Farming in the City: How Does the Altered Urban Environment Influence Cropping System Productivity, Ecology, and Profitability?

Final Report for GNC14-198

Project Type: Graduate Student
Funds awarded in 2014: $9,994.00
Projected End Date: 12/31/2015
Grant Recipient: University of Illinois Urbana-Champaign
Region: North Central
State: Illinois
Graduate Student:
Faculty Advisor:
Dr. Jack A. Juvik
University of Illinois Urbana-Champaign
Faculty Advisor:
Sam Wortman
University of Nebraska - Lincoln
Expand All

Project Information


Urban agriculture is a growing trend in the United States, but little is known about the effects of urban abiotic and biotic effects on agroecosystems in cities. Objectives of this study were to use six test gardens in the greater Chicago area to measure the urban microenvironment and compare these measures to crop production, insect abundance and population dynamics, and soil microbe indicators in raised bed pots. Results and objectives were presented at grower field days, presentations at community events, and scientific conferences. Results are in preparation for publication in peer reviewed scientific journals.  Micrometeorological towers measured atmospheric conditions. Temperature and CO2 were positively correlated with urbanization, whereas relative humidity and overall light interception were negatively correlated. Ground level ozone was higher in the peri-urban sites. Spring and fall planted cool season crop production was correlated with increased temperature and reduced ozone concentrations. Light and CO2 were most correlated to summer planted warm season crop production. Crop pests corn earworm and cabbage looper had highest abundance in the rural site, whereas European corn borer had similar abundance among the sites. Insect orders Diptera (flies) had lower populations in the peri-urban sites, whereas the order Thysinoptera (thrips) had higher populations in the peri-urban sites. All other orders collected, including common pollinator orders, were not different among the sites. Soil microbial biomass and microbial type did not differ across the urban to rural gradient within the raised beds, but there was differences between spring and fall sampling dates. Economic analysis showed highest overall profitability at an urban site mainly driven by tomato yield at that site, and higher overall profitability of urban sites and lowest overall for peri-urban sites due to low yield of these sites. Early and late crop species were generally more profitable in the urban environment than the peri-urban or rural. This research demonstrates that the urban environment is dynamic and urban farmers can choose crop species and varieties to increase profitability. Further research can focus on microenvironmental differences for further understanding light, temperature, and ozone dynamics, and studies on crop species and variety differences can enhance productivity in the urban setting.


Urban agriculture, the production of food crops within urbanized ecosystems, is a relatively common practice in developing nations and is becoming more prevalent in the U.S. Urban agriculture is viewed by many as a promising method for reducing urban externalities and improving diets, livelihoods, and community cohesiveness in blighted urban areas (Armar-Klemesu, 2000). Studies in Chicago, IL and Oakland, CA demonstrate an increasing concentration of home gardens, community gardens, and commercial farms in urban areas (Taylor and Lovell, 2012; McClintock, 2008). The North Central region of the US is especially well-suited to increasing urban food production because of the post-industrial depopulation of cities like Chicago, Milwaukee, and Detroit. Interest in urban agriculture in city centers in the North Central region has grown significantly as evidenced by the rise of local food movements, farmers markets, and organizations dedicated to urban agriculture.

Urban agroecosystems have unique challenges compared to rural counterparts. Carbon dioxide flux and ambient concentrations are higher in urban settings. Pollutant concentrations in urban areas can flux far beyond baseline levels in rural ecosystems (Trusilova and Churkina, 2008). Potentially harmful pollutant fluxes include primary (NOx, SO2, and PM10) and secondary (PAN, O3, and ethylene) pollutants. Each of these pollutants can, to some degree, limit plant production and alter agroecosystem function. Additionally, decreased albedo and evapotranspiration, and increased thermal admittance causes an urban heat island (UHI) effect, the increase in temperature of urban versus adjacent rural areas (Taha, 1997). Consequently, vapor pressure deficit (VPD) increases which leads to increased plant water stress.

Agroecological function can be effected by the altered urban environment. Effects of UHI are greatest in the spring and fall seasons, at night, and during inversions. Thermally regulated ecological processes, such as germination, flowering, senescence, hatching, or morphogenesis may be affected by UHI in urban agroecosystems (Ziska et al., 2007). Plants in urban agroecosystems are likely to benefit from additional growing degree days and frost free days, but also experience greater heat stress and vapor pressure deficit. In addition to UHI, wind speeds in urban areas tend to decrease due to increased surface roughness causing reduced plant stress, but also better conditions for plant pathogen infection.

Altered abiotic conditions in the urban agroecosystem will likely influence important biological functioning. Soil microbial community composition and diversity in urban ecosystems is markedly different from natural or rural ecosystems (Papa et al., 2010). Most notably, diversity of soil microbial communities tends to decrease with the level of urban soil disturbance. Healthy soil microbial populations are closely linked to soil quality in agroecosystems, but little is known about how soil microbial communities will respond to the complex suite of abiotic stressors in a managed urban environment. The unique microclimate and diversity of plant hosts in urban agroecosystems will also likely alter urban plant pathogen dynamics. Insect dynamics in the urban ecosystem are important because of the potentially beneficial ecosystem services provided (e.g., pollination) and damages incurred (e.g., herbivory) by insects. Insect diversity is generally reduced in urban ecosystems, but often spikes in the urban to rural transitional zone, a common phenomenon in natural ecosystems (i.e., the edge effect or ecotone) (Jones and Leather, 2012).

Armar-Klemesu, M. 2000. Urban agriculture and food security, nutrition and health. Grow, Cities Grow. Food Urban Agric. Policy Agenda, 99–118.

Jones, E. L., and Leather, S. R. 2012. Invertebrates in urban areas: A review. Eur. J. Entomol. 109:463–478.

Kaza, N. 2013. The changing urban landscape of the continental United States. Landsc. Urban Plan. 110:74–86.

McClintock, N. 2008. From Industrial Garden to Food Desert: Unearthing the Root Structure of Urban Agriculture in Oakland, California. Inst. Study Soc. Change.

Papa, S.,G. Bartoli, A. Pellegrino,and A. Fioretto. 2010. Microbial activities and trace element contents in an urban soil. Environ Monit Assess 165:193–203.

Taha, H. 1997. Urban climates and heat islands: albedo, evapotranspiration, and anthropogenic heat. Energy Build. 25:99–103.

Taylor, J. R., and Lovell, S. T. 2012. Mapping public and private spaces of urban agriculture in Chicago through the analysis of high-resolution aerial images in Google Earth. Landsc. Urban Plan.

Trusilova, K., and Churkina, G. 2008. The response of the terrestrial biosphere to urbanization: land cover conversion, climate, and urban pollution. Biogeosciences 5:1505–1515.

Ziska, L. H., George, K., and Frenz, D. A. 2007. Establishment and persistence of common ragweed (Ambrosia artemisiifolia L.) in disturbed soil as a function of an urban–rural macro-environment. Glob. Change Biol. 13:266–274.

Project Objectives:

In order to improve understanding of urban food production, this project aims to characterize environmental conditions of agroecosystems across an urban to rural gradient in Chicago, IL and use this environmental data to understand variability in crop production of several common vegetable crops, insect population dynamics, and soil microbial dynamics. This project leveraged several entities within the Chicago metropolitan area to set up and manage research plots for the characterization of crop production, soil microbes, and insect populations. Also an advisory board of urban farmers and stakeholders met in fall of 2014 to discuss results and improve coordination for better management and data collection. Farmers and community members met during two field days and two community presentations to listen to results from the trial and interact with the researchers. Researchers presented results at two scientific conferences to peer researchers in the urban ecology and agroecology fields. One scientific journal paper has been completed, and three additional peer reviewed journal articles are being prepared for publication. Further publications, stakeholder meetings, and scientific presentations will gauge performance of this project. 


Click linked name(s) to expand/collapse or show everyone's info
  • Dr. Carl Bernacchi
  • Dr. Jack A. Juvik
  • Joy Kaminski
  • Wade Kuipers
  • Dr. Sam Wortman


Materials and methods:

Experimental vegetable crop gardens were established along an urban to rural latitudinal transect (approximately 41° 50’) in Chicago, IL (see Figure 1). Experimental sites were 75 m2 and contained 40 raised-bed pots (0.4 m3) filled with a single mixed compost-soil-sand mixture. Annual soil samples of the mix showed sufficient macro and micronutrients. Soil moisture was measured continuously (Watermark sensors, IRROMETER, Inc.) and drip-irrigation maintained soil moisture near field capacity. Each site was divided into four blocks with eight replicates of each split-plot factor (two replicates per block). Main-plot crop species treatments included; spring planted (early April) Brassica oleracea (kale) and Allium cepa (onion), summer planted (late May) Solanum lycopersicum (tomato), Capsicum annuum (pepper), and Phaseolus vulgaris (snap bean), and fall planted (late July) Brassica oleracea var. gemmifera (Brussels sprouts) and Beta vulgaris (beet). Two cultivars of each crop species were included as split-plot treatments to assess genotype by environment interactions. Spring crops were followed by fall crops as a double crop, which is a common strategy in vegetable production and allowed maximization of experimental treatments per season.

Micrometeorological towers were located directly adjacent to each experimental site and equipped with micrometeorological and pollution sensors. Sensors included a temperature and relative humidity probe, cup anemometer, wind vein, pyranometer, CO2 infrared gas analyzer, and an ozone monitor. Data from each sensor was collected continuously and logged as average, maximum, and minimum every 60 minutes.

Plant physiological response across agroecosystems was quantified throughout the season with measures of relative growth rate, biomass production, and fruit yield and quality (e.g., sugar, micronutrient, and antioxidant content). Project cooperators (Master Gardeners and growers) and local student interns participated in weekly care, harvest, and data collection to reduce travel costs for the project coordinator between campus (Urbana, IL) and experimental sites in Chicago. Plant pathogens were scouted weekly, identified, and infection was quantified with disease rating scales. Soil samples were collected aseptically from each main-plot experimental unit to a depth of 4 cm at the beginning and end of each growing season to quantify soil microbial community composition across sites. Two crops were picked to sample from and 16 samples were taken per site. Phospholipid fatty acids (PLFAs) were extracted from soil samples and used as biomarkers to fingerprint microbial community composition, biomass, and fungal to bacterial biomarker ratios. Insects were collected throughout the growing season at each site using tanglefoot traps and pheromone baited nylon cone traps, identified, and counted to determine community diversity and abundance of functionally important species (e.g., pollinators and pests). Insect herbivory was scouted and quantified weekly with visual ratings. Species of collected insects include cabbage looper (Trichoplusia ni), European corn borer (Ostrinia nubilalis), and corn ear worm (Helicopvera zea). Student interns were trained in collecting insect, plant damage, and plant physiological measures.

Statistical analysis included a combination of generalized linear mixed models to assess the influence of site-years on biotic response variables and structural equation models (SEM) to explore complex relationships among microclimatic factors, pollutants, and biotic responses (plants, pathogens, soil microbes, and insects). Micrometeorological data was averaged or summed across the growing season for each crop. Overall site crop yield average was compared to climactic data in SEM. Model comparison of all micrometeorological factors to yield were used to determine best fit models.

Profitability analysis quantified fixed costs (land, raised beds, soil, irrigation, mulch and fencing) and variable costs (plants, seeds, water, labor, and supplemental fertilization) and compared costs to crop returns. Fixed costs are an average of the rural, peri-urban, and urban costs and included installation labor of fixed assets, depreciation schedule, and opportunity costs of capital. Variable cost analysis included general labor costs, labor costs of management, market value of plant starts, water rate, and farmer’s market fees. Crop returns are calculated from the yield of vegetables in the study compared to the market value of organic produce as an average over the Chicago area farmer’s markets (USDA Market News Service).

Research results and discussion:

Sites in this project were characterized as urban (Garfield Park and Growing Home), peri-urban (Cantata and Cantigny), and rural (St. Charles and Kuipers) (see Figure 1). The boundaries of these definitions are somewhat subjective, but the classifications are helpful in explaining plant and microclimatic responses in this study.

Crop Response

Vegetable crop yields from variety within site within year are presented in Table 1. Interactions of year by site by variety were significant across most trials and means separation were determined using the LSD method. Spring and fall planted kale and Brussels sprouts, which are both of the cool season B. oleracea species, had highest yields in the urban sites in 2013 and 2014, but in 2015 the rural site St. Charles had highest yields. Fall planted Brussels sprouts did not have measurable sprout yields outside of the urban site Garfield in 2013 and 2014 but in 2015 yields overall were much higher, and the rural site St. Charles had the highest yields overall. In 2013 spring planted onion had highest yield in the most rural site Kuipers, but in the other years the urban site Garfield had the highest yield. Fall planted beets had an unusual response, in 2013 the highest yield was in the peri-urban sites, but in 2014 and 2015 had highest yield in urban sites and lowest yield in rural sites. Summer planted tomatoes had highest yield in rural sites, followed by one peri-urban site Cantigny, except 2013, with highest yield in an urban site. Summer planted peppers yielded highest in urban sites in 2013 and 2015, but had highest yield at a peri-urban site Cantigny in 2014. Pepper yield was higher in 2014 than the other two years. The iso-line bean resistant variety had moderate yields across sites, with slightly higher yields in rural sites in 2013 and urban sites in 2014 and 2015. The susceptible iso-line had higher variability with highest yield in rural sites in 2013 and 2015 and an urban site in 2014. Overall crop yields were highest at the Garfield urban site in 2013 and 2014 and highest at the St. Charles rural site in 2015. The peri-urban site Cantata had the lowest overall yield in every year when combining all the individual crop responses.  Within years, crop varieties generally had the same trends among sites, although the magnitude of yield was different in kale, pepper, tomato and Brussel sprout.

Micrometeorological Measures

Micrometeorological measures are presented in Figure 2. This figure is a panel of individual results of each measure integrated from June to September of all three years. The bottom (horizontal) axis is the sites from rural to urban and the scale of the vertical axis is given in the title of the panel. Yellow boxes represent daytime (0530 to 1930) averages and grey boxes nighttime.

Temperature and subsequent growing degree day (given in Table 2) were lower overall in rural sites and highest in urban sites. Solar radiation and overall light interception (Table 2) were highest in the rural environments, and lowest in peri-urban sites. Canopy analysis with fisheye cameras gave us an idea of canopy light transmittance which is given in Table 2. Light transmission coefficients, a measure of light getting through canopy on a scale of 0-1, were low at peri-urban sites and high at rural and urban sites. Leaf area index, a measure of leaf canopy interception of light, had the opposite trend, being much higher at the peri-urban site Cantata.

Relative humidity (RH), vapor pressure deficit (VPD) and wind had very clear trends across rural to urban gradients. Rural sites had highest RH and closer to the urban area got progressively lower. The opposite trend was found in the VPD with highest averages in the city. Wind speed was highest in the rural areas, lowest in the per-urban and moderate in the urban sites. These trends were interesting, but RH, VPD, or wind speed had little effect on plant growth or production (presumably because crops were irrigated daily).

Aerial pollutants, CO2 and ozone, had very site specific trends. Plants need CO2 for growth, so higher concentrations could provide fertilization effects in plants that use the C3 photosynthetic pathway. Daytime CO2 levels were highest in the urban site and lowest in the rural sites and the variability was lower in the urban site. Nighttime CO2 was more variable and higher than daytime at the rural site, likely due to increased vegetative respiration.

Ozone is a highly toxic oxidant to life and near ground levels are in parts per billion. Plants are especially susceptible to ozone because they are constantly exposed to ambient levels. Ozone levels were moderately low in the rural site, higher in the peri-urban sites, especially the site Cantata, and low in the urban sites. Ozone is formed from complex reactions of pollution, light, and catalysts. This makes formation at the pollution source (the city) unlikely, but as the pollution moves out of the city and into vegetation canopies, it will start to form. Thus, the levels are higher outside the city center. Cantata is at the edge of the green belt of Chicago and this likely explains the high concentrations. Season long ozone measures (Table 2) show the same high amount of ozone exposure in the peri-urban sites.

Structural equation models

Belowground factors in this study were deliberately kept constant across sites to limit crop effects to aboveground factors. This method was relatively effective as the models using micrometeorological measures against plant responses showed most of the variation was explained by these climactic measures. Structural equation models (SEM) were used to compare plant measures to micrometeorological measures.

SEM models are pathway models that find causation of several predictor variables (latent or manifest variables) on response variables. The SEM models presented are finding casual pathways from meteorological variables (predictor) to crop production (response) values. Models of each crop are given in Figures 3-9 with response variables as the two varieties of each crop. All the models have very good fit statistics and R2 values are between 0.3 and 0.95 (The R2 is calculated from one minus the error term of the response variable). Overall fit statistics for crop and variety within crop SEM models is presented in Table 3.

Seasonality of the crops was important for understanding causation of meteorological measures to plant responses. From the model, factors that come from temperature measure (growing degree day and temperature averages) and ozone measures (AOT40 and average ozone) account for most of the significant causual pathways in spring and fall crops. Increasing ozone levels caused a negative crop response and increase in average temperature measures caused a positive crop response. Summer crops were not influenced by temperature measures and only beans were negatively affected by ozone. Summer season crop response was positively influenced by increasing light levels and CO2.

Crop responses in early and late season appear to be driven by temperature and ozone, whereas summer crops have more variation in what causes yield differences, but light is an important factor. Urban farmers can take advantage to the extra frost free days (averaged 15 extra over the three seasons) and higher temperatures in the fall and spring to extend their growing season, but must understand that specific locations in peri-urban areas may have elevated ozone, which is a limiting factor. Light attenuation is very important to summer warm season crops. Site selection is very important to reduce light attenuation from canopy for summer crop production. Carbon dioxide and distance to city center are significant factors in the models, but because the correlations are positive and negative, these factors are likely artifacts of other variation along the urban to rural gradient like biotic stress or unmeasured climatic factors.

 Insect population response

Insect population and insect pest captures are presented in Table 4. Insect pest species corn earworm and cabbage looper had the highest populations in rural sites in both 2014 and 2015; whereas aphids, and European corn borer did not show any difference between the six sites. Insect populations from tanglefoot traps showed consistent year over year trends. Flies and thrips had highest populations and variability among sites. Thrips had highest populations in peri-urban sites, followed by the rural site Kuipers. It had low overall populations in the most urban site Garfield and the rural site St. Charles. Fly populations were low in peri-urban sites and high in rural sites and the urban site Garfield. All other order of insects did not have large differences between sites in this study. Further studies on how site characteristics and micrometeorological measures correlate to insect population dynamics will be done in the coming months and reported. 

Soil microbe analysis

Phospholipid fatty acid analysis of soil samples revealed a healthy soil microbe community with above average microbe biomass and healthy ratios of bacteria to fungi and a high diversity index. Soil samples were taken in fall of 2014, spring of 2015, and fall of 2015. Table 5 has the results of PLFA analysis of the three sample dates. Time of sampling explained a significant amount of variability in the microbial data, while site and crop type explained very little variation. Principle component analysis of fatty acid markers and biological markers of diversity revealed loading scores that were highly grouped by sample date and randomly spread across site and crop type. Figure 10 has the comparison of the first and second principle component scores with labels for sampling time, site, crop type, and blocking factor. Sampling time differences have biological meaning because soil microbial dynamics change according to the temperature and moisture of the environment. These results suggest that the microclimate across the sites and crop type have limited effect on soil microbe population dynamics in raised bed crop production in the urban to rural transect.

Participation Summary

Educational & Outreach Activities

Participation Summary:

Education/outreach description:

Outreach activities included public demonstrations, grower and stakeholder involvement, and presentation of results to targeted audiences. Story boards were placed adjacent to each site and were in view of publically accessible paths at four of the six sites. Two field days at the St. Charles research center hosted 45 and 50 participants in 2014 and 2015, respectively, and peri-urban and rural farmers learned about project findings. In December of 2014, the urban advisory board met to discuss project findings and the future path of the research. In December of 2015, a presentation with 35 participants from the neighborhood was given at Garfield Park Conservatory. Then in January of 2015 a presentation was given to 40 community members and park volunteers at Cantigny Park in Wheaton, IL. Future planned presentations include a joint presentation to community members and staff at Cantata adult life services in Brookfield, IL with anticipated attendance of 45 individuals and a meeting of the Chicago Advocates for Urban Agriculture (AUA) spring meeting which typically draws 150-200 participants. Blog posts are going to be posted to the AUA website and the UIUC Urban Agriculture Lab in the next two months as results are finalized. 

Results have and will be presented at scientific symposiums and in peer reviews journals. Ecological Society of America (ESA) has sections for both agroecology and urban ecology. Yearly ESA meetings often have agroecology themed symposiums. In 2014 and in 2015 results from this project were presented in presentations and posters (Wagstaff et al. 2014, Wagstaff et al. 2015). The feedback and contacts made from these presentations have led to this project being repeated in Salt Lake City, UT and many other fruitful relationships. An inaugural conference for urban food systems will be held in Olathe, KS this year where further results will be presented. One paper (Wagstaff & Wortman, 2013) from this project has been published and three further papers are being prepared for submission to peer reviewed journals currently.

Wagstaff, RK, C Bernacchi, AE Davis, JA Juvik, and SE Wortman. 2014. Farming in the city: How does the urban atmospheric environment influence vegetable crop physiology? ESA Annual Convention, Sacramento, CA, 2014.

Wagstaff, RK, CA Clay, KM Ku, C Bernacchi, AE Davis, JA Juvik, and SE Wortman. 2015. Vegetable crop response and insect prevalence in urban agroecosystems. ESA Annual Convention, Baltimore, MD, 2014.

Wagstaff, R. K., & Wortman, S. E. 2013. Crop physiological response across the Chicago metropolitan region: Developing recommendations for urban and peri-urban farmers in the North Central US. Renewable Agriculture and Food Systems, FirstView, 1–7.

Project Outcomes

Project outcomes:

Impacts of this project include a greater understanding of the fundamental differences between the rural and urban agroecosystem and ecological function in respect to vegetable production. Future impacts include directed efforts to take advantage of altered urban environments to improve management and crop selection for urban farmers. Examples of this impact include selecting early season crops and planting earlier in the season to get produce to market earlier than rural counterparts, selecting varieties more suited for urban environments such as warmer climate zone vegetables and shade tolerant varieties. Additionally, breeding programs can focus efforts on ozone tolerant varieties of cool season crops. Further impacts would include improved policy to open up land and provide incentives to urban farmers due to production potential of urban farms compared to rural counterparts. The results of insect populations and soil microbes informs individuals and policy makers of the suitability of the urban environment for larger scale food production. Farmers considering crop production in the peri-urban environment can be informed of the challenges involved and plan for specific cropping systems to overcome reduced production potential.

Economic Analysis

All plots were managed according to USDA organic methods, and the type of production system used would typically serve CSAs or famers markets. Table 6 has fixed costs for each pot and variable costs for each crop and prices averaged from 2013-2015 across the Midwest for conventional and organic retail and farmers markets. Prices are from the USDA market news website and from Chicago farmers market forums. Table 7 has the yields per pot (a culmination of varieties) or each year and overall average from three years. These yield averages were compared to each predicted crop price category.  Fixed and variable costs were compared to net profit from produce price at farmers markets and overall profit for each crop from each site was determined.  Fixed cost were split between the double planted crops onion/Brussels sprout and kale/beet.

Fixed costs include costs outside of direct crop production, given in Table 6. Soil price is 0.5 yards of soil per pot at $30 per yard with depreciation over 5 years. Land rent is the opportunity cost of growing on owned land and direct cost of renting. This is from a 3-4% return expectation on land price per year according USDA land rent guidelines. The land price was estimated to be $50,000 per acre and the overall area of the plot was 750 ft2 thus 4% of land value is $2000 multiplied by 0.017 (the plot area in acres) to get a price of land. Then divide the price by the number of pots (40). Fertilization is done by topping the pots with more compost. About 4-6 inches, or about 15%, of the pots are used up as the organic material is broken down in a year, and this amount of compost is added to top the pots every year. Landscaping includes mulch, weed control, and fencing around the sites. Site irrigation systems cost around $200 per site and have a useful life of 5 years or more. Farmers market prices in the Chicago region are between free and $15 per week and some charge a 10% sales charge. The estimate was on the high side of $15 per week and 10% charge for a booth.

Variable costs are costs of seedlings, labor cost, and trellis material. All seedling costs are based on going rate of flats of small seedlings similar to seedlings used in this study. Labor costs are based on the amount of time needed to plant, weed, and harvest produce from each pot. Labor costs were highest for tomato and pepper because of trellising, training, and pruning. Trellis material is cages for tomatoes built from sturdy fencing and bamboo stakes for the peppers. Fixed costs were split between multiple crop companions onion/Brussel sprouts and kale/beets.

Profitability of crops, based on per pot production, was due more to crop type than site location (Table 7). Tomatoes were the most profitable crop because of the price and yield compared to other crops, even after having the highest variable costs due to extra management and trellising. Tomato profits were highest in a rural site due to highest yield of the heirloom variety in the rural sites. Pepper is another profitable crop that had limiting yield in the peri-urban sites due to light attenuation. Pepper profitability was highest in the most urban site and the two rural sites where light levels were highest. Spring planted kale and onion were crops with similar profitability to peppers, but the most profitable sites were in the urban areas. Kale and onion are both cool season crops and can be double cropped, making them potentially more profitable.

Beans, beets, and Brussels sprouts were less profitable, and had overall negative returns. Bean yield was low compared to other crops, and the price is moderately low per unit thus making it a less than ideal crop in space-limited systems. Additionally, beans can easily be double or triple cropped because they can be planted in early June and are done producing in late August. Beets are similar with low yields and low price, but beets have a very short growing season (4-6 weeks) and can fit in a shorter window of growth. Brussels sprouts did not grow well in 2013 and 2014. In 2015, yields were much greater but they were still not profitable in any site. Brussel sprouts take long seasons to grow and in this production system do not make a very profitable crop.

Overall site profitability across crops is given in Table 6. Profits were highest in the urban site Garfield in 2013 and 2014. This is due to high overall yields at that site. The rural site St. Charles had highest profitability in 2015 and highest overall profitability. The heirloom tomatoes did especially well at this site, which drove profitability. Disease pressure at the urban site Growing Home affected tomatoes and peppers, lowering profitability. The peri-urban site Cantata had lowest yield in most crops and overall the lowest profitability of any site. Overall, the urban environment has some economic advantages of extended growing season for early and late season crops, while the rural environment is better for producing tomatoes which drive rural site profitability. Even though this cropping system is not optimized for profitability there is information here that may help to inform design of a profitable system. The financial viability of urban agriculture will depend in part on planting the right crops and using the extended season afforded by the urban environment to add early and late season crops, and also planting crops that have the potential for higher profitability.

Farmer Adoption

No farmer adoption study was planned for this project.


Areas needing additional study

The complexity of the interactions between environment and plants in an urban agroecosystem or even outside the agroecosystem have had limited research focus. Considering an 80% urbanization rate in North America and future trends show 66% of world population will reside in cities by 2030, an understanding of the effect of the urban environment on crop production is warranted. Light and ozone dynamics in microclimates within the urban and peri-urban environment to better characterize site specific growing challenges can be studied. Insect and disease pressures in this study were low, but further pathology or insect studies can be done to characterize more site specific impacts of urbanization on biotic stresses.

More applied avenues of study can help urban farmers to plan and implement for productive and profitable enterprises. First, characterizing effective crop planting and rotation schedule to maximize plant production and economic return for farmers in the urban environment can be studied. Second, studies can be done to characterize crop variety differences with respect to the urban and peri-urban suitability. An unpublished study looking at participatory selection and breeding for tomatoes has been done in Chicago, IL by Dr. John Taylor. Grower participation in selection would make a study easily scalable across many microclimates and crops. Third, studies in best management practices for raised bed and cap and fill systems would prove helpful as many urban farms and community gardens are growing in these systems. Urban farming has many areas where research can help build a more stable and science-based production system.

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.