The impacts of anthropogenic climate change on the global agriculture industry are projected to become more severe through the end of the 21st century as a result of increased climatic variability (Newberry et al., 2016; Hatfield et al., 2014). In fact, it has been estimated that the warming global climate has already reduced global crop production by 1-5% per decade since the 1980s (Porter et al., 2014). In the United States, much of the annual fluctuation in agricultural production can be linked to seasonal variability in the local climate. This is particularly true in the southeastern United States, which accounts for nearly 17% of the total United States agricultural production and provides some of the most diverse groups of crops in the country (Asseng, 2013). However, the region is at an increased risk of potentially crop ruining natural hazards including tropical cyclones, drought, and wildfires. Furthermore, farmers and citizens in the region remain skeptical of climate change and are less likely to implement adaptation
strategies, furthering the regional economic risk.
Previously identified as a “warming hole,” the southeastern United States experienced a slight cooling trend in the second half of the twentieth century while temperatures over the rest of the country experienced significant warming (Ellenburg et al., 2016; Meehl et al., 2015). Much debate surrounds the possible attributable factors that have caused this anomalous regional phenomenon with explanations ranging from increased aerosols and clouds (Laseter et al., 2012; Weber et al., 2007) to internal variability in sea surface temperatures and atmospheric circulation patterns (Kumar & Wang, 2015; Kunkel et al., 2006). However, recent studies suggest a disappearance of this warming hole and consequently a regional warming trend since the 1970s with 2001-2010 being the warmest decade on record for the southeast (Meehl et al., 2015; Kunkel et al., 2013). Trends in precipitation for the southeast are more seasonal in nature, with summer trending more dry and autumn trending more wet (Knox et al., 2015). Evidence also suggests that the southeast has experienced an increase in the frequency of extreme precipitation days with events becoming more intense, albeit with shorter durations (Powell & Keim, 2014; Groisman & Knight, 2008). The variability of climatic trends in the southeast make a detailed analysis of crop yield response to extreme events particularly important as it allows those living in the region to better prepare and adapt for future unknowns.
Extreme events are particularly relevant to agricultural production in the southeastern United States as the crops grown in the region are highly heterogeneous with some of the greatest diversity of cultivated crops in the U.S. (Knox et al., 2013). However, some of the mainstays of the agricultural industry in the region including cotton, tobacco, and fruit are highly sensitive to extreme weather (Changnon & Changnon, 1999; Changnon, 1972). Tropical cyclones pose one of the greatest risks to agricultural productivity in the southeast as many of the specialty crops are harvested during the peak of hurricane season (August-October). As such, the timing and duration of hurricanes and other extreme weather events play a crucial role in determining just how much damage will occur within the agricultural industry. For example, Hurricane Florence made landfall in early September 2018, just weeks before the start of the traditional harvest season in North and South Carolina. Although some farmers began an early harvest as a result of the unprecedented warning time, the slow movement of Florence after landfall meant that fields across the southeast endured days of record rainfall and high winds. As a result, initial estimates from the state of North Carolina suggest that Florence caused at least $2.4 billion in agricultural losses with field crops such as sweet potatoes and soybeans taking the brunt of the damage (NCDPS, 2018). In comparison, the timing of Hurricane Hugo in late September 1989 meant that agricultural losses were minimized as a result of early harvesting (Janiskee, 1990).
Crops in the southeast are also highly sensitive to extreme heat, with many crops in the region already being grown near their thermal limits. One prominent example is soybeans, which has an ideal daytime temperature of around 29?C (85?F) and can experience pollen sterility and reduced seed sets during periods of heat stress. The impacts of heat can be intensified when coupled with drought conditions. Such was the case during the southeast drought of 1993, where the state of South Carolina lost 25% of the tobacco crop, 70% of the soybean crop, and 95% of the corn crop as a result of record warmth and extended periods with no precipitation (Lott, 1994). The impacts of heat stress can vary significantly across crops and is highly dependent on timing in regards to the development stage of the individual crop. For example, peanuts experience the greatest amount of heat stress during the development of floral buds and experience greater damage when temperatures are above 40?C (104?F) (Prasad et al., 1999). Although irrigation techniques may offset some of the impacts of heat stress and drought, the southeastern United States is more vulnerable to these impacts than the western U.S. and Midwest as the soils in the region have a lower water-holding capacity (McNider et al., 2011).
Despite significant scientific advancements in linking climate variability to agricultural production at national and state levels, it is critical to quantify the impacts of climate change at a finer spatial scale (Lobell & Asner, 2003). More importantly, analysis of the relationships between agriculture and climate change are dominantly focused on the major global crops of wheat, rice, maize, and soybeans (Zhao et al., 2017). As a result, there is a limited understanding of how the variability in regional and specialty crop production is linked to extreme climatic events. This is particularly important for areas of high heterogeneity in regional crop production, such as the southeastern United States, where crops have a variety of susceptibilities related to the local climate.
We used a linear regression approach to evaluate the relationship between crop yield, extreme heat occurrence, extreme precipitation occurrence, and drought in Georgia, and the Carolinas. Crop yield survey data, which is measured in units of weight harvested per acre planted, was downloaded from the USDA Quick Stats database for corn (grain), upland cotton, peanuts, soybeans, and sweet potatoes from 1981 through 2017. Counties were included in our anlysis if they had at least thirty years of crop yield data, determined seperately for each crop.