Numeric codes for developmental stages of oat apex in the growing conditions of Southern Finland

Scales that categorize the development of cereal apex have been introduced by several authors. Daylength markedly affects the rate of apex organogenesis and, hence, numeric codes for developmental stages of apices introduced for lower latitudes cannot be used in Finland without modifying them into a ratio scale. The present study introduces numeric codes that categorize the 22 developmental stages ofoat (Avena saliva L.) apex in the growing conditions of Southern Finland. Field experiments, including 14 oat cultivars and six breeding lines that differ in the duration of the period from initiation of double ridges to pollination, were carried out at the Viikki Experimental Farm of the University ofHelsinki (60°13’N) for two years. The numeric codes were established by relating them to cumulated degree days (CDD). This study showed that when estimating the developmental stages of oat apex by calculating CDD from sowing, the error of estimation was ±1 stage unit at the most.


Introduction
Temperature, daylength, and precipitation affect crop growth and development.In cereals, the higher the temperature and the longer the photoperiod, the higher the rate of development and the less leaf, spikelet, and floret primordia per head (Rashid et al. 1984, Kirby et al. 1985, Cao and  Moss 1989, Craufurd and Cartwright 1989,  Miglietta 1989,Peltonen-Sainio 1993).The effects of water deficit on apical development are, however, equivocal.In oats, moderate and severe water stress considerably reduced floret and grain set, but did not affect the duration of pre-anthesis phase (Peltonen-Sainio 1991).
Several scales describing the development of ce- real apex from initiation of doubleridges to pollina-tion have been introduced for use in crop manage- ment (Table I).Precise timing of inputs -e.g., application of supplemental nitrogen according to developmental stages ofapices (Darwinkel 1983,  Peltonen and Peltonen 1990, Peltonen 1992) - is an important strategy particularly in integrated crop production.Moreover, phenoxy acetic acid herbicides, such as MCPA and 2,4-D, whenapplied close to the double ridge stage, may cause apex abnormalities including twisted rachis, unilateral spikelet set, fusion of spikelets, branched spikes, and retarded differentiation (Andersen 1954,  Loubser and Cairns 1989).In addition, certain developmental stages of apex may be particularly sensitive to injuries caused by insect pests and pathogens.Markedly reduced grain set, following infection caused by barley yellow dwarf virus at the         Waddington et al. (1983) and Åfors et al. (1988)   1 Codes described for oats, 2 barley, 3 wheat, and 4 rye early reproductive phase of oats, indicates such sensitivity (Peltonen-Sainio and Karjalainen   1990).
When evaluating the effects of management practices and abiotic and biotic stresses on apex organogenesis, numeric codes are needed to cat- egorize the development of apex in a ratio scale.However, numeric codes earlier introduced for barley and wheat are not linearly related to any factor.Furthermore, the applicability ofnumeric codes for different genotypes has to be tested because, for example, in the studies of Kirby and Appleyard  (1980), barley cultivars responded differently to daylength.
The present study was carried out to simulate the uniform numeric codes for developmental stages of oat apices.The codes established can be used in the photoperiodic conditions of Southern Finland.Evaluation of performance of different oat lines was based on testing of 20 oat cultivars and breeding lines that differ in duration of the period from initiation of double ridges to pollination.

Material and methods
Plant material consisted of 14 oat cultivars (Table 2) and six breeding lines bred at the Hank- kija Plant Breeding Institute, Finland.The oat lines were tested in experiments at the Viikki Experimental Farm of the University of Helsinki (60°13' N) in [1989][1990].A completely random- ized block design with four replications in 1989and three in 1990 was used.Plot size was 10 m, and 500 viable seeds were sown per m .Planting date was 27 April in both years, and soil type was sandy clay.80 kg N ha" 1 (NH4NO3) was applied at sowing together with P, K, etc. fertilizers.Weeds were controlled with MCPA [(4-chloro-2methylphenoxyl)acetic acid] at a rate of 700 g ha 1 after the double ridge stage.
The developmental stages of the oat apices were recorded from 10 randomly sampled main stems per plot according to the scale introduced by Åfors et al. (1988).The stage of development was deter- mined on the terminal spikelet, which is the most advanced spikelet in oats.The interval of plant samplings was between two and seven days de- pending on the rate of differentiation.Cumulated daily growing degree days (CDD) to reach different developmental stages were calculated using +5°C as the base temperature.
Numeric codes, at ratio scale, for each developmental stage of apex were established to replace the letter codes of Åfors et al. (1988).The letter code A was changed to 1, B to 2, C to 3 and so on.Correlations between these integers that correspond to the letter codes A-Y and the weighted means of CDD required to reach the developmental stages were calculated over years and cultivars (MSTAT  Development Team 1989).Regression coefficients ) and the codes for the different developmental stages adjusted until correlation was absolute (R"=1.00 ). Numeric codes (CODE) over genotypes were established using the formula: [1 ] CODE = -6.08 + 0.067 * CDD where CDD is cumulated degree days required to reach different developmental stages of apex from transitionapex to pollination.

Results and discussion
When evaluating the effects of management practices and abiotic and biotic stresses on apex organogenesis, numeric codes are needed to categorize the development of apex in a ratio scale.In this study, numeric codes were established over 20 genotypes (Table 3).The developmental stage, when carpel primordium is initiated, was not recorded in this study.This stage of apex is rapidly transient and hard to distinguish from the preceding one.
When testing the validity of the regression equation [ 1 ] for different years and genotypes, our res- ults showed that CDD from sowing explained 88-99% of the variation in the developmental stage of apices in 1989 and 95-98% in 1990.The median was 98% in 1989 and 97% in 1990.Precipitation was higher at pre-anthesis in 1989 than in 1990 (Fig. 1).In 1990, degree days cumulated faster at early growth stages, and double ridges initiated about one week earlier than in 1989.Since then, degree days cumulated more slowly in 1990, and oats pollinated five days later than in 1989.
Comparison ofregression coefficients for different oat genotypes by F-test showed that apical de- velopment of 11 genotypes did not differ signific- antly between years.The 11 genotypes are all bred in Finland, whereas the nine genotypes that were excluded from further analyses included, for in-stance, the foreign cultivars.Data on both years for the 11 genotypes were combined and new regres- sion equations were established.Then, CDD from sowing explained 96-98% of the variation in the development of apices (Table 4).The more ad- vanced the developmental stage, the higher was the deviation from the uniform trend (Fig. 2).Further grouping of oats, e.g., into old and modem cultivars and breeding lines, was not statistically justified.In general, differences in intercepts and regression co- efficients between the eleven oat genotypes were modest (Table 4).The relatively homogeneous re- sponse of lines to CDD, recorded in this study, can be attributed to resemblance of genetic background of oats bred in Finland (REKUNEN 1988, JUSSILA et  al. 1992).
Regarding estimation of the developmental stage of apex by calculating CDD from sowing, our res- ults showed that root mean square errors (Root MSB) for different genotypes ranged from 0.94 to 1.39.It means that estimation of developmental stages of apices by calculating CDD from sowing resulted in error of ±1 developmental stage unit at the most.The residual scatter plot of cultivar Veli, shown in Figure 3, is similar to that of other culti- vars examined.The three developmental stages that may be incorrectly estimated, i.e., they mixed with each other, are when styles began to elongate until stigmatic branches elongated (Fig. 3).Because these stages of apex are particularly rapidly tran- sient, they may mix with each other.Logarithmic modifications of the variables did not improve the   validity of estimation of the three developmental stages of apex.However, CDD estimates accurately the double ridge stage and, hence, phenoxy acetic acid herbicides should not be sprayed until CDD >l7O °C in oats to avoid apex abnormalities and possible crop losses.

Fig. 2 .
Fig. 2. Regressions between development of apex and CDD in 11 oat lines.Broken lines indicate confidence interval at 95% level.

Fig. 3 .
Fig.3.Residual scatter plot in Veli.Residual was calculated by subtracting the simulated developmental stage from the observed one.
of numeric and letter codes that describe development of apices.

Table 2 .
Oat cult!vars tested in the study, year of release, and country of origin.

Table 3 .
CDD from sowing to reach different developmental stages of apex and estimated numeric codes(CODE)for growing conditions of Southern Finland.

Table 4 .
Intercepts and regression coefficients between established numeric codes for developmental stages of apex and CDD in 11 oat cultivars.