Agronomy Journal Grow Your Career With ASA
HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS
QUICK SEARCH:   [advanced]


     

This Article
Right arrow Abstract Freely available
Right arrow Figures Only
Right arrow Full Text (PDF) Free
Right arrow Alert me when this article is cited
Right arrow Alert me if a correction is posted
Services
Right arrow Similar articles in this journal
Right arrow Similar articles in ISI Web of Science
Right arrow Alert me to new issues of the journal
Right arrow Download to citation manager
Right arrow reprints & permissions
Citing Articles
Right arrow Citing Articles via HighWire
Right arrow Citing Articles via ISI Web of Science (15)
Right arrow Citing Articles via Google Scholar
Google Scholar
Right arrow Articles by Calderini, D. F.
Right arrow Articles by Slafer, G. A.
Right arrow Search for Related Content
PubMed
Right arrow Articles by Calderini, D. F.
Right arrow Articles by Slafer, G. A.
Agricola
Right arrow Articles by Calderini, D. F.
Right arrow Articles by Slafer, G. A.
Related Collections
Right arrow Statistics
Right arrow Crop Physiology & Metabolism
Right arrow Wheat
Right arrow Crop Models
Agronomy Journal 92:895-901 (2000)
© 2000 American Society of Agronomy

MODELING

Physiological Maturity in Wheat Based on Kernel Water and Dry Matter

Daniel F. Calderini, Leonor G. Abeledo and Gustavo A. Slafer

Dep. de Producción Vegetal, Facultad de Agronomía, Univ. de Buenos Aires, Av. San Martín 4453, 1417 Buenos Aires, Argentina

dfcalder{at}agro.uba.ar


   ABSTRACT
TOP
 ABSTRACT
INTRODUCTION
 Materials and methods
 Results
 Discussion
 REFERENCES
 
Estimation of the time of physiological maturity could be beneficial to avoid yield penalties due to lodging, sprouting, hail, and other harvest risks. The aim of our study was to evaluate a simple empirical relationship (model) to determine physiological maturity by simultaneously analyzing the water and dry matter dynamics of wheat kernels. An experiment was conducted in which two cultivars with different kernel mass potential were sown on four dates. Fresh and dry kernel mass from different positions in the spike were measured twice weekly. To validate the regression model, measured and calculated data from different cultivars, growing seasons, and kernel positions were compared. A negative linear relationship between kernel dry matter (relative to final kernel mass) and kernel water concentration was determined. This showed that in wheat, physiological maturity is reached at 37% of kernel water concentration. Validation of the regression model was done using data from field experiments in Argentina and Mexico, and from controlled-conditions experiments reported in the literature. The regression model successfully simulated results from field experiments . In addition, data from controlled-conditions experiments showed the same negative linear relationship between relative kernel dry matter and kernel water concentration , and the model achieved a good fit for measured data . This regression model is proposed for use by farmers and crop managers, who can simply measure grain humidity with grain moisture meters.

Abbreviations: Kn, kernel position n • KWC, kernel water concentration • RKDM, relative kernel dry matter • Sn, sowing date n • Vn, validation experiment n


   INTRODUCTION
TOP
 ABSTRACT
 INTRODUCTION
Materials and methods
 Results
 Discussion
 REFERENCES
 
KNOWLEDGE OF THE TIME of physiological maturity could be critical under some circumstances. At physiological maturity, the crop has reached maximum possible grain yield, and kernels, which are no longer growing, merely lose water; from there on, the crop is subject to an increasing risk of yield penalties due to damage from different sources (e.g., lodging, preharvest sprouting, hail, and biological stresses). When this risk is high and/or when there are certain expectations that kernel water cannot be reduced at a high rate simply by leaving the crop in the field, determination of physiological maturity is relevant. In addition, there are many farming areas in which the practice of double-cropping is possible. For example, in the Pampa region of Argentina, soybean is sown immediately following the wheat harvest (Hall et al., 1992). In this case, soybean is sown two months later than the recommended date. Therefore, the earlier the wheat harvest, the sooner the soybean can be sown (Andrade, 1995). Under such circumstances, an accurate estimation of physiological maturity would allow farmers to anticipate the harvest (using desiccants or windrowing) without reducing wheat yield, thereby providing better returns in double-crop systems.

The most precise method of determining the time of physiological maturity in crops is to follow kernel dry matter accumulation after anthesis. By weighing oven-dried samples after physiological maturity, a regression model can be determined that calculates when that stage was actually reached. Following this procedure, we can normally tell when physiological maturity is reached no earlier than 2 wk after it actually occurred, which does not allow us to decide on the earliest possible harvest time. In addition, this methodology requires weighing equipment that is not commonly available on most farms.

Development of alternative, simple methods to determine with a single measurement if a crop is at physiological maturity would be useful. For example, the use of the milk line and black layer are an indirect but frequently used indicator of physiological maturity in maize (Muchow, 1990). In wheat (as in other crops), there are no simple early visual morphological signs strongly correlated, for a wide range of environmental conditions, with the cessation of kernel dry matter accumulation. The kernel water concentration (KWC; i.e., percentage of water in kernels) is a potentially useful trait. Farmers and crop managers can simply estimate it in the field with such relatively inexpensive equipment as grain moisture meters (0.5% accuracy), which are frequently owned by farmers. This suggestion is borne out by different studies where clear negative associations between dry matter and water concentration have been shown during kernel filling in wheat (e.g., Sofield et al., 1977; Millet and Pinthus, 1984; Tashiro and Wardlaw, 1990), maize (Cheikh and Jones, 1994), soybean (Egli, 1994), and pea (Ney et al., 1993). Although water and dry matter dynamics are associated (see Sofield et al., 1977; Schnyder and Baum, 1992), different values of KWC have been reported at physiological maturity. For example, while Dodds et al. (1979) concluded that the practice of windrowing in wheat could be conducted at 30 to 35% of KWC without serious loss of yield in wheat, Schnyder and Baum (1992) obtained a value of 46% at physiological maturity. However, Schnyder and Baum (1992) included only basal kernels of central spikelets within the spike rather than all kernels. Recently, Egli and TeKrony (1997) estimated a KWC of 43% at physiological maturity, measuring kernels from central spikelets of the spike. In addition, cultivar differences in KWC at physiological maturity (between 13 and 28%) have also been reported (Hanft and Wych, 1982).

The purpose of this study was to develop a statistical model to determine the time of physiological maturity so that growers can decide the earliest possible time to harvest wheat without yield penalty from lack of complete kernel filling (e.g., spraying desiccants).


   Materials and methods
TOP
 ABSTRACT
 INTRODUCTION
 Materials and methods
Results
 Discussion
 REFERENCES
 
Site and Treatments
A field study was conducted during 1995 and 1996 at the experimental field of the Faculty of Agronomy, University of Buenos Aires, Buenos Aires City, Argentina (34°35' S, 58°29' W, elevation 25 m above sea level) on a silty clay loam soil (Aeric Argiudoll) with an organic matter content of 2.4%, a pH of 5.2, and a C:N ratio of 9.3.

The study involved the factorial combination of two high-yielding wheat cultivars (ProINTA Federal and Buck Ombú, developed by Instituto Nacional de Tecnología Agropecuaria, Marcos Juarez, Pcia. de Córdoba, Argentina, and Criaderos Buck, La Dulce, Pcia. de Buenos Aires, Argentina, respectively) with different potential kernel mass, and four sowing dates: 21 July 1995 (S1); 4 Sept. 1995 (S2); 18 Dec. 1995 (S3), and 27 Mar. 1996 (S4). Although the last two sowings were beyond those agronomically acceptable, they were included so as to expose the crops to extreme conditions during the kernel filling period. The cultivars were chosen because of their differences in potential kernel mass, with B. Ombú as the cultivar with potentially heavier kernels than P. Federal (Pedrol and Castellarín, 1989; Calderini et al., 1999a). Treatments were arranged in a split-plot design with three replications. Main plots were assigned to sowing dates and subplots to cultivars. The experimental units (subplots) consisted of seven rows, 0.20 m apart and 3 m long, with a north–south orientation.

Plant Husbandry
Sowing rates ranged from 350 to 450 seeds/m2 in all plantings. One week after seedling emergence, plants were hand-thinned to 300, 320, 400, and 350 plants/m2 in S1, S2, S3, and S4, respectively, to compensate for differences in tillering rate. Plots were fertilized at sowing with 60, 100, 100, and 120 kg N/ha in S1, S2, S3, and S4, respectively. These N rates were based on soil tests for nitrates, and the aim was to increase total soil N to 180 kg N/ha. Water stresses were avoided by maintaining the plots with adequate water availability throughout the study. For this purpose, plots received water two or three times weekly (depending on the season) from either natural rainfall or irrigation. When irrigated, the plots received water to field capacity (climatic data corresponding to this experiment is provided in Table 1) .


View this table:
[in this window]
[in a new window]
  		Table 1 Monthly maximum and minimum temperature, solar radiation, and rainfall for the period from the first sowing to maturity of the fourth sowing of the experiment conducted in Buenos Aires during the 1995–1996 growing season. Months during which sowings were conducted are shown in italics

 
The experiments were maintained free of biotic stresses. Weeds were periodically removed by hand. Diseases (powdery mildew) and insects (aphids) were controlled by application of tebuconazole ({alpha}-[2-{4 chlorophenyl} ethyl-{alpha}-{1-1- dimethyl-ethyl}-1H-1,2,4 triazol-1-ethanol) and triadimephon (1[4 chloro phenoxyl-3,3 dimethyl 1-{1H-1,2,4 thriazol-1-il}-2-butomene]), respectively.

Measurements
The date of anthesis (stage 65) was recorded for each cultivar using the scale proposed by Zadoks et al. (1974). From anthesis onward, one main stem spike was harvested from each treatment at least twice weekly. Three kernels from two central spikelets were removed to study moisture and dry matter dynamics. Kernel positions were defined as closest to the rachis (K1), second from the rachis (K2), and furthest from the rachis (K3). Spikes were harvested at 1200 h and kernel fresh mass was measured immediately after harvest. Kernel dry mass was measured after drying the samples for 48 h at 75°C. Kernel fresh and dry mass were measured with a precision balance (Sartorius, Germany; 0.1-mg resolution).

Kernel dry mass and physiological maturity (stage 95; Zadoks et al., 1974), assessed as the time when kernel growth ended, were estimated using a linear model subject to boundary conditions (i.e., kernel mass is described by three equations with two boundaries, c and e). To fit the kernel mass data over time, the following equations were used:

(1)

(2)

(3)
where KM is kernel mass (mg), a is the intercept (mg), b is the rate of kernel filling (mg/°Cd) for the period of slow dry matter accumulation (commonly termed the lag phase), c is the thermal time at which the lag phase ended (°Cd), d is the rate of kernel filling (mg/°Cd) for period of higher dry matter accumulation, e is the thermal time at which the linear dry matter accumulation ended (°Cd) (i.e., physiological maturity), f is the rate of kernel filling (mg/°Cd) after physiological maturity (i.e., rate of kernel filling = 0), and x is the thermal time after anthesis (°Cd). The duration of phases in thermal time units was calculated as the sum of daily average temperature [(maximum temperature + minimum temperature)/2] at a base of 0°C. The base temperature used in our work was 0°C (the value commonly used in wheat; see Hay and Kirby, 1991; Slafer et al., 1994; Calderini et al., 1996). The same procedure was used for independent data sets, described below, to validate the model.

Parameters described above (KM, a, b, c, d, e, and f) were iteratively calculated by fitting least squares until no improvement was obtained with further iterations using the optimization routine of Table Curve (Jandel, 1991). Estimates of the kernel filling duration and final kernel mass were derived from the fitted model (see Miralles et al., 1996).

Kernel water concentration of each sample was calculated as:

(4)
where KWCi is kernel water concentration (%) at time i, KWCoi is kernel water content (mg) at time i, and FKMi is fresh kernel mass (mg) at time i. Kernel water content was calculated as kernel fresh mass (mg) - kernel dry mass (mg).

Relative kernel dry matter was calculated as:

(5)
where RKDMi is relative kernel dry matter at the time i relative to that at maturity (%), KDMi is kernel dry matter (mg) at time i, and FinKM is kernel mass at maturity (mg) derived from the three-equation model (Eq. [1–3]).

To analyze the relationship between KDM and KWC for each cultivar and kernel position, a two-equation regression model was used:

(6)

(7)
where {alpha} is the intercept (mg), ß is the rate of kernel filling per unit of decrease of KWC (mg/%), {phi} is KWC when the kernel reaches final kernel mass (%), and x is KWC (%). As with the three-line model, the two-line model was fitted to the data using the optimization routine of Table Curve (Jandel, 1991).

The relationship between water and dry matter dynamics of the kernels was used because KWC can easily be measured under field conditions using a grain moisture meter.

Experiments for Validating the Regression Model
Data from three independent experiments (V1, V2, and V3) were used to validate the regression model between water and dry matter dynamics of the kernels (Eq. [6] and [7]), and to test its ability to calculate physiological maturity. The first experiment (V1), aimed to evaluate the effect of detaching basal florets (K1 and K2 positions) on final kernel mass at K3 position, was conducted at the experimental field of the Faculty of Agronomy (University of Buenos Aires) in 1996. Treatments consisted of two levels of floret detachment in the two central spikelets of the spike at heading (with and without detachment of florets in positions K1 and K2). The experiment was sown on 20 July in a completely randomized block design with three replicates. The cultivar was Buck Ombú, plant density was 300 plants/m2, and plots were fertilized and maintained free of biotic and abiotic stresses. Water stresses were avoided as described above for the experiment where four sowing dates were studied. At heading (stage 55; Zadoks et al., 1974), 40 main shoot spikes were selected in each plot. Florets were detached from kernel positions K1 and K2 on 20 spikes. Florets on the remaining 20 spikes were not detached and were used as controls. Twice weekly, one spike of each treatment was harvested at 1200 h, and fresh and dry matter (dried at 75°C for 48 h) of kernels corresponding to the two central spikelets were weighed with a precision balance (Sartorius, Gottingen, Germany; 0.1-mg resolution) and the weights were recorded.

The other experiments (V2 and V3) were conducted under field conditions in 1998 at the International Maize and Wheat Improvement Center (CIMMYT), El Batán, Mexico (19°31' N, 98°50' W, elevation 2249 m above sea level). Sowing dates were 22 May (V2) and 17 June (V3). Within each date, treatments consisted of two high-yield-potential cultivars released by CIMMYT (Bacanora and Rayón) used in a completely randomized design with three replicates. In both experiments, seeds were sown on raised beds spaced 80 cm apart. Plots contained two beds 5 m long, each with two rows 20 cm apart and 33 plants per m of row. Experimental plots were fertilized and maintained free of biotic and abiotic stresses. The trials were surface-irrigated at sowing and irrigation was continued as required, at approximately 50% depletion of available water, until physiological maturity. Sample procedures and fresh and dry matter determinations of kernels were similar to those described for experiment V1; however, in experiments V2 and V3, spikes were harvested early in the afternoon (1500 h) and kernel fresh and dry mass were measured with an electronic balance (Mettler, Zurich, Switzerland; 0.1-mg resolution) and recorded.


   Results
TOP
 ABSTRACT
 INTRODUCTION
 Materials and methods
 Results
Discussion
 REFERENCES
 
Physiological Maturity Estimation
In the study conducted in Buenos Aires during 1995–1996, the average air temperature of the kernel filling period was substantially affected by sowing date (Calderini et al., 1999b). The highest and lowest average air temperatures for the kernel filling period were recorded for S3 (24.3°C) and S4 (15°C), respectively. The average air temperature for the kernel filling period was similar between cultivars within each sowing date because the cultivars reached anthesis and physiological maturity on similar dates (Calderini et al., 1999b). Conversely, final kernel mass was clearly affected by cultivar (P < 0.001) and kernel position (P < 0.001). It was also affected by sowing date, but at a lower statistical level (P < 0.10) (Table 2) . Final kernel masses showed a wide range of values (between 22.6 and 44.4 mg; Table 2). The relationship between final kernel mass and average air temperature during kernel filling was curvilinear in P. Federal and linear in B. Ombú (Calderini et al., 1999b).


View this table:
[in this window]
[in a new window]
  		Table 2 Final kernel mass corresponding to the kernel positions closest to the rachis (K1), second from the rachis (K2), and third from the rachis (K3) for cultivars P. Federal and B. Ommbú in the first (S1), second (S2), third (S3), and fourth (S4) sowing dates of the experiment conducted in Buenos Aires during the 1995–1996 growing season. Values in parentheses are the standard error of the mean

 
Although KWC was not affected by cultivar and kernel position when this trait was plotted against days after anthesis, sowing date clearly modified kernel water dynamics (Fig. 1) : The decrease in KWC was clearly slower in S4 than at other sowing dates, and S2 had a faster rate of KWC loss than the other dates, particularly in P. Federal (Fig. 1A).



View larger version (14K):
[in this window]
[in a new window]
  		Fig. 1 Relationship between kernel water concentration (KWC) and days after anthesis for the cultivars (A) P. Federal and (B) B. Ombú during the first, second, third, and fourth sowing dates (S1, S2, S3, and S4, respectively) of the experiment conducted in Buenos Aires during the 1995–1996 growing season. Within each sowing date, data are the average of grain positions K1, K2, and K3

 
Differences between sowing dates in KWC were similar to those in dry matter accumulation. Therefore, it was possible to analyze the relationship between these two traits for each cultivar and kernel position independently of the sowing dates (Fig. 2) . Thus, all sowing dates had similar rates of increase of kernel dry matter associated with the decrease of KWC for each cultivar and kernel position (Fig. 2, Table 3) . All these relationships were highly significant (P < 0.001) and showed similar KWC both at the beginning of the kernel filling period and at the time when kernels ended their dry matter accumulation (Table 3). The only difference between cultivars or kernel positions was the rate of dry matter accumulation (Table 3).



View larger version (20K):
[in this window]
[in a new window]
  		Fig. 2 Relationship between kernel mass and kernel water concentration (KWC) for the cultivars (A) P. Federal and (B) B. Ombú during the kernel filling period of the four sowing dates corresponding to the experiment conducted in Buenos Aires during the 1995–1996 growing season. Data is shown for kernel position K1 (closed circles), K2 (open squares), and K3 (open triangles). Lines show the regression analysis fitted by a linear model with one boundary (see Table 3) for each kernel position: K1 and K3 (top and bottom solid lines, respectively) and K2 (dashed line)

 

View this table:
[in this window]
[in a new window]
  		Table 3 Intercept ({alpha}), slope (ß), end of dry matter accumulation of kernels ({phi}), and coefficient of determination (r2) of the linear model with one boundary for the relationship between kernel mass (mg) and kernel water content (%). Data show the traits corresponding to the three kernel positions closest to the rachis (i.e., K1, K2, and K3) for cultivars P. Federal and B. Ombú, for the four sowing dates of the experiment conducted in Buenos Aires during the 1995–1996 growing season

 
To develop a regression model to estimate physiological maturity independently from final kernel mass, the RKDM, instead of absolute dry matter, was considered. After transforming the dry matter values to RKDM for each situation, both cultivars and all kernel positions showed a single relationship between KWC and RKDM (Fig. 3) . To obtain this relationship for each cultivar, within the range where kernels were losing water but were still accumulating dry matter, only data with lower KWC than that at which kernel mass leveled off (see Fig. 2 and Table 3) were considered. For all cultivars and kernel positions, this corresponded to KWC values lower than 36% (Table 3). The vertical line illustrates moisture loss after 100% kernel dry matter was reached (Fig. 3).



View larger version (20K):
[in this window]
[in a new window]
  		Fig. 3 Relationship between kernel water concentration (KWC) and relative kernel dry matter (RKDM) corresponding to the three grain positions closest to the rachis of cultivars (A) P. Federal and (B) B. Ombú during the kernel filling periods of the four sowing dates in the experiment conducted in Buenos Aires during the 1995–1996 growing season. The line shows the regression analysis of this relationship for kernel water concentrations higher than 36%. The line of kernel water concentrations lower than 36% was fitted by hand, and these data were not included in the regression analysis

 
Based on the equations shown in Fig. 3, the proposed regression model for calculating current kernel mass (as a proportion of final kernel mass) between anthesis and physiological maturity is

(8)
where RKDMi is relative kernel dry matter at time i relative to that at maturity (%) and KWCi is kernel water concentration (%) at time i. The only measurement required is the determination of KWC. This value is then substituted into Eq. 8.

Model Validation
Our proposed model (Eq. [8]) to determine RKDM from KWC at any time from anthesis to physiological maturity was tested with independent data from experiments V1 (Fig. 4A and 4B) , V2 (Fig. 4C and 4D), and V3 (Fig. 4E and 4F). In all cases except Fig. 4B, both the RKDM and KWC were calculated for all kernels per spike (weighted average of kernel positions).



View larger version (25K):
[in this window]
[in a new window]
  		Fig. 4 Relationship between relative kernel dry matter (RKDM) and days after anthesis for measured (solid symbols) and calculated (open symbols) data. Data correspond to (A, C, D, E, and F) the weighted average of all kernel positions from the central spikelets of the spike in experiments V1, V2, and V3, and (B) kernel position K3 of experiment V1. In this experiment, kernel position K3 for control (circles and triangles for measured and estimated data, respectively) and detached (squares and diamonds for measured and estimated data, respectively) treatments are shown

 
The dynamics of RKDM were well calculated for experiment V1 (Fig. 4A) using the proposed regression model (Eq. [8]). This was true even in the case in which only the lightest and most variable (in dry mass) distal kernels (K3) were analyzed (Fig. 4B). In addition, the model also fitted data from the detached kernel treatment (Fig. 4B), despite the fact that this treatment significantly (P < 0.05) increased final kernel mass for position K3 (41.5 vs. 35.4 mg in the control).

In experiments V2 and V3, two cultivars different from those used to construct the model were evaluated. In addition, these cultivars set one more kernel per spikelet in the two central spikelets of the spike, and exhibited an even wider range of final kernel mass, than those measured in experiment V1 (Table 4) . Despite the large differences in cultivars, number of kernels per spikelet, and environmental conditions, the model generated a good estimation of RKDM throughout the kernel filling period (Fig. 4C–4F). However, in experiment V3, the model slightly underestimated measured values. The general good fit of the model to calculate RKDM in these independent studies (Fig. 4) is shown by the calculated-to-measured relationship with slope not statistically different from 1 (P < 0.05), and intercept not statistically different from 0 (P < 0.01).


View this table:
[in this window]
[in a new window]
  		Table 4 Final kernel mass corresponding to the three or four kernel positions closest to the rachis (i.e., K1, K2, K3, and K4) for cultivars B. Ombú (experiment V1), Bacanora, and Rayon (experiments V2 and V3). The experiments were conducted in Buenos Aires during 1996 (experiment V1) and in Mexico during 1998 (experiments V2 and V3) growing seasons

 
Finally, independent published data obtained under controlled conditions were used to provide a more general evaluation of our proposed regression model. The slope of the relationship between KWC and RKDM , calculated from data reported by Sofield et al. (1977), Millet and Pinthus (1984), Tashiro and Wardlaw (1990), Nicolas et al. (1985), and Stone and Nicolas (1995), was similar to that of the present study (represented by the solid line in Fig. 5) . Good agreement was found between the regression model proposed in the present study and the data of the studies conducted under controlled conditions when the calculated-to-measured relationship was evaluated. The slope of this relationship was not statistically different from 1 (P < 0.01), and the intercept was not statistically different from 0 (P < 0.01). Despite the general agreement, the model tended to slightly underestimate RKDM for the cultivars Oaxley and Egret, which are from the study by Stone and Nicolas (1995).



View larger version (19K):
[in this window]
[in a new window]
  		Fig. 5 Relationship between kernel water concentration (KWC) and relative kernel dry matter (RKDM) for data obtained in different experiments conducted under controlled conditions. Data correspond to proximal kernels (K1) from central spikelets of the spikes (there were no data for distal kernels; i.e., K3) of cultivars Triple Dirk (Sofield et al., 1977), Banks (Tashiro and Wardlaw, 1990), H-18 and B.L. 24 (Millet and Pinthus, 1984); Oxley and Egret (Stone and Nicolas, 1995), and Warigal (Nicolas et al., 1985). The line shows the relationship between kernel water concentration and relative kernel dry matter obtained in the present study (see Fig. 3)

 

   Discussion
TOP
 ABSTRACT
 INTRODUCTION
 Materials and methods
 Results
 Discussion
REFERENCES
 
Our results, as well as results from Schnyder and Baum (1992), showed that the values of KWC and dry matter are consistently related during most of the kernel filling period. However, whereas Schnyder and Baum (1992) found that this steady relationship was maintained only between 75 and 46% KWC, our results show that the relationship is linear over a wider range of KWC (75 and 37%).

The regression model (Eq. [8]) for estimating RKDM has shown a good fit to data from different cultivars, kernel positions, kernel mass potential, and environmental conditions (see Fig. 4 and 5). No cases were found in which the estimated date of physiological maturity could imply loss of potential yield due to a premature harvest time.

Differences reported between KWC at physiological maturity in literature (e.g., Hanft and Wych, 1982; Schnyder and Baum, 1992) could be related to the methodological approaches used by the authors. As dry matter accumulation slowly ceases when kernels are reaching physiological maturity, only small changes can be measured shortly before this stage, while the KWC changes markedly (see Sofield et al., 1977; Egli and TeKrony, 1997). Thus, calculation of KWC at or around physiological maturity could be strongly influenced by the experimental approach, number of data points, and environmental effects. However, the contrasting environments, different cultivars, and different kernel positions evaluated in the present study have shown little effect on KWC at physiological maturity. Moreover, the time of the day (1200 h or afternoon) at which the spikes were harvested apparently had little impact on water and dry matter dynamics. This does not imply that this value would be constant throughout the day or that samples of kernels in the field should be taken during the same interval (about 1200 to 1700 h) to that used for constructing the model. On the contrary, it is to be expected that values of KWC measured early in the morning would be higher than at 1200 h.

The regression model proposed in the present work for estimating physiological maturity was based on and validated with data from irrigated experiments. Despite the fact that drought during grain filling does significantly reduce final grain mass in wheat (e.g., Nicolas et al., 1984), it is speculated that the relationship between relative dry matter and KWC would be equally effective in predicting physiological maturity under dryland conditions. This speculation is borne out by the fact that a linear relationship between final grain mass and maximum water content of grains was found in well-watered and water-stressed plants of wheat and maize (Westgate, 2000). This relationship confirms the correspondence between the dry matter and grain water dynamics of grains grown under different environmental conditions.

The slight underestimation found for data obtained in experiment V3 was probably related to the unusual rainy season experienced in El Batán during the last days of the kernel filling period of this experiment. This may have produced a higher value of fresh kernel mass and, consequently, a higher KWC near physiological maturity. Therefore, an increase of KWC could be expected during rainy days. In addition, differences in synchrony among tillers should also be considered when estimating physiological maturity at crop level. For this reason, it could be convenient to measure KWC at different tiller strata if the number of spikes per plant is large (normally, the first two tillers mature simultaneously with the main shoot).

According to the equation proposed above (Eq. [8]), KWC at physiological maturity averaged 37%. This value is similar to the highest value of the range (30–35%) suggested by Dodds et al. (1979) for windrowed hard red spring wheat without serious loss of yield in Western Canada. Considering the 95% confidence limits of the proposed equation, the relative KWC at physiological maturity could range between 33 and 41%. For those cases evaluated in which physiological maturity coincided with the lower extreme of the confidence interval (33%), the average loss of potential yield (Fig. 4) would be 3.5% (depending on the actual final kernel mass, the highest loss was 7.5% and the lowest negligible). On the other hand, the decision to harvest at 37% KWC in cases where the actual value is the upper limit of the confidence interval (41%) would produce no yield penalties, but could cause a slight delay in harvesting with respect to the actual date of physiological maturity. The exact length of this delay cannot be accurately stated, as it would partially depend on variable environmental conditions such as air temperature, humidity, and wind (see Brooking, 1990; Otegui and Slafer, 1996).

The most common indirect method of determining physiological maturity in wheat is to assume that it coincides with the complete loss of green color from the peduncle or flag leaf (see Hanft and Wych, 1982). This method has the disadvantage of depending on a qualitative trait (it is not always simple to determine when the green color is completely lost); that is overcome by the quantitative method proposed here. In addition, the proposed model could also be used to estimate the proportion of dry matter that remains to be accumulated if the measurement is made previous to physiological maturity. This may be considered as an indirect estimation of the time between the time when the KWC is measured and physiological maturity.

The methodology for estimating physiological maturity in wheat presented here (the regression model between RKDM and KWC) could be useful in other crops where there are not simple early visual signs of the cessation of kernel dry matter accumulation. However, the resulting regression models (i.e., the values of the coefficients shown in Eq. [8]) could differ from those calculated here due to their differences in KWC at physiological maturity (see Egli, 1998).


   ACKNOWLEDGMENTS
 
We thank Dr E.H. Satorre for critically reviewing the manuscript. We also thank Sonia Hall for revising the English usage and Luís Hercum (Faculty of Agronomy, University of Buenos Aires) and the personnel of CIMMYT's Wheat Physiology Lab for their important technical assistance. This work was supported by grants from the University of Buenos Aires and Fundación Antorchas as well as an overseas scholarship of the University of Buenos Aires (Argentina).

Received for publication April 8, 1999.
   REFERENCES
TOP
 ABSTRACT
 INTRODUCTION
 Materials and methods
 Results
 Discussion
 REFERENCES
 




This article has been cited by other articles:


Home page
 Crop Sci.Home page
R. G. Sala, F. H. Andrade, and M. E. Westgate
Maize Kernel Moisture at Physiological Maturity as Affected by the Source-Sink Relationship during Grain Filling
Crop Sci., March 1, 2007; 47(2): 711 - 714.
[Abstract] [Full Text] [PDF]

Home page
 Agron. J.Home page
D. Farrer, R. Weisz, R. Heiniger, J. P. Murphy, and M. H. Pate
Delayed Harvest Effect on Soft Red Winter Wheat in the Southeastern USA
Agron. J., April 11, 2006; 98(3): 588 - 595.
[Abstract] [Full Text] [PDF]

Home page
 Crop Sci.Home page
B. L. Gambin and L. Borras
Sorghum Kernel Weight: Growth Patterns from Different Positions within the Panicle
Crop Sci., January 31, 2005; 45(2): 553 - 561.
[Abstract] [Full Text] [PDF]

Home page
 ANN BOT (LOND)Home page
L. BORRAS, M. E. WESTGATE, and M. E. OTEGUI
Control of Kernel Weight and Kernel Water Relations by Post-flowering Source-sink Ratio in Maize
Ann. Bot., June 1, 2003; 91(7): 857 - 867.
[Abstract] [Full Text] [PDF]

Home page
 Crop Sci.Home page
D. F. Calderini and I. Ortiz-Monasterio
Grain Position Affects Grain Macronutrient and Micronutrient Concentrations in Wheat
Crop Sci., January 1, 2003; 43(1): 141 - 151.
[Abstract] [Full Text] [PDF]

This Article
Right arrow Abstract Freely available
Right arrow Figures Only
Right arrow Full Text (PDF) Free
Right arrow Alert me when this article is cited
Right arrow Alert me if a correction is posted
Services
Right arrow Similar articles in this journal
Right arrow Similar articles in ISI Web of Science
Right arrow Alert me to new issues of the journal
Right arrow Download to citation manager
Right arrow reprints & permissions
Citing Articles
Right arrow Citing Articles via HighWire
Right arrow Citing Articles via ISI Web of Science (15)
Right arrow Citing Articles via Google Scholar
Google Scholar
Right arrow Articles by Calderini, D. F.
Right arrow Articles by Slafer, G. A.
Right arrow Search for Related Content
PubMed
Right arrow Articles by Calderini, D. F.
Right arrow Articles by Slafer, G. A.
Agricola
Right arrow Articles by Calderini, D. F.
Right arrow Articles by Slafer, G. A.
Related Collections
Right arrow Statistics
Right arrow Crop Physiology & Metabolism
Right arrow Wheat
Right arrow Crop Models

HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS
The SCI Journals Crop Science Vadose Zone Journal
Journal of Natural Resources
and Life Sciences Education		 Soil Science Society of America Journal
Journal of Plant Registrations Journal of
Environmental Quality		 The Plant Genome