Quasi-biennial corn yield cycles in Iowa
Introduction
The quasi-biennial oscillation (QBO) of the equatorial stratospheric winds is among the most well-known quasi-biennial climate patterns. This phenomenon consists of easterly and westerly wind regimes in the tropical stratosphere with a mean period of 28 months (2.3 years; Baldwin et al., 2001). Surface air temperatures also exhibit a quasi-biennial component, which may be associated with the North Atlantic Oscillation (NAO; approximately 2.2 year period; Mann and Park, 1994). Spectral analysis has helped detect quasi-biennial signals in precipitation (Rajagopalan and Lall, 1998), surface air temperature (Mann and Park, 1994), sea ice cover (Gloersen, 1995), tree rings (Rao and Hamed, 2003), and indices of ENSO (approximately 2.5 year period; Ghil et al., 2002). Despite the evidence for significant quasi-biennial variability in climate, previous studies have argued against a similar timescale crop signal in the U.S. (Black and Thompson, 1978).
Fig. 1 shows six counties in Iowa that cover nearly 9000 km2 within the surface geologic region known as the Des Moines Lobe. This is an important corn producing region because it is among the highest corn producing and yielding regions in Iowa (Fig. 1), Iowa is generally the leading corn yielding and producing state in the U.S. (USDA-NASS, 2008), and the U.S. has produced approximately 40% of the world's corn since 2000 (FAOSTAT, 2008). Also, these six counties have the principal soil association of Clarion–Nicolette–Webster (ISU, 2004), which require artificial drainage for corn production. Artificial drainage coupled with a high corn yielding environment contribute to streams within the Des Moines Lobe to be among the greatest sources of nitrogen loading to the Mississippi River Basin (Goolsby et al., 2001), which has been implicated as a cause of hypoxia in the northern Gulf of Mexico. In general, anthropogenic perturbation of the global nitrogen cycle is of increasing concern (Gruber and Galloway, 2008), and food production is the major contributor (Galloway et al., 2003). Corn yield variability could affect nitrate flux because small changes in corn yield may have greater effects on N loss in artificially drained soil (Malone and Ma, 2009).
Long-term U.S. corn yield variability is often associated with weather variability such as temperature and rainfall (e.g., Lobell and Asner, 2003, Hu and Buyanovsky, 2003). Large-scale climate signals such as ENSO have been linked to corn yield variability in the U.S. corn-belt because of its association with growing season temperature and precipitation variability (e.g., Phillips et al., 1999, Carlson et al., 1996). Combinations of climate signals (ENSO and the NAO) have been found to affect agro-pastoral production in Africa (Stige et al., 2006). For example, in southern Africa strong associations were found between year-to-year variability of ENSO and corn yield. Also year-to-year NAO variability was associated with slaughter weights of goats in western Africa and rice yield in northern and central Africa. However, the combined effects of several climate indices on U.S. corn yield variability remain fairly unexplored. Also, the effects of annual variation in ground level solar radiation during the growing season on long-term corn yield in the U.S. remain fairly unexplored.
Here we analyze long-term corn yield from the Des Moines Lobe region of Iowa with daily temperature, rainfall, solar radiation, and monthly indices of NAO, SOI, and QBO. The Southern Oscillation Index (SOI) provides a quantitative measure of the ENSO cycle and the SOI correlates with future rainfall in some regions (Stone et al., 1996). This analysis should help answer several questions: does long-term corn yield variability in the U.S. contain a significant quasi-biennial component, what weather drives this phenomenon, is it related to large-scale climate variability, and what are the quantitative effects?
Section snippets
Data
Table 1 summarizes the data used in this analysis, which includes county-level corn yield from Iowa, climate indices (SOI, QBO, and NAO), daily temperature, daily precipitation, and daily solar radiation. The QBO was briefly described above; the NAO and SOI consist of monthly records of fluctuation in north-south North Atlantic atmospheric pressure gradient (NAO) and the surface air pressure difference between Tahiti and Darwin, Australia (SOI). In Iowa, corn is generally planted in April or
Spectral analysis of detrended observed corn yield
Fig. 2c shows the average annual corn yield fraction from 1952 to 2006 of the six Iowa counties (Fig. 1) with the long-term linear and polynomial trends removed (Fig. 2a and b). Fig. 3a shows the spectrum of Fig. 2c using the multitaper method. The largest spectral magnitude is centered on a frequency of about 0.4 cycles per year (2.5 years). The F-statistic in proximity to this frequency suggests two significant periods of 2.3 and 2.6 years (frequencies of 0.43 and 0.38 cycles per year; p <
Conclusions
Our results suggest the existence of a quasi-biennial pattern in long-term Iowa corn yields related to large-scale climate variability organized on this timescale. This conclusion should be treated as an impetus for further research. However, we have demonstrated that a statistical model based on underlying climate variables yields skillful predictions of interannual variation in Iowa corn yields. Given the importance of this cereal crop, refined versions of this model might prove to be of
Acknowledgements
We thank R. Carlson, J. Cornette, and P. Heilman for critical discussions. We are also grateful to the reviewers for their helpful comments.
References (48)
- et al.
An integrated North Florida dairy farm model to reduce environmental impacts under seasonal climate variability
Agric. Ecosyst. Environ.
(2006) - et al.
Potential benefits of climate forecasting to agriculture
Agric. Ecosyst. Environ.
(2000) - et al.
Multiple-taper spectral-analysis – a stand-alone C-subroutine
Comput. Geosci.
(1995) - et al.
Empirical analysis and prediction of nitrate loading and crop yield for corn-soybean rotations
Geoderma
(2007) Estimation of maximum possible daily global solar radiation
Agric. Forest Meteorol.
(1997)- et al.
Interannual variability in western US precipitation
J. Hydrol.
(1998) - et al.
Response of three eastern tree species to supplemental UV-B radiation: leaf chemistry and gas exchange
Agric. Forest Meteorol.
(2003) - AutoSignal User's Manual. 2003. SeaSolve Software Inc. Framingham,...
- et al.
The quasi-biennial oscillation
Rev. Geophys.
(2001) - et al.
Some evidence on weather-crop-yield interaction
Am. J. Agric. Econ.
(1978)