|
|
||||||||
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
Center for Sustainability and the Global Environment (SAGE), The Nelson Institute for Environmental Studies, Univ. of Wisconsin-Madison, 1710 University Ave., Madison, WI 53726
* Corresponding author (Kucharik{at}wisc.edu)
Received for publication May 18, 2006.
| ABSTRACT |
|---|
|
|
|---|
Abbreviations: CTIC, Conservation Technology Information Center DOY, day of year NCEP, National Centers for Environmental Prediction
| INTRODUCTION |
|---|
|
|
|---|
Previous research has noted that a variety of technological factors including the development of hybrids with a higher tolerance to suboptimal growing conditions (e.g., cold, wet soils), increased resistance to diseases and pests, and improved equipment (planter) functioning have led to a large-scale shift toward much earlier corn planting today than several decades ago (Bruns and Abbas, 2006; Lauer, 2001). For example, Bruns and Abbas (2006) cited that in the Mid-South USA, the 50% corn planted date has shifted earlier by about a month (from early May to early April) over the past 30 yr. Other studies by Duvick (1989) and Lauer (2001) also state that significant shifts in corn planting, to a much earlier date, have occurred across the Corn Belt region. However, there has not been a comprehensive data analysis that documents such trends over the past few decades.
Because agroecosystems comprise a significant portion of the total land area across the Corn Belt, a dramatic shift in the timing of crop planting similar to that noted in the Mid-South could significantly affect the seasonality of C and water exchange, and will influence the large-scale phenology and surface albedo of the region (Twine et al., 2004). Thus, documentation of land management trends and vegetative response are particularly important to C cycle science, vegetation-climate feedbacks, and the interpretation of drivers of change in remote sensing observations of the land surface, particularly during the last 30 or so years. Furthermore, any trend toward an earlier initiation and completion of planting could have contributed to a portion of the yield increases that have occurred since the 1950s, and therefore analysis of these data are relevant to studies of future yield increases and the underlying drivers of such change.
The objectives of this research were to: fully examine and document large-scale patterns and changes in corn planting dates across the central USA for the previous 30 yr, and investigate whether any changes were induced by large-scale climate trends. In this study, we present and analyze state level U.S. Department of Agriculture (USDA) weekly crop planting progress data (defined as the percentage of corn acreage planting that has been completed at any point in time) from 1979 to 2005 for 12 states that currently occupy approximately 82% of the total U.S. corn harvested area. The National Centers for Environmental Prediction gridded reanalysis climate data of daily minimum and maximum temperature and precipitation rate were also analyzed to determine whether trends in springtime weather conditions over the 27-yr period were potential drivers of any observed trends in the USDA data.
| MATERIALS AND METHODS |
|---|
|
|
|---|
In this study, all available weekly data, grouped by state, were input into the SAS-JMP statistical discovery software package (SAS Institute, 2002) and a neural net algorithm was used to fit a S-shaped numerical function through each growing season's weekly data for each year and each state independently (refer to Fig. 1 as an example). This statistical curve fitting was performed assuming a temporal resolution of 1 d so that a continuous daily record of planting progress could be constructed and therefore easily compared across all years because the survey reporting dates varied by year. The date of occurrence for planting progress at four thresholds of completion (10, 25, 50, and 75%) were subjected to linear regression analysis (using SAS-JMP) to detect any statistical trends (and level of significance or P value) for each individual state over the 27-yr study period. Regional (12 state) average dates of occurrence for each planting threshold were formed by weighting each state's calendar date (day of year) by the fractional planted corn acreage relative to the total acreage for all 12 states. These weighted regional average values for the 27-yr period were also subjected to linear regression analysis.
|
| RESULTS AND DISCUSSION |
|---|
|
|
|---|
|
|
|
13 d earlier now compared to the early 1980s (Table 3, Fig. 3
). This result corroborates a previous study that suggested corn-planting dates across the entire Corn Belt became earlier by approximately 4 wk between 1930 and the 1980s (Duvick, 1989), and is in agreement with trends previously shown for Wisconsin corn planting from 1976 through 1999 (Lauer, 2001).
|
|
0.3 to 0.8 d yr1) of state level trends for 25, 50, and 75% planted date thresholds (Table 1, Fig. 2B) were similar to the computed linear trend for the 10% threshold, however, the number of states that showed statistically significant trends of earlier planting (P < 0.05) generally decreased as the value of the planted thresholds analyzed increased. For example, 10 of 12 states had significant (P < 0.05) trends for the 25% planted threshold, but only 8 and 4 states yielded significant results for the 50 and 75% thresholds, respectively (Table 1). However, very large changes were apparent even for the 75% planted date in some states, as farmer management changes in Kentucky, Kansas, Minnesota, and Nebraska have caused this threshold to be reached 12 to 18 d earlier during the course of the last 27 yr (Tables 1 and 2; Fig. 2B). In the early 1980s, 75% of corn planting was completed in a timeframe between 18 May and 28 May; currently, this is now occurring between 5 May and 21 May (Table 2, Fig. 2B). The varied results (e.g., magnitude of the trends) produced as a function of the planted thresholds analyzed suggests that while farmers have worked toward an earlier start to corn planting, they did not always complete their planting by an equal number of days. In fact, when trends for the total number of days elapsed between the 10% planted and 75% planted thresholds were analyzed over the 27 yr, the state data that produced statistically significant trends (P < 0.1; Iowa, Indiana, and Kentucky) suggested the duration of time needed to progress from 10 to 75% planting completed was taking 0.2 to 0.3 d yr1 longer.
The number of days that elapsed between the 10 and 75% completion of planting stages varied from an average of 15.2 (±5.6) d in Minnesota to a maximum average of 31.6 (±12.2) d in Michigan. This range of planting duration is likely driven by a variable number of workable field days in each state during spring (influenced by weather), farm size and equipment, and the typical length of time in spring that farmers have to get full-season hybrids in the ground before yield losses are likely to occur due to delayed planting.
The regional, weighted average planting date trends were 0.45, 0.39, and 0.37 d yr1 for the 25, 50, and 75% planted thresholds, respectively, and all were significant (P < 0.05, Table 3). Thus, the overall regional trend is for planting to be initiated and progressing toward completion faster today than several decades ago, but the most significant trends are on the initiation end of the planting progress continuum (Table 3, Fig. 3).
We investigated the idea that regional climate change, specifically warmer and drier springtime weather conditions, were a driving factor to earlier planting. Lauer (2001) questioned whether earlier corn planting dates in Wisconsin were a testament of real technological progress in farming, or were being driven by global warming. It would be easy to presume that a push toward earlier planting has been induced by climate change as evidence suggests the onset of spring is starting earlier across much of North America and is generally linked to warming in the Northern Hemisphere over the last half of the 20th century (Schwartz et al., 2006).
There does not appear to be strong evidence for widespread springtime warming from 1979 to 2005 to support the idea that climate change has been the most important factor driving the earlier planting trends encompassing the majority of the Corn Belt. An analysis of the NCEP daily mean temperature reanalysis suggests that linear trends of the date that 15-d running mean temperatures reached 10°C and 12.78°C (an assumption and surrogate used here for approximate optimal planting times, and the coolest temperatures that seed germination would occur) (Dwyer et al., 2000) were not statistically significant (P < 0.1) across the large majority of the study region (Fig. 4A , 4B).
|
However, while these air temperature indices are useful for a general interpretation of whether climate may have influenced planting behavior, a much more difficult quantity to assess (and more relevant to driving trends in springtime planting) are how the number of "workable field days" has changed. Central U.S. farmers typically plant over a period of a few weeks to a month during stints where weather conditions are optimal to perform fieldwork, which may only amount to 2 to 4 d each week during April and May (Nafziger, 2002). The sequence of weather patterns and the level of soil "trafficability" will partially determine whether a farmer is able to initiate getting seed into the ground.
Following on this idea, once a grower initiates corn planting, weather and soil conditions among other factors such as completion of other field activities, equipment breakdowns, and operational efficiency, begin to all play an integrated role in how quickly all planting is completed. The choice of the first planting "wave" appears to be getting earlier by about 8 to 22 d (e.g., the 10% planted date) at the state level, meaning growers are choosing to head to the fields earlier to initiate planting. However, growers aren't always nearing completion of their planting (e.g., the 75% planted threshold) by a similar number of days earlier, and this may be indicative of the following: (i) the number of workable field days during each week changes after the start of planting due to weather pattern changes; (ii) growers aren't pushing as hard to finish all of their planting because they achieved an early start and perceive themselves to be slightly ahead of schedule; or (iii) differences in farm operations and average farm sizes lead to varied planting behavior (after initiation) across the Corn Belt. In reference to the last point above, highly sophisticated, large farm operations that are responsible for hundreds to more than a thousand acres might be more likely to invest in cutting-edge technology so that they have the best chance to complete all planting before yield potential declines. In contrast, small operators who farm less than 81 ha (200 acres) and are doing so as a part-time activity, often plant when it is most convenient and might not have the ability to respond quickly to brief planting windows when they occur.
If climate change doesn't point strongly to supporting earlier planting since 1979 over the entire region, what is the more likely reasoning for the unified, advancing planting dates? Other data suggests that farmers are becoming much more efficient at field operations, and a growing number are gradually adopting management practices that can facilitate earlier planting (Lithourgidis et al., 2005). Perhaps the most significant changes are a trend toward performing tillage immediately after fall harvests, and the adoption of reduced-tillage or no-tillage management practices over the past three decades. Data from the Conservation Technology Information Center (CTIC) showed that the percentage of U.S. cropland managed with conservation tillage practices increased from 26.1% in 1990 to 40.7% in 2004, while the fraction of land acreage that was subjected to conventional tillage practices declined from 48.7% in 1990 to 37.7% in 2004 (CTIC, 2004). The greatest change was through the adoption of no-till tillage, which gained 18.4 million ha (45.5 million acres) from 1990 to 2004, mostly at the expense of conventional tillage, which decreased by 13.1 million ha (32.3 million acres).
Reduced tillage practices lead to a significant reduction in time and resources (i.e., labor and fuel) needed to prepare soils in spring before planting can take place (Arvidsson et al., 2000; Lithourgidis et al., 2005). The use of more advanced equipment that is more efficient in operation that can complete several tasks in one pass may also be a contributor to planting earlier in springtime (Popp et al., 2000). However, reduced tillage practices and the increasing adoption of leaving corn stubble on the soil surface reduces the rate at which soils will warm during the early spring compared to conventional management practices (Hayhoe et al., 1996).
To help offset some of these factors that can lead to a slower warming of soils during spring and to continue promotion of earlier planting, advances in technology and plant breeding have also enhanced the heartiness of plants and their ability to deal with less than ideal growing conditions that persist in early spring (Bruns and Abbas, 2006; Lauer, 2001; Lee et al., 2002). New hybrids are becoming better suited to fight off disease and pests that may attack the plant during the early growth stages and are gradually being developed to handle suboptimal temperatures (nonfreezing) that plants may encounter after early planting, particularly in northern Corn Belt locations (Dwyer et al., 2005; Lee et al., 2002). The continued advancement of fungicide and insecticide chemistries used in seed development and production also have added to early season plant heartiness so that seedlings can withstand disease, increasing the likelihood of survival (Lauer, 2005). Finally, the new development of seeds with a temperature-activated polymer coating may continue to allow for earlier planting (of a few weeks) in cooler and wetter soils (Gesch and Archer, 2005).
| CONCLUSIONS |
|---|
|
|
|---|
One potential causal affect of earlier planting may be a contribution to the increase in corn yields that have occurred since the 1950s. Regardless of the exact causes that have supported earlier corn planting, this trend could have supported a switch to hybrids suited for a longer growing season, which would encourage yield gains (Gupta, 1985; Lauer et al., 1999; Nielsen et al., 2002; Swanson and Wilhelm, 1996), or it would lengthen the time window that growers had to complete their field activities before significant delays would instigate switches to shorter season hybrids with lower yield potential. Earlier planting reduces the risk of fall frost damage to corn and decreases the likelihood that the crop will suffer injury from late insect and disease pest problems such as European Corn Borer (Ostrinia nubilalis) and gray leaf spot (Cercospora zeae-maydis; Ohio State University Extension, 2005). Furthermore, if the first crop planted were to be damaged, in many cases there would still be enough time to replant (Nafziger, 2002). However, it is difficult to separate the relative contributions of specific agronomic practices and other breeding changes to yield trends because of their interrelated nature, but it is likely that these planting date trends have contributed partially to the increasing yields over the past several decades (Cardwell, 1982). Finally, earlier planting contributes to lower grain moisture leading to reduced drying time and energy costs, which is now a major concern of growers due to rising fuel prices (Ohio State University Extension, 2005).
Farmers may eventually reach a point where it will become more difficult to increase operational efficiency and plant earlier, especially for those that are already adopting time saving strategies. However, continued development of new corn genotypes and the planting of polymer coated corn seed may continue to allow for earlier planting to take place in the future. However, it is unrealistic to expect that a trend toward earlier planting will continue indefinitely because planting will ultimately be limited by freezing temperatures, snow cover, and frozen soils in many locations.
| ACKNOWLEDGMENTS |
|---|
| REFERENCES |
|---|
|
|
|---|
This article has been cited by other articles:
![]() |
H. A. Bruns A Survey of Factors Involved in Crop Maturity Agron. J., January 8, 2009; 101(1): 60 - 66. [Abstract] [Full Text] [PDF] |
||||
![]() |
D. B. Egli Comparison of Corn and Soybean Yields in the United States: Historical Trends and Future Prospects Agron. J., May 7, 2008; 100(Supplement_3): S-79 - S-88. [Abstract] [Full Text] [PDF] |
||||
![]() |
C. J. Kucharik Contribution of Planting Date Trends to Increased Maize Yields in the Central United States Agron. J., February 26, 2008; 100(2): 328 - 336. [Abstract] [Full Text] [PDF] |
||||
![]() |
E. A. Lee and M. Tollenaar Physiological Basis of Successful Breeding Strategies for Maize Grain Yield Crop Sci., December 18, 2007; 47(Supplement_3): S-202 - S-215. [Abstract] [Full Text] [PDF] |
||||
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |