Portal de Periódicos da CAPES

Impacts of Climate Change on Milk Production in the United States

2014; Routledge; Volume: 67; Issue: 1 Linguagem: Inglês

10.1080/00330124.2014.921017

1467-9272

Guillaume Mauger, Yoram Bauman, T.D. Nennich, Eric P. Salathé,

Climate Change and Health Impacts

AbstractClimate change is likely to affect milk production because of the sensitivity of dairy cows to excessive temperature and humidity. We use downscaled climate data and county-level dairy industry data to estimate milk production losses for Holstein dairy cows in the conterminous United States. On a national level, we estimate present-day production losses of 1.9 percent relative to baseline production and project that climate impacts could increase these losses to 6.3 percent by the end of the twenty-first century. Using present-day prices, this corresponds to annual losses of $670 million per year today, rising to $2.2 billion per year by the end of the century. We also find that there is significant geographic variation in production losses and that regions currently experiencing the greatest heat-related impacts are also projected to experience the greatest additional losses with climate change. Specifically, statewide average estimates of end-of-century losses range from 0.4 percent in Washington to a 25 percent loss in annual milk production in Florida. Given that the majority of these losses occur in the summer months, this has the potential to significantly impact operations in hotter climates.乳牛对于极端气温和溼度具有敏感性, 因此气候变迁有可能会影响牛乳的生产。我们运用降尺度的气候数据和郡县层级的酪农业数据, 评估美国大陆荷斯坦牛的牛乳生产损失。我们估计, 目前全国层级相较于基本产量, 有百分之一点九的生产损失, 并推测气候冲击可能在二十一世纪末, 使该损失增加至百分之六点三。若运用目前的价格计算之, 此将等同于当下有每年六亿七千万美元的损失, 该损失并将于本世纪末增加至每年二十二亿美元。我们同时发现, 生产损失具有显着的地理差异, 而目前正受到热浪冲击最为严重之区域, 预计将同时成为在气候变迁之下承受最大额外损失之地。尤其是在州层级方面, 本世纪末的各州平均年度牛乳生产损失, 将从华盛顿州的百分之零点四, 到佛罗里达州的百分之二十五不等。有鉴于这些损失多半发生在夏季的月份, 上述情形将有可能显着影响较炎热气候中的酪农业经营。Es probable que el cambio climático afecte la producción de leche debido a la sensibilidad de las vacas lecheras a la temperatura y humedad excesivas. Utilizamos datos micro-climáticos e información de la industria lechera a escala de condado para calcular las pérdidas en la producción lechera con vacas de raza Holstein en los Estados Unidos contiguos. A nivel nacional, las pérdidas de producción las calculamos para la actualidad en el 1.9 por ciento relativo a la línea de base de la producción y proyectamos que los impactos del clima podrían incrementar estas pérdidas hasta el 6.3 por ciento a finales del siglo XXI. Con base en precios actuales, esto representa pérdidas anuales de $670 millones de dólares, lo cual alcanzaría los $2.2 mil millones de dólares por año al terminar el siglo. Encontramos también que hay una variación geográfica significativa en las pérdidas de producción y que las regiones que actualmente experimentan los impactos más altos relacionados con calor se proyectan también como las que experimentarían la más altas pérdidas adicionales con el cambio climático. Específicamente, los cálculos de pérdidas para los diferentes estados a finales del siglo van desde el 0.4 por ciento para Washington a un 25 por ciento de pérdida anual en producción lechera para Florida. Si se considera que la mayoría de estas pérdidas ocurren en los meses de verano, esto tiene el potencial de impactar significativamente las operaciones en los climas más cálidos.Key Words:: climate changeclimate impactsdairy productioneconomic impactsheat stress.关键词:: 气候变迁气候冲击酪农生产经济冲击热压力。Palabras clave:: cambio climáticoimpactos del climaproducción lecheraimpactos económicosestrés por calor Notes* The authors acknowledge the skilled assistance of Robert Norheim in providing geographic information systems analysis and maps used in this article, of Hye Jin (Jenny) Chang in gathering dairy industry data, and of the anonymous reviewers for their thoughtful comments on the article. This research was supported in part by the Climate Impacts Group and the Joint Institute for the Study of the Atmosphere and Ocean under NOAA Cooperative Agreement No. NA17RJ1232.1In principle one could recalibrate Equation 2 so that it can be applied using monthly data. This would necessarily reduce the accuracy of LOSS estimates and could result in errors that vary across different geographic regions. Furthermore, use of monthly average data would implicitly assume that the range of daily temperatures about the monthly mean remains the same going into the future. This is unlikely to be the case—for example, decreases in soil moisture associated with greater evaporative stress are expected to result in an increase in daily temperature variability (see, e.g., Seneviratne et al. Citation2010). Although our calculations (not shown) suggest that it might be possible to recalibrate the LOSS equation, further investigation would be needed to confirm that these concerns are not realized.2Gridding observed data entails several approximations (e.g., constant lapse rate, interpolation to data-sparse regions). Such assumptions can result in errors that vary with season and geography. Maurer et al. (Citation2002) used these data to drive a hydrologic model and showed that the results match observations quite well over large basins. Validation over large river basins does not address the question of reliability over small areas, however, and it is likely that representation of heat stress is sensitive to different errors than stream flow. Nonetheless, this data set includes the best daily-resolved continental-scale climate data that are available at sufficiently high resolution to be applied to dairy farming.3Gridded humidity data are not available at daily time steps: The use of monthly data is necessitated by the scarcity of surface humidity observations, which are much less common than observations of temperature.4The reported weekly averages of observed daily minimum and maximum temperature and humidity were used to compute LOSS. This introduces a possible inconsistency in that weekly average milk loss is not necessarily given by weekly average THI due to the nonlinearity of Equation 2. To test the validity of weekly values for this calculation, we used a Monte Carlo method and the reported standard deviations of the observed quantities to derive daily values and then computed the corresponding weekly average production loss. We found that the results based on weekly averages are in reasonable agreement with the Monte Carlo results, which follows from the near linearity of LOSS with temperature for the hot, dry conditions reported in the Ahmed and El Amin (Citation1997) study. As a result, only the former are shown for clarity.5Milk loss per cow is a weighted average of county-level data based on dairy population; state totals reflect all cows in each state but county-specific populations were not available for all counties, so “unattributed” cows—a percentage ranging from 1 percent in California, Wisconsin, and New York, to 28 percent in Florida—were allocated to the “unattributed” county with the smallest amount of milk loss per cow. This makes our estimates more conservative but does not affect our conclusions.Additional informationNotes on contributorsGuillaume MaugerGUILLAUME MAUGER is a research scientist with the Climate Impacts Group at the University of Washington, Seattle, WA 98112. E-mail: gmauger@uw.edu. His research centers on climate impacts assessment, including efforts to evaluate climate change impacts on hydrology and ecosystems.Yoram BaumanYORAM BAUMAN is an environmental economist previously affiliated with the University of Washington, Seattle, WA 98112. E-mail: yoram@standupeconomist.com. He now makes a living doing stand-up comedy about economics.Tamilee NennichTAMILEE NENNICH is an Associate Professor in the Department of Animal Sciences at Purdue University, West Lafayette, IN 47907. E-mail: tnennich@purdue.edu. Her research interests focus on management strategies to increase the utilization of nutrients on dairy farms while improving animal performance and decreasing nutrient excretion.Eric SalathéERIC SALATHÉ is an Associate Professor in the School of Science Technology Engineering and Mathematics, University of Washington, Bothell, WA 98011. E-mail: salathe@uw.edu. He conducts research on regional climate change and climate change impacts.