Where grow food? A suitability analysis of where to establish a new bio-intensive farm plot in Southeast Virginia

 Section I: Introduction

            As nations around the world become more urbanized, and as space suitable for agriculture amounts to a premium, people living in urban areas will need to seek alternative methods to acquire fresh produce.  A lack of ready access to fresh produce has both environmental and human health consequences.  A study by Pearce, et.al. (2007) revealed a strong indication of “food deserts” in economically deprived urban neighborhoods in New Zealand.  Food deserts are areas where people have a long travel distance to full-service grocery stores compared to fast-food restaurants and convenient stores.  Their study revealed total distances to full-service grocery stores were longer than distances to fast-food restaurant within economically poor urban areas in New Zealand. 

There exists a general lack of understanding of where food comes from, as well.  A reliance on large grocery stores with expensive food coming from great distances has revealed a populous not fully understanding the agricultural origin for their produce.  Leopold (1970) expresses the naïve view regarding agriculture:

            “There are two spiritual dangers in not owning a farm.  Once is the danger of supposing that breakfast comes from the grocery, and the other that heat comes from the furnace.” 

Clearly the source of food for meals originates from farms and that the heat from the furnace must come from a combustible source such as natural gas.  As urbanized populations become disconnected with the process of agricultural production and natural resource extraction, the misunderstanding of food sources and origins becomes pervasive.

            In response to the dominance of corporate agriculture and the daunting effects it can have on the environment, cities across the United States have been developing community gardens to provide food to lower-income, socioeconomically disadvantaged peoples.  Community gardens have been cited for their various benefits, including the elimination of urban food deserts, addressing nutrition problems, reducing the cost of fresh food for families, and engaging people in their neighborhoods to become involved with community development and growth (Grow Pittsburgh, 2013). 

Grow Pittsburgh, an urban agriculture non-government organization in the city of Pittsburgh, Pennsylvania, has utilized abandoned lots in the post-industrial landscape of this western Pennsylvania community to grow crops using biointensive methods of farming.  Biointensive farming is a method of agriculture with a focus on producing a high yield of crops via organic agriculture methods using only a small amount of space, usually an 8’X 3’ garden plot (Micro Farms, 2013).  The modest amount of cultivated land combined with modern organic farming methods allows for a nearly continuous output of crops on a yearly basis without compromising space.  It also utilizes a chemical-free crop management practice intended to reduce impacts to natural resources. 

This study examines potential sites to establish a new biointensive farm location in two densely populated cities in southeast Virginia: Norfolk and Virginia Beach (see Map 1).  The Micro Farms, a subsidiary of the non-for-profit Ecocycling, an organization dedicated to sustainable community development in this region, has been constructing biointensive farm plots in various locations.  At the time this paper was written, the Microfarms have 5 biointensive farm plots in the city of Norfolk, as well as a conglomerate of plots on a farm in Virginia Beach (see Map 2). 

The cities of Norfolk and Virginia Beach are characterized as a mostly urban landscape, with socioeconomically disadvantaged people across many areas (see Map 3).  Norfolk in particular has a large concentration of minority people, as indicated by Maps 3-5.  This suitability analysis for a new Micro Farms biointensive farm plot will take into account racial and socioeconomic factors using a Geographic Information System (GIS).  The criteria include:

·         A location within an impoverished neighborhood.  The National Poverty Center (2013) has determined that the poverty rate for a family of four in the United States is $22,113. 

·         A location where the population is greater than 50% racial minorities. ‘Racial minorities’ in this context will be defined as any population that is non-white. 

·         Within 2 miles of a public school. 

·         Greater than 2 miles from an already established farm or garden. 

Part of the goal of community gardens and biointensive farm plots is to provide ready access to fresh food to those who suffer “food insecurity” as a result of a low-income status (Grow Pittsburgh, 2013).  Therefore it would be ideal to establish a farm location within an economically disadvantaged neighborhood.  An area dominated by minority populations is also ideal, as environmental justice advocates have cited a lack of quality environmental health resources in regions dominated by minorities (Warren, 2003).  A location near public schools could help make the farm plot visible to young people, and could perhaps help establish community norms about growing food in urban areas.  Guidelines regarding the establishment of community norms in environmental sustainability projects and its rate of success is outlined by McKenzie-Mohr et al (1999), and is the inspiration for this particular criteria.  Finally, to further narrow down a new plot location, a site that is a reasonable distance (2 or more miles) from one of the currently established plots will be included in the analysis. 

Section II:  Methodology

Census tract data for Norfolk and Virginia Beach were acquired from the U.S. Census Bureau’s Topologically Integrated Geographic Encoding and Referencing (TIGER) website, which has census tract data for the 2010 census in a GIS-compatible format.  These GIS data layers include population information by race.  A percentage of minority (non-white) people was calculated for each census tract in Norfolk and Virginia Beach based upon 2010 population data.   This newly calculated data field was then imported back into the GIS.  The GIS was used to create chloropleth map visualizing the percent of minority populations per census tract for both cities (Map 3).  An additional chloropleth map (Map 4) was produced to show the ratio of minorities to the amount of white people per census tract,

N_m / N_w

where N_m is the number of minorities per census tract and N_w is the number of white people per census tract. 

To determine where a concentration of minority people live in the study area, the minority population totals were submitted to a geostatistical hotspot analysis (Getis-Ord G_i*) using ESRI’s spatial statistics package, an extension of their ArcGIS software platform.  This procedure shows the concentration of a phenomena based upon the spatial relationships of a related feature.  In this case, a chloropleth map was produced to show the concentration of minorities in the study area (Map 5). 

Upon visual inspection of the hotspot analysis displayed in Map 5, it was then determined that only Norfolk will be further examined for suitability, as the greatest concentration of minorities occurs primarily within census tracts in the city of Norfolk.  Areas of Norfolk in red shading in the western region of Map 5 were isolated into its own data layer.  Census tracts with a percentage greater than 50% minorities within Norfolk were isolated into its own data layer, as well.  Average household income information, acquired from the TIGER website, was then overlaid with the Norfolk census tracts.  Census tracts whose average household income was at or below the poverty level of $22,113 was isolated into its own data layer and is displayed in Map 6.  A list of Norfolk public school addresses was acquired from the Norfolk Public Schools website (2013) and then geocoded using an address matching service built into ESRI’s ArcGIS software platform.  To determine the 2-mile distance from these schools, a service area analysis based upon the local street network was performed via the network analyst extension in ArcGIS.  The result was an output polygon layer showing two mile distances from each school based upon accessible roads.  The school distances layer, the poverty layer (Map 6), the proportional minority layer, and the minority population concentration layer were then merged into one output in the GIS. 

A layer showing the locations and addresses of the farm and garden locations was loaded into the GIS, and a service area analysis of 2 miles was utilized.  Since areas within 2 miles needed to be excluded from the results, the ‘erase’ function in the GIS was used to omit areas within 2 miles from an existing farm and garden.  Finally, the results were compiled together using overlay analysis in the GIS, producing an output showing areas (as a shaded polygon) that meet all of the above criteria (Map 7).  It should be noted that there was originally a polygon near the northwest corner of Map 7, but it was manually removed since this area is a militarized zone.  

A summary of the entire methodology workflow is presented in Figure 1.   

Section III:  Results

As shown in Map 7, suitable areas for an additional biointensive farm plot were predominantly in the northern and southeastern sections of Norfolk, as indicated by the shading.  Also, the extreme southern stretches of the city might be suitable.  Areas not shaded either did not meet the selection criteria, or were located outside of the city. 

The percentage of racial minorities presented in Map 3 provided too vague a distribution of minority populations, and was therefore omitted from the analysis.  It did however provide a visual clue in isolating suitable areas to Norfolk, as the dark red polygons on the map show where a higher number of minorities are located compared to the total population of the study area (and most of the higher values were located in Norfolk). 

Section IV:  Discussion and Conclusion

            The results presented in Map 7 are by no means intended to be exhaustive in their area coverage.  Other factors, such as soil moisture, soil contamination, land cover, land ownership, and other geographical criteria will need to be taken into account. 

The analysis result itself was imperfect.  The shaded area of suitability in Map 7 also covers water bodies, roads, buildings, and other sites incompatible with biointensive agriculture.  

As such, there exists a possibility for further research in determining site suitability.   More detailed criteria, such as only including sites that are abandoned lots, or where a specific plot of land is known to be available, could help improve location determination.  A more detailed spatial analysis such as excluding unusable areas like as bodies of water, roads, buildings, and other non-suitable areas could have provided a more detailed result than what was presented in Map 7.  Nevertheless, a generalized overview of suitable sites for a new biointensive farm plot based upon the above-mentioned criteria can assist The Micro Farms with their location selection. 

Section V:  References

Grow Pittsburgh. (2013 June 17). Why grow food?. Retrieved from www.growpittsburgh.org/why-grow-food/

McKenzie-Mohr, D., & Smith, W. (1999). Fostering sustainable behavior: An introduction to community-based social marketing. (pp. 71-81). Gabriola Island, BC: New Society Publishers.

Micro Farms. (2013, June 28). Micro farms. Retrieved from http://ecocycling.org/activities/projects/microfarms/whatwedo/

National Poverty Center, University of Michigan Gerald R. Ford School of Public Policy. (2010). Poverty in the United States: Frequently asked questions. Retrieved from http://www.npc.umich.edu/poverty/

Norfolk Public Schools. (2013). All schools list. Retrieved from http://www.nps.k12.va.us/index.php/nps-schools

Pearce, J., Blakely, T., Witten, K., & Bartie, P. (2007). Neighborhood deprivation and access to fast-food retailing. American Journal of Preventive Medicine, 32(5), 375-382.

United States Census Bureau. (2013, June 24). Tiger products. Retrieved from http://www.census.gov/geo/maps-data/data/tiger.html

Warren, L. (2003). American environmental history. (1 ed., pp. 298-322). Malden: Blackwell.

 

Map 1:  Study area.

 

Map 2:  Current distribution of biointensive 

farms and community gardens in Norfolk and Virginia Beach.

 

Map 3:  Percent of minority people versus total population per census block.

 

Map 4:  Ratio of minorities to white people per census tract in Norfolk and Virginia Beach.

Map 5:  Concentration of minority people in Norfolk and Virginia Beach.  

Map 6:  Census tracts in Norfolk at or below an average household income of $22,113.  

Map 7:  Analysis results

Figure 1:  Methodological summary for a suitability analysis for a location for a new biointensive farm plot in Norfolk, Virginia.  See text for details.   Click to enlarge.