| | Abstract | Introduction | Materials and methods | Results and discussion | Conclusions
Abstract
Physiological maturity in small grain cereals is defined as the cessation of nutrient movement to the grain. It usually occurs when the grain is at a moisture content of 35%. Using temperature data from nearby weather stations, growing degree days (GDD= [(Tmin+Tmax)/2] were calculated during the maturation phase of barley, wheat and triticale. Grain samples were collected from several cultivars of each grain from about 50% down to 10% grain moisture. During the linear phase of dry-down, moisture content was determined on the basis of GDD. For barley, an overall relationship was found such that estimation of GDD from sowing to physiological maturity can be made from moisture content measurements at harvest: GDDphysmat=GDDharvest- (35-Moistureharvest)/0.05933. Introducing total growing season precipitation data into the equation made a slight improvement in the fit of the data, such that the average relationship became GDDphysmat=GDDharvest-[(35-Moistureharvest)-0.04293(PrecipitationTotal)]/0.07124. A program based on SASRTM has been developed to determine days to maturity based on temperature, precipitation, moisture content of the grain at harvest, and the dry-down equations.
Introduction
For over thirty years, grain moisture content at harvest has been used at Field Crop Development Centre to estimate maturity. Each year, maturity indicator rows for winter and spring cereal types have been grown out and harvested over the dry-down phase of maturity (from about 60% moisture down to 10% moisture).
The grain moisture contents of these indicator rows have been used to estimate the rates of dry down on a location-year-species basis, and these rates of dry-down have been used to calculate when all harvested material would have reached physiological maturity, estimated to occur at 35% grain moisture, for that type within a location-year.
Materials and Methods
The moisture data from the maturity indicators (Table 1) for five locations (Calmar, Lacombe, Olds, Stettler and Trochu) was taken from 1999 to 2003 to determine rates of dry-down on a Growing Degree Day (GDD) basis. Not all location years were available due to loss of sites due to hail, winterkill, etc.
GDD were determined as the sum of mean daily temperatures (C) over 0oC from LicorRTM weather stations located at the plot site. When daily mean temperatures were not available (usually due to equipment failure), mean temperatures were calculated from minimum and maximum daily temperature available for the nearest weather station from Environment Canada http://climate.weatheroffice.ec.gc.ca/climateData/canada_e.html
Total precipitation data from seeding to harvest was also measured by LicorRTM weather stations or obtained from Environment Canada.
Table 1. Maturity indicator cultivars
| Type | No. of samples | Cultivar |
| Barley | 596 | AC Lacombe, Condor, Falcon, Harrington, Kasota, Seebe |
| Spring triticale | 321 | AC Alta, AC Certa, Pronghorn, Sandro, Wapiti |
| Spring wheat | 212 | AC Barrie, AC Karma, AC Vista, Katepwa |
| Winter triticale | 110 | Bobcat, Pika, Wintri |
| Winter wheat | 100 | AC Readymade, CDC Clair, CDC Osprey, Norstar |
| Winter rye | 78 | AC Rifle, Musketeer, Prima |
Results and Discussion
Protocol
Moisture data were fit to GDD using PROC GLM of SASRTM on a location-year-cultivar basis.
Where the r-square value was equal to, or greater than 0.80, data was determined to be acceptable; otherwise the data was scrubbed so that the non-linear phases of the dry-down curve were deleted. If the fit was still unacceptable, these data were discarded.
A general relationship of moisture to GDD for each type was determined using 3 methods (Table 2):
1) based on acceptable data using all data for the type
2) by determining the linear relationship on a location-year-cultivar basis and determining the mean rate of drydown
3) the data were then also run against GDD and Precipitation (Total) over all acceptable data by type.
Table 2. Dry down rates on a GDD basis
a = the coefficient from Moisture=Intercept+a*GDD over all acceptable data
am = the mean of ‘a’ from Moisture=Intercept+a*GDD on a location-year-cultivar basis
ap = the coefficient from Moisture=Intercept+a*GDD+b*Precipitation
b = the GSP(growing season precipitation) coefficient
Equations to predict maturity
We used grain moisture content at harvest (MoistureHarvest) to predict how many GDDDiff had passed since the crop reached 35% moisture (MoisturePM) our indicator of physiological maturity).
MoistureHarvest=%H2O =100 * (WET-(DRY * ((100- dryH2O)/100))) /WET; dryH2O defaults to 10
MoisturePM=35%.
Method 1 or Method 2: Determining GDDDiff based on GDD only:
- GDDDiff=(MoisturePM-MoistureHarvest)/a
- Where GDDDiff is the GDD from physiological maturity to harvest;
- MoisturePM is the moisture at physiological maturity (35%);
- MoistureHarvest is the moisture at harvest; and
- a is the GDD dry-down coefficient (a using Method 1 or am using Method 2).
Method 3: Determining GDDDiff based on GDD and Growing Season Precipitation (GSP):
- GDDDiff=[(MoisturePM-MoistureHarvest)-(b*GSP)]/a
- Where GDDDiff is the GDD from physiological maturity to harvest;
- MoisturePM is the moisture at physiological maturity (35%);
- MoistureHarvest is the moisture at harvest;
- a is the GDD dry-down coefficient;
- b is the GSP coefficient; and
- GSP=growing season precipitation (sowing to harvest).
Using GDDDiff, we were now able to calculate the GDD at physiological maturity as:
GDDPhysmat=GDDHarvest-GDDDiff
The GDDPhysmat can then be used to determine the calendar day on which Physiological Maturity occurred and days to maturity from sowing can be determined.
Verification and validation of estimations of maturity using GDD and Precipitation
Maturities were determined by type using:
1) the dry-down terms based on all acceptable data (a);
2) the mean of all acceptable data (am);
3) all acceptable data corrected for total growing season precipitation (ap+b).
These maturities were correlated with the historical maturity data from our 1999 to 2003 yield trials.
As well as tight correlations, we wanted the range of the new maturities to be reflective of the historical data.
We manipulated the total growing season precipitation to last 30, 25, 20, and 10 days to see what effect this would have on the maturities determined by (ap+bn) where n=number of days.
Table 3. Correlations for barley (n=24,454)
*** = Significant at P<0.001
Z = Correlations for ap+bn were not as good as total precipitation.
For barley (Table 3), using either the mean dry-down coefficient and the dry down adjusted for total growing season precipitation gave excellent fits to the historical data with similar ranges and means.
Table 4. Correlations for spring triticale (n=4200+)
*** = Significant at P<0.001
For spring triticale (Table 4), using the mean dry-down coefficient gave an excellent fit to the historical data; while using only the last 25 days of growing season precipitation gave higher correlations, using total growing season precipitation was more reflective of the range and mean of the historical data.
Table 5. Correlations for spring wheat (n=1400+)
*** = Significant at P<0.001
For spring wheat (Table 5), using the mean dry-down coefficient gave an excellent fit to the historical data; while using only the last 25 days of growing season precipitation gave higher correlations, using total growing season precipitation was more reflective of the range and mean of the historical data.
Table 6. Correlations for winter triticale (n=445)
*** = Significant at P<0.001
Z = Days from January 1
For winter triticale (Table 6), using the overall dry-down coefficient gave a better fit to the data than the mean coefficient; using the last 25 days gave a better fit than using total growing season precipitation.
Table 7. Correlations for winter wheat (n=2597)
*** = Significant at P<0.001
Z = Days from January 1
For winter wheat (Table 7), while the correlations were significant, they were not as tight as for the other types; both overall and mean coefficients gave similar results but using the last 25 days of precipitation was better than using total growing season precipitation.
Table 8. Correlations for winter rye (n=121)
*** = Significant at P<0.001
Z = Days from January 1
For winter rye (Table 8), using the mean dry-down coefficient gave a better fit to the data than the overall coefficient; using the last 25 days gave a better fit than using total growing season precipitation.
Conclusions
Scrubbing the data gave us very robust coefficients for the linear phase of drydown.
In some types, adding seasonal precipitation gave a better fit than using just GDD whether determined by using overall data or by a mean; and in some cases using just a subset of the last 20 to 25 days of growing season precipitation gave a better fit, as well as having a range and mean more similar to the historical data.
Future: In 2004, maturities will again be determined using both the historical method and the new dry-down coefficient methods. We will select the coefficient method that best reflects maturities. In the winter cereals, we may need to continue using the historical method until the data set is more robust.
P.E. Juskiw, J.H. Helm, D.F. Salmon and J.M. Nyachiro
Field Crop Development Centre
Presented at the Canadian Society of Agronomy Meetings, The Science of Changing Climates Conference, University of Alberta, Edmonton, July 20-23, 2004 |
|