Urban planners have an opportunity to address food insecurity and other urban food system challenges including production, consumption, waste management and recycling by co-creating solutions with urban farmers through participatory processes and investing in community-led solutions. In our systematic review of the literature on whether urban agriculture improves urban food security, we found three key factors mediating the effect of UA on food security: the economic realities of achieving an economically viable urban farm, the role of city policy and planning, and the importance of civic engagement in the urban food system . A radical transformation toward a more equitable, sustainable and just urban food system will require more responsible governance and investment in UA as a public good, that is driven by active community engagement and advocacy.The effects of natural events such as a disease outbreak are typically difficult to measure since simultaneous shifts can occur along several dimensions. The analysis of longitudinal data may reveal dynamic change that would be hard to recognize when only cross sectional data are used . With panel data, one can examine when depression on production occurs -if there is any-, vertical grow weed either at the time of the outbreak, soon after or even before, and for how long such depression occurs.
Previous studies have addressed the impact of porcine reproductive and respiratory syndrome , but have not provided detailed information regarding how the disease affected farm performance. Here, we used longitudinal data routinely collected from sow farms from a US firm between 2014 and 2015 to explore the intensity and extension of outbreaks of PRRS. We then evaluated the effects on revenue due to a decrease on output production using the pre-, during, and post-outbreak periods. This approach allows us to evaluate if outbreaks were reported on time, as well as the extension and length of the impact on production. Endemic animal diseases can affect farm profit by reducing output, increasing production costs, and reducing product price . For example, PRRS, which was first identified in the 1980s, has become one of the most important endemic animal diseases in the US . It affects the swine industry and disease control is difficult due to factors inherent to the disease and the nature of the swine production system. The causal agent is an RNA-virus from the Arteriviridae family that is highly mutagenic and resistant to the low temperatures registered in Midwestern areas of the US, where a significant proportion of the US swine industry is located. On the other hand, the disease is highly transmissible and can persist for long periods in chronically infected animals and in the environment, if contaminated through secretions and excretes . Because PRSS has no effect on humans and has little impact on international trade, PRRS is a non-reportable disease in the US. There are no official programs for its control, but producers in some regions have begun collaborative programs to exchange information on PRRS outbreaks in the hope that coordinated action might reduce disease effects .
PRRSV spreads between and within farms via airborne transmission, the introduction of infected animals and contaminated fomites, often associated with the failure of bio-security protocols . PRRS may increase abortion and mortality rates in pre- and post-weaning pigs, lead to reproductive failure in sows, and lower feed conversion in feeder pigs , thus affecting several stages of the swine production cycle. However, the severity and length of the impact at each production stage are still unclear. In high farm density areas, PRRSV eradication is not the main target. Indeed, farmers prefer to maintain homogeneous levels of immunity in breeding herds using vaccination or, although less common, exposing animals to live virus . Herd closure and rollover is one of the most common strategies to eradicate PRRSV from sow farms. It consists in stopping introducing new sows as replacements in addition to remove seropositive animals for at least 24 weeks . A study showed that production of PRRSVnegative pigs was reached 27 weeks after herd closure started, although an important variation between farms was observed . Alternatively, whole-herd depopulation and repopulation strategy is the most effective strategy described but in many cases is financially impracticable . Two studies have estimated the economic impact of PRRS using data from a set of farms and then extrapolating their results to the entire US swine industry . They reported total annual losses of ~$560 million and ~$664 million, respectively . Although the two studies estimated losses similar in magnitude, they significantly differed in the proportion attributed to losses on sow farms. While Holtkamp et al. estimated that 46% of total losses occurred on sow farms , Neumann et al. estimated that only 12% of total losses occurred on sow farms. The causes of the differences in their loss estimates are not explicitly explained, but may occur because of differences in the epidemiology of the disease at different times, the diversity of clinical outputs in infected animals and/or differences in types of farms. We observe that the effects of disease vary slightly across the farms in our study despite a common management approach, but the availability of data from multiple farms is likely to provide a better estimate of impact than would the use of data from only one farm.
We measure the effect of disease using data prior to the outbreak as the baseline and find that disease impact varies over time, with output declining rapidly initially following the outbreak and then recovering slowly and non-monotonically. We gain additional insights into the progression and recovery of disease by measuring changes in seven other performance indicators. Our methodology can be used to characterize disease impact at the farm and/or firm level, as it provides information on the timing of disease effects, the pathways through which PRRS affects production, and the total time needed for recovery. We anticipate that the results presented will help in the development of more accurate models for evaluating alternative PRRS prevention and control strategies in the US.We screened production records from a large, vertically integrated swine firm that includes farms in each stage of the swine production cycle, i.e., breeding and growing. All of the farms are located in the Midwestern region of the US. Numerous of the sow farms experienced PRRS outbreaks during 2014–2015. An outbreak was reported when animals showed PRRS-compatible clinical signs that were subsequently confirmed through PCR testing. We chose for analysis only sow farms that had not experienced a PRRS outbreak for at least one year prior to the outbreak studied in this analysis. In addition, we excluded from the analysis any farm that experienced cases of porcine epidemic diarrhea virus during the eight months before the PRRS outbreak to avoid confounding disease effects. Thus, all farms were classified as positive-stable without undergoing elimination at the time of this study. This firm also used a common disease management protocol for all its farms.f weaned pigs , which subsequently was used to estimate the decline in the value of output due to a PRRS outbreak. Likewise, we used weekly data for seven statistics, drying rack cannabis referred here as performance indicators, to more comprehensively assess how the disease affected weaned pig output. These indicators are: the number of live births per litter -or litter size- , the number of stillbirths per litter , the number of pre-weaned pigs dead , the number of sows dead , the number of sows aborting , the number of sows with repetition of service , and the number of sows farrowing .The estimated baseline is used to measure PRRS’ effect on production after the outbreak, i.e., between t and t + 35. We then used the same procedure to analyze, separately, the baseline values and post-outbreak effects for each performance indicator . Using longitudinal data allows us to reveal PRRS dynamics that might be difficult to identify if using cross-sectional data. In this case we evaluate the net effect of the outbreak on production within a selected set of farms. The use of fixed effects also permits us to manage the unobserved heterogeneity within farms whose omission could bias the estimated coefficients. We assume that time invariant effects are unique to each farm and are not correlated with effects on other farms. In addition, the expectation that individual farms have stable characteristics over time and the recognition our sample set has not been selected randomly led us to prefer a fixed effects rather than a random effects model. We used the Hausman test to determine whether the unobserved effects are distributed dependently of the regressors . We used Stata Statistical Software V13.1 to perform all these statistical computations and graphic designs .Dummy coefficients for weeks were statistically different than 0 indicating that the inclusion of Tt as fixed time effects in equation is appropriate. Although we found no significant trend in weaned pig production during the pre-outbreak period, holding the number of sows and season constant when using equation , we observed a consistent decrease in weaned pig production relative to the baseline in the week t − 1, immediately before the outbreak was reported in week t . The decreases ranged between 1 and 12%. As our regressions based on equation 1) showed no significant trend in weaned pig production even when week t − 1 was included, we did not remove week t − 1 from the baseline period.
Had we done so, the baseline would have been very slightly higher and the estimated damages from PRRS slightly greater, as discussed subsequently. We estimate that aggregate weaned pig production for the 16 farms decreased from the baseline production of 2,094 per week to 1,600 in week t + 5, when output was a full 23% lower than the baseline. Table 1. The results show that farm production decreased monotonically from t − 1 to t + 5, and then began to recover . Output recovered moderately from t + 5 until t + 11, at which point another significant decline in production occurred to t + 17 . Eight of the 16 farms then recovered monotonically to their baseline production levels by t + 33, but a slight drop occurred again in t + 34 and t + 35 with 15 farms producing lower than the baseline. In the aggregate, observed production approached the baseline value by the end of t + 35, when our sample ended. Estimated output appears slightly lower than the pre-outbreak level, but the difference is not statistically significant . Similar to the production of weaned pigs, the seven performance indicators did not fully recover to their pre-outbreak means. Week-to-week comparisons revealed changes in all performance indicators, with some variation in timing and intensity. For each performance indicator, the recovery of each farm fluctuated around a rising trend estimated for all farms, and again showed a non-monotonic recovery . As expected, some performance indicators presented a lag with respect to the trend observed in weaned pig production. A significant increase in the number of preweaned pigs dead was detected at t with an average expected rise of 79 deceased animals relative to the baseline, reaching a maximum increase at t + 1, with 143 expected extra losses . While litter size did not show a significant decline at t, the expected number of live births decreased by around 1 animal between t + 1 and t + 18, reaching a maximum decline at t + 2 and t + 3 and a new deterioration at t + 14. The number of stillbirths increased between t and t + 16, reaching a maximum at t + 12, with 2 stillbirths per litter . Although there was no immediate increase in the number of sows designated for repeated service the week of the outbreak report, by t + 6 the number of sows that were designated to repeat service increased from 11 sows in the estimated baseline to 31 sows. Likewise, the number of pigs farrowed declined after t + 1. The number of abortions significantly increased at week t− 1, doubling the number of sows that aborted prior to t − 1. The number of sows with abortions peaked in the week of the outbreak report at a level five times higher than the baseline level.