A Multi-Criteria Analysis of Arable Land on Bennington College Campus as it Relates to the Relocation of the Student Garden

From BenningtonWiki
Jump to: navigation, search

Return to Bryan Markhart

Bmarkhart 10:25, 12 December 2011 (UTC)


The Bennington Sustainable Food Project (BSFP), a student group at Bennington College concerned with food related issues is working on relocating the current student garden to a larger more permanent space. In this study I used QGIS to perform a multi-criria analysis of all arable land on Bennington College campus that weighed access by roads, proximity to irrigation sources, soil quality, and slope against each other to create a score for each section of land as it relates to that spaces viability as a student garden. The study results show that the greater area around Jennings, commons lawn, the area just south of third street, and part of shingle field are the best locations on campus for a student garden. Due to some design errors however, I do not find the results of this study to be a comprehensive or accurate prediction of the quality of different land on campus.


As apart of the growing food movement in the United States, sustainability oriented student gardens and farms have begun to pop up on college and university campuses all over the country. Not unlike these many other schools there is student interest at Bennington College in small scale farming that helps to decrease environmental and ecological impacts of an agricultural system while fostering community and helping people to understand their food system and appreciate what it takes to get them their three meals a day with snacks in between. During the ’09-10’ school year the BSFP, a student group at Bennington College concerned with food issues on campus and in the greater community began working on the latest attempt to have a student garden on campus. The BSFP was able to secure two 30x30ft plots in the community garden on campus for it’s first growing season in the summer of 2010, and then a third plot during the summer of 2011. Thus far, this space in the community garden has worked well for the student garden in its infancy but interest and ambition has outgrown the limitations of this space. In order for the student garden to grow into the working college farm that Bennington College, as a progressive school that prides itself on always being the next big thing, needs to support. The student garden must be in a space that can be built upon year after year with room for expansion. During the summer and fall of 2011 the BSFP began working for a new location for the student garden that better satisfies it growing needs, specified in the Permanent Garden Space Proposal (Permanent Garden Space Proposal), to the Bennington Sustainability Committee.

This GIS based research project was designed to perform a multi-criteria analysis of arable land on the Bennington College campus in order to determine the agricultural value of this land as it relates the needs of the student garden. From this one can determine the optimum lands for the relocation of the student garden. My analysis was based on four criteria: proximity to college roads and irrigation sources, slope, and soil quality.


For this analysis I compiled the following preexisting datasets: the USGS soils survey as provided by the Vermont Center for Geographic Information, the Bennington County roads map as provided under the datasets section of Kerry Wood’s faculty website, as well as a DEM for the North Bennington area found in the same place. I also created a point shapefile showing the location of all irrigation sources on campus. I used the slope tool in QGIS to create a slope layer for lands on campus. I also ran the buffer tool on the roads map, disregarding off campus roads, and the irrigation map; this created layers showing all land within 100m of both roads and irrigation sources separately. These additional manipulations yield the input maps found in Figures 1-5 below.


Table 1. Scoring system applied to my input maps

I created a scoring system that gave value to the above criteria based on importance of the success of the student garden. I saw soil quality and slope to be the two most important criteria as they are two things that are much more difficult to change than water sources and access. Based on the AgVal data rated in the USGS soils map as seen in Figure 3, one being the best, 11 being the worst, I created a scoring system with three points as the highest and minus two points as the lowest. For areas with drainage issues, signified as a ‘d’ next to the number (ie. 3d or 11d), I subtracted two points from the score awarded based on value number, as seen in Table 1. For slope I chose my values based on information provided by the USDA regarding agricultural viability relating to slope. This information claimed land with a slope steeper than 8% would prove difficult for agricultural production. I used this slope as my cut off, and gave it a very low score of minus two. I then split up the areas with slopes shallower than 8% into those greater and less than 4% slope and gave them values of zero and one respectively. With regards to irrigation and roads I simply said anything within 100m had a score of one and anything further away had a score of zero.

After applying the scoring system I have described above to my data-layers I created a master shapefile which added together all of the values above, creating a new shapefile with a polygon for each unique value. After many technical problems I managed to do this by rasterizing each shapefile, loading them in saga and combining the grids into one. I then exported it back to a GeoTIFF format and loaded it back into my QGIS mapset and polygonized it. I then clipped this new master polygon file with a shapefile of all the arable land on campus that I had traced from a 2009 aerial photograph (Figure 4). This work yielded the map shown in Figure 6 with a gradation of the most and least suitable for a student garden.


Agricultural Value of all Fields on Bennington College Campus
File:Final Ma.gif
Agricultural Value of all Fields on Bennington College Campus

Bases on this final map it is clear where the best land is on campus. All of the fields in the broader area surrounding Jennings, the soccer field and the area at the end of third street, and most of Shingle fields, all scored in the four and five range; while the entrance field, the greater end of the world, and the Orchard field all scored below minus two.

More precisely the only locations scoring a top score of six points were small patches located in the area around Jennings, while areas scoring five points included commons lawn, the area just around Longmeadow cottage, a small portion of the Ohio field, the circus field, and part of shingle field, making these areas the locations I will focus on in my discussion.


This study gives us a very rough idea of where to look on campus when relocating the student garden. The results of this study suggest that the area to the north of Longmeadow is, has the largest continuous tract of quality land on campus, supporting that it is the optimum location for food production on campus. The circus field, could be a very strong candidate for the student garden relocation, it has scored well and is fairly a large space. While commons lawn, along with the area past the end of third street, also have very respectable scores of five, they can not be considered as candidates for a garden for other reasons: commons lawn gets too much traffic and likely turn other students off of the idea of the garden if it were to take away that space from them, and the area past the end of third street, although it does not show up on the USGS soils map, has been anecdotally observed to have significant drainage issues. Part of the Ohio field also scored well in this study, however due to aesthetic reasons that are rather difficult to include in a study of this kind, I would not strongly suggest this location.

I feel that in order to use this study in any real way to inform the decision making process of relocating the student garden, it requires some tweaking to make it more precise, and more thorough. First, the scoring system may need some reweighing done in order to make it more balanced. As it is right now I feel that the minus two score for areas with slope steeper than 8% may be a little too significant. Perhaps I could score land with slopes between 8% and 12% at minus one, and then anything over 12% slope at minus two. The same could be applied to proximity to roads and irrigation, perhaps make land with in 100m two points, and land within 150m one point.

The way in which this study looks at proximity to irrigation is somewhat flawed to begin with, for the sake of a more complex study, I chose to include water spigots on buildings as viable irrigation sources, which they have been in the gardens first two seasons, and will likely continue to be for several years to come, but as production increases and irrigation demand with it, the price of city water will become a significant factor. For this reason, we must look for areas that can be irrigated from rainwater catchment sources or dug wells, two things that are not incorporated into this study. Another factor that is not looked at in this study is agriculturally significant ecological characteristics of the land surrounding the garden. In order for the garden to successfully practice organic, and ecologically based farming techniques it must be located in close proximity beneficial habitats that allow predator insects and pollinators to access the fields. This is something that I could not find any quantifiable data for as far as where the best beneficial habitats are nor how to rate them.

There are other important characteristics that the BSFP is looking for in their new student garden location that are not taken into account in this study. It is important for the garden to be located in a space that is accessible to and enjoyable for students to be in. This means fairly close to student foot traffic, and in aesthetically pleasing surroundings, these are two factors that are difficult to quantify in any sort of unbiased and balanced way and were therefore not included in this study.

One all of the adjustments mentioned above have been incorporated into the study, further analysis may also be helpful in deciphering the final map that is generated. It could be good to look at the average student garden value of each field on campus, and also to incorporate land area into the decision making.


This study was a good preliminary look at the agricultural value of land on campus as it relates to the relocation of the student garden, however the criteria are not comprehensive enough, nor the scoring system balanced enough to draw and legitimate conclusions from these results. With some tweaking to include more information that gives a better representation what would be the optimal location for the student garden, this study could be very helpful in this decision making process. However, as of yet I do not feel confident in my results.