splitting operation , and a small pollinator-friendly garden. Given this dynamic spatial distribution of the hives, Jack uses a mobile application specifically designed for tracking hives and other beekeeping activities. Data contained includes: location, hive numbers, and the date the hives were moved to location. This allows Jack to keep track of his spatially distributed agricultural system. Figure 4.6 shows four spatial representations used in different types of agricultural systems. The hand-drawn Pullman Biodynamic Vineyard map was originally intended to communicate to visitors to the on-site winery what varieties of grape are grown, and which different systems exist to support the environmental sustainability of the property .
As location of the vines doesn’t change frequently, such a model would not need to be updated on a regular basis. However, the farmers reported that this map had been opportunistically used to coordinate activities, such as pruning and harvesting, among the four family members who work on the farm.In contrast, Marsden Organic Farms have detailed field layout diagrams for each property based on satellite images that are available through a customized FileMaker database application. While the rest of the application is heavily used to coordinate sales, customer orders, and harvest-related activities, these maps are entirely unused as the turnover of crop in each field is high.Glass Organic Orchards and Osborne Organic Greenhouses seem to have found a middle ground in representing useful information through spatial representations. Glass Organic Orchards use a box and text layout diagram of the L shaped property to demarcate irrigation lines, power lines and tree varieties. At Osborne Organic Greenhouses, they use a similar box and text layout diagram to mark where pest-related interventions, such as mousetraps, have been placed. A duplicate of this map exists for indicating flytraps. These maps are used both for communication among farm workers, as well as for CCOF reporting. As the crops themselves are on a two-week harvest cycle they do not use this map for crop locations. Structurally static farms do not typically use spatial representations in daily farm activities. For example, Iyer Orchards was comprised of four non-contiguous 40-acre fields within an approximately 4-mile radius and with almost identical layouts at each site. Each was planted a few years after the next as Irfan made staggered land purchases. While his orchards are geographically dispersed, they are each treated as independent systems, vertical grow system with little interaction between the fields across space. Both the farming and modeling practices are simply replicated at each location with minimal customization as each of the fields are very similar in geography and environmental context. Littlewood Fields is arguably similar in spatial complexity. The primary location consists of several rice fields, each approximately 40 acres and tens of acres of pasture. The second location serves only as a winter pasture for Larry’s cows. Here too, no spatial representations are used on daily basis with Larry relying on a visual survey of his fields to determine field status and plan the next set of activities. Both Iyer Orchards and Littlewood Fields were also closer to the conventional side of the farming spectrum. At each farm, approximately the same inputs were applied to the fields each season, and the farm layouts were designed to optimize yield.
This uniformity dimin-ished the need for much spatial representation, with most of their records revolving around input tracking for internal assessments and regulatory reporting.Prior to each new planting, the farm manager traces an outline of the farm layout onto a new sheet of paper as shown in Figure 4.10. He marks the date, and labels each field with what crops are currently there, signing off the field as it is harvested. They have a binder containing farm layouts dating back several years so that they may track what has been grown, and where. This allows them to plan out relevant soil building activities , report data required for regulatory and certification-related purposes, and engage in daily farm management.Spreadsheet tools are commonly used to track changes on farms, including changes to the land, crops, or animals. At Marsden Organic Farms two temporally sensitive items are tracked: crop locations in specific fields, and animal movement . It shows a spreadsheet used to track flowers grown in a specific field during the 2014 season. Each row in the spreadsheet represents an actual row in the field. Each year, the template is duplicated, instantiated with flower types, printed out, and annotated as the season progresses. This table-based representation is analogous to the map-based crop tracking at Brooks Organic Farms : both track the movement of crops across the farm over time.Animal movement at Marsden Organic Farms is similarly tracked . While it may seem like an innocuous activity, this log is instrumental to the farm maintaining its organic certification. While animals are only eating cover crop and crops that were not harvested, there is a requirement for no animals to have been in a field for 120 days before crops for human consumption can be grown organically. This requirement makes it particularly hard for many farms to keep both animals and row crops.A consistent challenge across farms exists in representing variegated spatial complexity in terms of system composition and structure. In all the nut orchards, for example, alternating rows of trees within a single field contained different varieties for pollination purposes. Vertical integration of the farm, i.e., how much of the farm-to-table pipeline is integrated into the farm’s scope of activities, plays a role in increasing the overall complexity of the farm. While some farms only grow crops or raise livestock, there are many others that we found processing, packaging, even trucking produce to consumers. The different layers of activity must also be considered: i.e., activities concerning the soil layer; the water and irrigation layer; the air and emissions; machinery and building infrastructure; the human layer; and not to forget, the crop layer itself. There is a tension between models that farmers use to create snapshots of their system in space , and those used to represent time . This tension has impelled farmers to create models to track changes in farm composition over time . How this inherent complexity of the farm is represented directly affects how it can be managed, modeled, and subsequently analyzed.All farms encapsulate many time cycles. For example, staggered plantings of row crops like tomatoes are common to allow for a longer harvest period, resulting in several cycles associated with each planting overlapping. Yet more dynamism is introduced on farms containing multiple commodity types such as at Marsden Organic Farms. Due to changes occurring in a farm, models are also often out of sync with reality. This dynamic complexity on farms makes the representation of time, activities, and tracking changes a significant challenge. Farms are not only subject to human induced changes, but also at the whims of natural evolution and decay.Farms do not exist in isolation. They have challenging and diverse relationships with their neighbors, society, and the natural environment. There are often interconnections between farms, whether it involves resource sharing, output or waste reuse or creating natural buffers between farms. These informal connections are rarely captured in farm models. The capture of context is challenging as farms have complex flows of resources within and among systems and subsystems. When assessing the environmental performance of farms, a further complication is that farm boundaries are fuzzy; where does the farm end and the natural environment begin?Many policies and certifications require data to be both collected and structured using specific forms, formats, or tools.