HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS		QUICK SEARCH:   [advanced] Author: Keyword(s): Year:  Vol:  Page:

This Article Abstract Figures Only Full Text (PDF) Free Alert me when this article is cited Alert me if a correction is posted Services Similar articles in this journal Alert me to new issues of the journal Download to citation manager Citing Articles Citing Articles via Google Scholar Google Scholar Articles by Vocanson, A. Articles by Jeuffroy, M.-H. Search for Related Content PubMed Articles by Vocanson, A. Articles by Jeuffroy, M.-H. Agricola Articles by Vocanson, A. Articles by Jeuffroy, M.-H. Related Collections Other Legumes Crop Growth and Development Other Crop Management Other Cropping Systems Field-Scale Studies

LEGUMES

Agronomic Performance of Different Pea Cultivars Under Various Sowing Periods and Contrasting Soil Structures

UMR 211 Agronomie, INRA AgroParisTech, BP 01, 78850 Thiverval-Grignon, France

^* Corresponding author (jeuffroy{at}grignon.inra.fr).

	ABSTRACT

TOP NOTES ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION CONCLUSIONS REFERENCES

 
Yield variability of spring pea (Pisum sativum L.) in farmers' fields is mainly due to soil compaction at sowing and abiotic stresses during the reproductive period. Winter peas flower earlier, and thus should be less sensitive to abiotic stress at the end of the cycle, but because of their sensitivity to frost they must be sown late in autumn when soils are very wet. Pea breeders are working on new winter cultivars that are more resistant to frost and highly sensitive to photoperiod and that could be sown earlier in autumn under better soil conditions. Our aim was to measure crop growth, development, and seed yield of different types of pea cultivars using factorial experiments that evaluated the effect of soil structure, sowing period, and cultivar at two sites in France. Compaction reduced seed yield by 18% at Grignon and 6% at Estrées-Mons, but had no significant effect on crop development. November sowings resulted in increased seed yield of about 1 t ha^–1 at both sites compared with spring sowings, and were associated with larger numbers of reproductive nodes and seeds m^–2. Cultivars exhibited contrasting characteristics in terms of mean seed dry weight, seeds m^–2, and number of reproductive nodes, but final seed yields were similar within sowing periods. The ability of the crop to convert crop growth rate to seed number was independent of environment but decreased linearly with mean seed dry weight of the cultivar. This research suggests that pea breeders should develop new winter pea cultivars with earlier flowering to enhance yield.

Abbreviations: BF, beginning of flowering • BM, aboveground dry biomass accumulation • DAS, number of days after sowing • dd, degree-days • EF, early fall • HI, harvest index • LF, late fall • LW, late winter • Ngunit, seed number per unit of crop growth rate • PM, physiological maturity • SD, standard deviation

	NOTES

TOP NOTES ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION CONCLUSIONS REFERENCES

 
All rights reserved. No part of this periodical may be reproduced or transmitted in any form or by any means, electronic or mechanical, including photocopying, recording, or any information storage and retrieval system, without permission in writing from the publisher.

Received for publication November 2, 2005.

	INTRODUCTION

TOP NOTES ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION CONCLUSIONS REFERENCES

 
PEA YIELD IS HIGHLY VARIABLE between years and between fields (Doré et al., 1998). The high yield variability and the lack of increased national average yield across years are major reasons for the great decrease in the amount of peas grown in France between 1993 (737,500 ha) and 2004 (354,000 ha; Union Nationale Interprofessionnelle des Plantes Riches en Protéines, 2005). Moreover, a large proportion of this area is sown with spring peas (90% in 2004; Union Nationale Interprofessionnelle des Plantes Riches en Protéines, 2005). The yield variation of spring peas is highly correlated to the amount and distribution of the spring rainfall (Uzun and Açikgöz, 1998), as peas are highly sensitive to water stress during the period of seed formation (Lecoeur and Guilioni, 1998; Guilioni et al., 2003; Ney et al., 1994). Furthermore, Jeuffroy et al. (1990) reported that high temperatures during the same period are detrimental to yield, causing seed abortions. For spring cultivars, heat and water stress occur, on average, 1 yr out of 2 in the Parisian basin, where large areas are sown with peas (Guilioni and Jeuffroy, 2005; Lecoeur and Guilioni, 2005).

The beginning stage of flowering is the visible start of the seed set period and thus a key stage in yield formation. Winter peas sown in autumn generally flower 2 to 3 wk earlier than spring-sown peas in France (Lejeune-Hénaut et al., 1999). Therefore autumn sowings should provide earlier seed set, thus escaping the main abiotic stresses. But the effects of changing the sowing season on pea crop growth are debated. For instance, Ranalli et al. (1997) found no significant advantage from autumn sowing on seed yield because of a large interaction between sowing date and year. In contrast, Uzun and Açikgöz (1998) showed that autumn, rather than spring, sowings produced the highest biomass and seed yield at maturity, regardless of the cultivar used. These studies compared the effect of sowing season only on seed yield and its components but not on crop growth. Furthermore, the comparison has always been between spring and autumn sowings, but no study compared the effect of various sowing dates within the autumn season on crop growth and development.

The commercial winter pea cultivars must be sown very late in autumn (not before mid-November in northern France) to maintain frost resistance until the end of winter (Biarnès et al., 2004). New winter pea lines are being bred that have a longer period of resistance to frost and thus can be sown at the beginning of October in France (Etévé et al., 2004). These cultivars are called "Hr" because they have, in their genome, a single dominant gene, Hr, which is responsible for their higher sensitivity of flowering date to photoperiod than most commercial pea cultivars (Lejeune-Hénaut et al., 1999). The earlier period of flowering should allow avoidance of the period of abiotic stress at the end of the crop cycle. Their highly indeterminate growth habit should allow compensating for fewer seeds on some nodes due to adverse conditions during seed formation but greater seeds on other nodes, thus leading to lower variation in seed yield. Moreover, as shown by Mokashi et al. (1997) and Gosse et al. (1986) for many species, yield potential is often increased by a longer growth cycle. It is therefore assumed that these new Hr lines could lead to more stable and higher yields, but their agronomic performances under field conditions have never been quantified, in comparison with commercial varieties.

As discussed by Jeuffroy and Ney (1997), current data on plant development and crop growth and their effect on yield formation are available for spring peas and can be used to analyze the performances of various genotypes. For instance, Dumoulin et al. (1994) used the general pattern of pea plant development proposed by Ney and Turc (1993) to compare the variable characteristics of 10 contrasting genotypes and their effects on seed yield. For example, the number of nodes with pods ranged between 4.5 for the spring cultivar Solara and 12.4 for the winter cultivar Alex and the mean seed dry weight ranged between 0.12 g seed^–1 for the winter pea cultivar Frisson and 0.27 g seed^–1 for the spring pea cultivar Solara (Roche, 1998). Moreover, Poggio et al. (2005) reported cultivar differences in aboveground biomass, seed yield, and pod and seed numbers m^–2. Because pea cultivars exhibit various sensitivities to environmental conditions (e.g., weather conditions, photoperiod), it is hypothesized that sowing date and cultivar should influence seed yield.

To avoid water and heat stress during reproductive stages, farmers try to sow spring peas as early as possible after the cold and wet winter and thus sometimes when the soil is too wet, resulting in compaction (Doré, 1992). For a spring sowing and a spring cultivar, the yield reduction due to soil compaction can reach 30% depending on weather during the growing season (Crozat et al., 1992). Because of differences in environmental conditions during the crop cycle between autumn and spring sowings, soil compaction should not be as great a factor in crop growth and development for different sowing dates. But to date, no research has been performed.

In this study, our objective was to compare crop development, growth, and yield of various pea cultivars in response to soil compaction, sowing date, and cultivars.

	MATERIALS AND METHODS

TOP NOTES ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION CONCLUSIONS REFERENCES

 
Sites and Treatments
Field experiments were performed at Grignon, France (48°53'60'' N, 1°53'60'' E) and at Estrées-Mons, France (49°42'0'' N, 3°23'60'' E) during two growing seasons, 2002–2003 (respectively, Exp. 1 and Exp. 3) and 2003–2004 (respectively, Exp. 2 and Exp. 4). At both sites, the soil was a clay loam soil (Hapludalf): (21:70:9 clay:silt:sand in Exp. 1 and 2 and 21:74:5 clay:silt:sand in Exp. 3 and 4). Soil depths, defined as the depth of soil above bed rock, were 1.1 m in Exp.1 and 2 and more than 10 m in Exp.3 and 4.

Four pea cultivars were sown at each site: a commercial spring type, Baccara (0.28 g seed^–1, Florimond-Desprez, France); two commercial winter types, Cheyenne (0.20 g seed^–1, GAE-Recherche, France) and Dove (0.19 g seed^–1, Agri-obtentions, France); and a winter photo-reactive line, Hr200 (0.13 g seed^–1, INRA). Each cultivar was sown during three sowing periods: early fall (EF), late fall (LF), and late winter (LW). The precise dates for each site-year experiment are given in Table 1 .

The sowing density was 90 seeds m^–2. Each plot was 1.75 m wide by 30 m long in Exp. 1 and 2 and 1.25 m wide by 24 m long in Exp. 3 and 4. The row spacing was 0.17 m at both sites. Diseases, insects, and weeds were controlled by standard local pesticide applications.

At both sites, the crops were sown in randomized split-plot designs with three replications. At Grignon, the main plots were soil structure and the subplots were sowing period and cultivar. At Estrées-Mons, because of other experimental constraints, the main plots were sowing period and soil structure, and the subplot was cultivar.

Development
As soon as the first flowers began opening, we counted open flowers every 2 d on 10 stems randomly chosen per replicate. The beginning of flowering (BF) stage was determined as soon as 50% of the stems in a plot had at least one open flower (Maurer et al., 1966).

Pod life is divided into two successive phases: seed set and seed filling. During seed set, seed abortions can occur. At the end of this first period, called the final stage in seed abortion (Pigeaire et al., 1986), the final seed number of the pod is fixed. The total seed number per stem is fixed when the last pod of the stem reaches this stage. Because this stage is not easy to observe at the field scale, it was calculated ex-post from the final number of reproductive nodes. Since the thermal time (temperature base of 0°C at 2-m height) between the BF and the final stage in seed abortion of the first pods of the stem (corresponding to the beginning of seed filling at the stem level) is 265 degree-days (dd) for spring cultivars and 300 dd for winter cultivars (Dumoulin et al., 1994), and the thermal time between the final stage of seed abortion on two successive nodes is constant and equal to 30 dd (Ney and Turc, 1993), the thermal time between the BF and the final stage in seed abortion of the stem can be estimated using the following equation:

where NRN = mean number of reproductive nodes per mean stem; dd_bf–bsf and dd_bf–fssa = cumulative dd (temperature base of 0°C) from BF to beginning of seed filling and final stage in seed abortion, respectively.

For Exp.1 and 2, maturity was estimated in the field for each treatment when 50% of the plants were completely yellow (Chaillet and Biarnès, 2004). In some treatments, maturity was then verified by measuring the water content of the seeds of the last node on 10 stems randomly chosen in the plot. The measured maturity stage was reached when 50% of the last pods had their seeds containing less than 0.55 g g^–1 water in their fresh weight (Le Deunff, 1988).

Plant Growth
To measure aboveground biomass of peas, plants were harvested once a month from emergence to maturity by taking one sample (0.85 m² in Exp.1 and 2 and 1 m² in Exp.3 and 4) per replicate. A sample was also taken precisely at the BF. For each sample, stems were cut at the base of the green part of the stems to harvest only aerial matter. Samples were then washed and packed, dried in an oven for 48 h at 80°C, and then weighed.

Harvest
At maturity, the mean number of reproductive nodes per stem was measured for the samples. For 10 stems randomly sampled per replicate, the number of reproductive nodes bearing at least one pod with seeds was counted. Seeds were then separated from the vegetative parts of the sample. The vegetative biomass was dried for 48 h at 80°C in an oven and then weighed.

Dry seed yield and its components (seeds m^–2 and mean seed dry weight) were estimated by harvesting large areas within each plot (10 m long x 1.75 m wide in Exp. 1 and 2, and 5 m long x 1.25 m wide in Exp. 3 and 4). After harvest, seeds were separated and weighed to get their fresh weight. For each sample, approximately 0.50 kg of seeds was precisely weighed and then dried for 48 h at 80°C in an oven and weighed. Dry seed yield was then calculated from fresh seed weight and seed water content. For all seed samples, the number of seeds was counted using a Contador (Pfeuffer) to determine seed number m^–2 and mean seed dry weight. Harvest index (HI) was calculated as the ratio between seed yield and total dry biomass (straw and seeds) at maturity.

As observed by several authors (Egli, 1993; Guilioni et al., 2003; Jeuffroy and Ney, 1997), seed number m^–2 is highly correlated with crop growth rate during the period of seed formation, as described by the following equation:

where SN = seed number m^–2; BM_fssa and BM_bf = aboveground dry biomass at final stage in seed abortion and BF, respectively (g ha^–1); dd_bf and dd_fssa = cumulative dd (temperature base of 0°C) from sowing until BF and final stage in seed abortion, respectively; Ngunit = seed number per unit of crop growth rate (seeds g^–1 dd^–1). To determine the aerial biomass at the final stage in seed abortion, the biomass measurement taken on the closest sampling date (before or after this stage) was considered; it was never more than 8 d from the calculated date.

Weather Conditions
Minimum and maximum daily air temperatures (°C), daily rainfall (mm), daily Penman evapo-transpiration (mm), and mean daily total radiation (MJ m^–2 d^–1) were measured by permanent weather stations located within 200 m of each experimental site.

To quantify the effect of high temperature stress on crop growth and development, the number of days with maximum daily air temperatures above 25°C between the BF and the final stage in seed abortion on the plot was counted (Jeuffroy et al., 1990).

A daily water balance was calculated to determine the intensity of water stress (Brisson et al., 1992):

where WR(d) = the quantity of water available for crop requirements in the soil on day d; Irr(d) = the amount of water from irrigation applied on day d; R(d) = daily rainfall; ETP(d) = daily evapotranspiration. The balance was initialized on 1 January assuming the soil was saturated. The number of days with WR(j) = 0 was calculated during the period of seed formation and the period of seed filling.

Characterization of Soil Compaction
Soil compaction was characterized using image analysis of soil profiles of the surface 20-cm layer following the methodology of Roger-Estrade et al. (2000). A soil pit (1.5 m wide in Grignon and 1.0 wide in Estrées-Mons, 0.5 m deep at both sites) was dug. In the plowed layer, the severely compacted zones with no visible macroporosity, called in the present paper (Manichon and Gautronneau, 1987), were marked using a knife on the observation surface of the soil pit (Roger-Estrade et al., 2000). The percentage of soil with a structure was then calculated for the plowed surface area using image analysis (software Optimas 6.0.) on photographs of the observation surface of the soil pit taken with a digital camera (Sony, DSC-S75). Soil profiles of the plowed layer were characterized on 8 Apr. 2003 for Exp. 1, 11 May 2004 for Exp. 2, 20 Mar. 2003 for Exp. 3, and 21 Apr. 2004 for Exp. 4 for the uncompacted treatments at EF, LF, and LW and the compacted treatments at LF and LW. Only one cultivar per sowing date and site was observed, because this measurement is time consuming and the soil compaction did not differ according to the cultivar.

Statistical Analysis
Statistical analyses were performed using the GLM procedure (type III) (SAS Institute, 1987) separately for Exp. 1 and 3 and Exp. 2 and 4 to take into account differences in experimental design at the two locations. Since experiments were conducted in different fields during the 2 yr at each site, a block effect was introduced into the analysis of variance that took into account the combined effects of weather, field, year of the experiment, and replicate. Finally, four factors were studied in the analysis (block, soil structure, sowing date, and cultivar) and all the first-order interactions between soil structure, sowing date, and cultivar.

For each treatment, the aboveground dry biomass accumulation (BM) was characterized by nonlinear regressions fitted to the aboveground dry biomass measurements of the three replicates using the NLIN procedure of SAS (SAS Institute, 1987). The beta function of Yin et al. (2003) was used as a reference:

where t = time, days after sowing (DAS); BM_max = maximum value of BM, t ha^–1; t_e = time when BM_max is reached; t_m = the abscissa of the inflection point, where the growth rate reaches its maximum value (Fig. 2). The growth function of Yin et al. (2003) was used because it is flexible in describing various asymmetrical sigmoid patterns by varying the value of t_m. It is useful in our case because the length of the initial phase of slow growth differs greatly between autumn and spring sowings. The parameter BM_max was the mean of the maximum biomass values observed at maturity in most situations. The maximum growth rate, c_m (t ha^–1 d^–1), was given by

Because the parameters of this equation correspond to plant growth traits, the comparison of plant growth among treatments was accomplished by quantifying the effects of soil structure, sowing date and cultivars on these parameters.

View larger version (15K): [in this window] [in a new window]   Fig. 2. Aboveground dry biomass (BM, t ha–1) according to the number of days after sowing. (A) Exp. 3 uncompacted LF Cheyenne, R2 = 0.98; (B) Exp. 4 uncompacted LW Dove, R2 = 0.99. The points represent the observed values per replicate and the curve is the fitted curve of the beta growth function. BF = beginning of flowering; FSSA = final stage in seed abortion; PM = physiological maturity; BMmax = maximum value of BM; te = time when BMmax is reached; tm = abscissa of the inflection point at which the growth rate reaches its maximum value; cm = growth rate at tm. An asterisk indicates the parameters that were experimentally determined.

	RESULTS

TOP NOTES ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION CONCLUSIONS REFERENCES

View this table: [in this window] [in a new window]   Table 2. Weather characteristics for the four experiments.

View larger version (35K): [in this window] [in a new window]   Fig. 1. Proportion of the plowed layer with clods in the four experiments according to soil structure (UC: uncompacted; C: compacted) and sowing date (EF: early fall; LF: late fall; LW: late winter). Treatments with the same letter above the bars are not significantly different (P < 0.05) within an experiment.

View this table: [in this window] [in a new window]   Table 3. Date of beginning of flowering for each experiment according to the sowing date, the soil structure, and the cultivar.

The three commercial cultivars differed in terms of earliness to flower: for a given sowing date, the spring cultivar Baccara flowered between 1 and 12 d before the winter cultivar Dove, and Dove flowered between 0 and 6 d before Cheyenne. On average for the four experiments, taking into account the recommended sowing date for each cultivar, Dove sown in LF flowered on 7 May (standard deviation [SD] of 7.3 d), Cheyenne sown in LF flowered on 11 May (SD of 6.1 d), and Baccara sown in LW flowered on 24 May (SD of 7.9 d).

The number of days with maximum temperatures above 25°C between the BF of the earliest cultivar and the physiological maturity (PM) of the latest differed among experiments: 25 d in Exp. 1, 10 d in Exp. 2, 16 d in Exp. 3 and only 4 d in Exp. 4. For the three commercial cultivars, the number of days with maximum temperatures above 25°C during the seed set period depended on sowing date and varied between 0 and 3 d for EF, 0 and 6 d for LF, and 3 and 14 d for LW (Table 4 ). For Hr200, because of a late BF, regardless of sowing date, seed set period occurred during a risky period for heat stress (between 4 and 16 d in all conditions).

View this table: [in this window] [in a new window]   Table 4. Number of days with maximum temperature higher than 25°C between the beginning of flowering and the final stage in seed abortion in each experiment according to the sowing date, the soil structure and the cultivar.

View this table: [in this window] [in a new window]   Table 5. Number of days with no water available in soil for crop requirements between the beginning of flowering (BF) and the final stage in seed abortion (FSSA) and between this stage and the physiological maturity (PM), in each experiment according to the sowing date, the soil structure, and the cultivar.

The eight models of variance analysis, explaining the four parameters of the growth curve on the two locations, accounted for 66% to 99% of the variance of the four studied parameters (Table 6 ). At both sites, the best fitting models were for the prediction of t_e and t_m. At both sites, the parameters t_e and t_m differed significantly between years (P < 0.05). In contrast, at both sites, BM_max did not depend on the year (P > 0.05). Only at Grignon was c_m significantly higher in Exp. 1 than in Exp. 2 (P < 0.05).

View this table: [in this window] [in a new window]   Table 6. Results of the analysis of variance: effects of experiment, soil structure, sowing date and cultivar (B = Baccara, C = Cheyenne, D = Dove, H = Hr200) on the four parameters of the beta growth function.

For the three commercial cultivars, the growth cycle lasted 92 d more (at Grignon) and 114 d more (at Estrées-Mons) for LF than for LW. The time between t_m and t_e ranged from 26 to 29 d for LF and from 21 to 25 d for LW. Finally, BM_max was significantly higher for LF than for LW (P < 0.05). The parameter c_m was similar for LF and LW in Grignon (P > 0.05), whereas it was significantly higher for LW than for LF at Estrées-Mons (P < 0.05). For Hr200, the earlier the sowing date, the lower BM_max. For EF, the cycle was 20 and 128 d longer than LF and LW, respectively. The time between t_e and t_m was similar for EF and LF, but higher than for LW. The sowing date also affected c_m; the highest value was observed for LW in comparison with EF and LF. Considering all treatments, t_m occurred between 16 d before and 9 d after the final stage in seed abortion for EF and LF (on average, 2 d before the final stage in seed abortion), whereas it was between 0 and 16 d before the final stage in seed abortion for LW (on average 8 d before the final stage in seed abortion).

The effect of cultivar on the four parameters varied according to site. At Grignon, the four cultivars exhibited a similar BM_max, whereas in Estrées-Mons, Cheyenne and Baccara had a significantly higher BM_max than the two other cultivars (P < 0.05). The longest crop cycle was observed for Hr200 at all sites (P < 0.05), but no significant difference was observed among the three other cultivars. At Grignon, t_m was the highest for Hr200, whereas no significant difference was observed at Estrées-Mons. The four cultivars were significantly different with respect to c_m (P < 0.05); the highest values were observed for Cheyenne and Dove. For the three commercial cultivars, t_m occurred between 16 d before and 3 d after the final stage in seed abortion. Only the interaction between sowing date and cultivar was significant (P < 0.05) at Grignon for BM_max and c_m.

Seed Yield and Its Components
Experiment, soil structure, sowing date, cultivar, and interactions accounted for 78% of the variance of seed yield at Grignon and 60% at Estrées-Mons (Table 7 ). Seed yield differed significantly between years only at Grignon (P < 0.05). Compaction significantly reduced seed yield (P < 0.05) at both sites. At Grignon only, the interactions between compaction and sowing date or cultivar were significant (P < 0.05). In three experiments out of four for LW, compaction significantly reduced seed yield, whereas this happened in only one out of four for LF. In three situations out of four for Baccara, compaction significantly reduced seed yield, but in only one out of four for Dove. Mean yield was significantly higher for LF treatment than LW but this depended on the cultivar, as manifested by the significant interaction between these two factors (P < 0.05). In Exp. 2 for Baccara, yield was lower for LF than LW, whereas for Hr200, yield was similar for both sowing dates in Exp. 2, 3, and 4. For Hr200, at Grignon, the highest yield was observed for LF (4.7 t ha^–1) and the lowest for EF (3.7 t ha^–1). The three commercial cultivars had similar seed yields at both sites. In contrast, seed yield of Hr200 was significantly lower at Grignon (P < 0.05), but not at Estrées-Mons (P > 0.05). The coefficient of variation differed among cultivars and sites: at Grignon, the most stable yield was observed for Baccara (coefficient of variation of 12%, Table 7), whereas, at Estrées-Mons, the most stable yield was for Hr200 and Dove (coefficients of variation of 4% and 6%, respectively).

View this table: [in this window] [in a new window]   Table 7. Seed yield (SY), seed number m–2 (SN), mean seed dry weight (SDW), number of reproductive nodes per plant (NRN), and harvest index (HI) for each experiment, soil structure, sowing date, and cultivar.

Mean seed dry weight was the variable best explained by the factors studied, since the coefficients of determination were 98% at Grignon and 95% at Estrées-Mons (Table 7). It was the most stable seed yield component as it did not depend on the experiment, soil structure, and sowing date at either site (P > 0.05). In contrast, the cultivar effect was significant (P < 0.05).

The factors studied and their interactions accounted for 81% of the variation in number of reproductive nodes per stem at Grignon and 79% at Estrées-Mons (Table 7). Experiment had a significant effect at both sites (P < 0.05). The number of reproductive nodes per stem was significantly lower for compacted than for uncompacted treatments only at Grignon (P < 0.05). A significantly higher number of reproductive nodes per stem was observed for LF than LW at both sites (P < 0.05). For Hr200, the number of reproductive nodes per stem was significantly higher in EF than in LF and LW. At both sites, cultivar had a significant effect on number of reproductive nodes per mean stem (P < 0.05).

Experiment, soil structure, sowing date, cultivar, and interactions explained 82% of the variance of HI at Grignon and 77% at Estrées-Mons (Table 7). Harvest index was significantly different among experiments (P < 0.05) but not between soil structures and sowing dates (P > 0.05). For Hr200 at Grignon, HI increased with later sowing dates. Finally, Baccara and Dove had the highest HI at both sites, the ranking of the two other cultivars depended on the site.

For each cultivar, seed yield was positively correlated to seed number m^–2, the coefficient of determination ranged from 0.72 to 0.98 (Fig. 3A ). The highest slope value was observed for Baccara and the lowest slope for Dove. In contrast, no significant relationship was observed between seed yield and mean seed dry weight, except for Hr200 (R² = 0.80, Fig. 3B).

View larger version (23K): [in this window] [in a new window]   Fig. 3. Relationships between seed yield (SY, t ha–1) and (A) seed number m–2 (SN) and (B) mean seed dry weight (SDW, g seed–1) for each cultivar in all experiments.

View this table: [in this window] [in a new window]   Table 8. Results of the analysis of variance: effects of experiment, soil structure, sowing date, and cultivar on Ngunit (seeds g–1 degree-day–1) according to the site.

View larger version (20K): [in this window] [in a new window]   Fig. 4. Relationship between mean cultivar seed dry weight (SDW, g seed–1) and Ngunit (seeds g–1 degree-day–1). The regression line was calculated using the data of the five cultivars: Baccara, Cheyenne, Dove, Frisson, and Solara. Vertical bars are the standard deviation of Ngunit.

	DISCUSSION

TOP NOTES ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION CONCLUSIONS REFERENCES

In our experiments, the parameter Ngunit (ratio between the seed number m^–2 and the crop growth rate during the period of seed set) did not vary between years, soil structures, and sowing dates, which agrees with the results of Guilioni et al. (2003). In contrast, the crop growth rate close to the final stage in seed abortion was at a maximum for November sowings, whereas it had decreased by about 8 d for the February–March sowings. This reduction was probably due to heat and water stresses (Lecoeur and Sinclair, 1996; Guilioni et al., 2003) and explained the lower seed numbers m^–2 in LF and LW sowings. The higher biomass production for November sowings in comparison to LW sowings was already observed by Peksen et al. (1977), and is probably due to differences in cumulative incident solar radiation (Monteith, 1977) linked with a longer crop cycle (5 and 3 mo for November and February–March sowings, respectively). A higher BM was also observed by Gesch et al. (2002) for Cuphea, while Uzun and Açikgöz (1998) observed the same for pea, and it was linked to an increased branch formation. In contrast, the lowest BM_max was observed for the October sowing due to severe frost during the 2002–2003 winter, and the aboveground foliage was severely damaged, reducing the spring pea growth. Moreover, in addition to increased risk of frost damage, sowings in early autumn are subject to increased risk of diseases (Heenan, 1994) and weed competition (Rasmussen, 2004), two main problems that were not quantified in our study.

The HI varied from 0.33 to 0.72, within the range of values observed by Lecoeur and Sinclair (2001). As observed by Uzun and Açikgöz (1998), there was little difference in HI between autumn and spring sowings for the commercial varieties. For the cultivar Hr200, HI was higher for spring sowings, maybe because of more severe stress during seed set, as indicated by high temperature and water stress indicators. This result may be due to the later flowering period for this cultivar.

Effect of Cultivar
Pea cultivars differ in their resistance to frost (Lejeune-Hénaut et al., 1999; Ranalli et al., 1997). For Exp. 1 EF at Grignon, because of a warm autumn, the phase of frost tolerance for the three commercial cultivars was already finished when the frost occurred, and only Hr200 survived. In Exp. 3, all the cultivars sown in EF died. In Exp. 3 LF, only the cultivar Baccara died because of the cold winter.

The c_m differed significantly among the cultivars, probably because of differences in the rate of leaf appearance (Poggio et al., 2005; Roche, 1998) and radiation use efficiency (Lecoeur and Ney, 2003). Few differences in the date of BF were observed between the three commercial cultivars, perhaps due to their similar responses to photoperiod and temperature as observed by many authors for legumes (Bourion et al. (2002) for pea; Acosta-Gallegos et al. (1996) for common bean). In spite of differences in seed number m^–2, mean seed dry weight, and number of reproductive nodes per stem, the three commercial cultivars had similar average yields at all sowing dates: 6.1 t ha^–1 at Grignon and 6.3 t ha^–1 at Estrées-Mons. In contrast, the lower yield of Hr200 could be explained by its late BF in comparison with the three other cultivars; yield potential was more affected by water and heat stresses. Breeders are currently working to advance the BF of photo-reactive cultivars like Hr200 to escape high temperature and late water stress (Lejeune-Hénaut et al., 1999).

Linear relationships between seed yield and seed number m^–2 were found, as in Poggio et al. (2005), but slopes and intercepts varied among cultivars, according to the mean seed dry weight of the cultivar. For the three commercial cultivars, the higher the mean seed dry weight, the lower the number of reproductive nodes per stem and the steeper the slope of the regression line. Cultivars with a low mean seed dry weight should exhibit seed yields highly dependent on seed number m^–2. Nevertheless, as reported by many authors (Doré, 1992; Guilioni et al., 2003; Lecoeur, 1994), seed number m^–2 depends on environmental conditions, particularly soil and weather conditions. At Grignon, heat stress was more severe (in frequency and intensity) than at Estrées-Mons. Baccara flowered 1 to 12 days before Dove and had a shorter period of seed set due to a lower number of reproductive nodes, so this cultivar had the most stable yield at Grignon because it was less affected by abiotic stress. At Estrées-Mons, the wetter weather led to a less severe water stress than at Grignon, regardless of the cultivar. Cultivars with more indeterminate growth, such as Dove, would be expected to minimize the effects of abiotic stress due to compensatory effects between nodes (Guilioni et al., 2003). More studies are needed to confirm these results in a larger range of growing conditions. However, this research suggests that cultivars with a high number of reproductive nodes should be grown in areas characterized by low risks of abiotic stress, as they lead to greater yield stability across years.

Similar to results obtained for winter wheat by Leterme et al. (1994), no relationship was found between mean seed dry weight and seed yield, except for Hr200. The later period of seed filling due to a later BF explains the greater sensitivity of seed yield to mean seed dry weight for this cultivar.

The absence of the effect of soil structure and sowing date on the ability of the crop to convert crop growth rate into seed number (Ngunit) is consistent with the results of Guilioni et al. (2003) who observed that water and heat stresses had no effect on this efficiency. Guilioni et al. (2003) did not find any effect of the cultivar on this parameter, in contradiction to our observations. This may be due to the narrower range of mean seed dry weight (from 0.20 g seed^–1 to 0.30 g seed^–1) in their experiment. Moreover, for commercial pea cultivars, the negative relationship between mean seed dry weight and Ngunit is important for breeding cultivars with higher yield potential. Hr200 is a line still being bred, which probably explains its large variability.

Effect of Soil Structure
The significant reduction in seed yield caused by soil compaction is in agreement with results reported by Crozat et al. (1991). At Estrées-Mons, the smaller effect on seed yield and the absence of an effect on BM were probably due to nonsignificant differences between uncompacted and compacted soil structures. In Exp. 3, the excessive soil water at sowing in spring increased the percentage of zones in the uncompacted treatment by up to 38% (compared with 64% in the compacted treatment). In Exp. 4, insufficient soil water at the time of compaction resulted in a low percentage of zones in the plowed layer for the compacted treatments at all sowing dates. Only 34% and 42% of zones were found in the compacted treatments of the autumn and the spring sowings, respectively. However, as reported by Lindemann et al. (1982) for soybean, although seed yield is not significantly affected by compaction, a consistent trend is discernable. In our study, compaction did not affect mean seed dry weight, in agreement with the observations of Dawkins et al. (1984).

At Grignon, the reduction in seed number m^–2 and the associated reduction in the number of reproductive nodes by compaction could be explained by reduced N (Jeuffroy and Ney, 1997). Doré (1992) observed a negative linear relationship between crop N uptake during the reproductive period and the proportion of zones in the plowed layer. The fraction of N fixed is also reduced by compaction (Crozat et al., 1992). Compaction reduces soil exploration by roots (Bengough and Mullins, 1990; Tsegaye and Mullins, 1994) and hence the ability of the crop to extract soil water (Tardieu and Manichon, 1987). Water stress could therefore have been greater in compacted than uncompacted treatments, with a larger resulting effect on number of reproductive nodes per mean stem. The effect of soil compaction on seed number may also be explained by crop growth. At Grignon, compaction reduced c_m by 20%, and c_m occurred close to the final stage in seed abortion, the period during which the risk of water stress was high, especially under compacted conditions as suggested by Tardieu and Manichon (1987). Finally BM_max was reduced by 16% under compaction, in agreement with the results of Lindemann et al. (1982) for soybean. Compaction had no effect on HI, the assimilate partitioning thus being unchanged between vegetative and reproductive parts of the crop.

	CONCLUSIONS

TOP NOTES ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION CONCLUSIONS REFERENCES

 
For commercial cultivars, yields are higher for November sowings than spring sowings, regardless of the cultivar. However, in these situations, compacted soils can lead to high yield loss. To reduce the risk of soil compaction, it is necessary to sow in EF, with photoreactive cultivars. However, the agronomic performance of cultivar Hr200 was poor in our study, mainly because of its late flowering date and thus its high sensitivity to heat and water stress during the period of seed set and seed filling. Breeders are trying to breed pea lines with earlier flowering stages, and this appears necessary to reduce the occurrence of late seasonal stress. With these breeding lines, changing cultural practices should stabilize and enhance yield.

	ACKNOWLEDGMENTS

 
We would like to thank J.Troizier and his team for assistance with trials at the experimental Unit at Grignon, and J.J. Stempniak and his team for the trial at Estrées-Mons. We also thank G. Grandeau, V. Tanneau, B. Le Fouillen, A. Chauveau, and M. Bazot for their help with the measurements on the crop. Financial support was obtained from Union Nationale Interprofessionnelle des plantes riches en Protéines, GAE-Recherche, and Institut National de la Recherche Agronomique (Action Impact). We would like to thank Alan Scaife for English language revision.

All rights reserved. No part of this periodical may be reproduced or transmitted in any form or by any means, electronic or mechanical, including photocopying, recording, or any information storage and retrieval system, without permission in writing from the publisher.

	REFERENCES

TOP NOTES ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION CONCLUSIONS REFERENCES

 

• Acosta-Gallegos, J.A., P. Vargas-Vazquez, and J.W. White. 1996. Effect of sowing date on the growth and seed yield of common bean (Phaseolus vulgaris L.) in highland environments. Field Crops Res. 49:1–10.
• ARVALIS–Institut du végétal. 2003. Pois protéagineux de printemps et d'hiver: Guide de culture. (In French.) ARVALIS, Paris.
• Bengough, A.G., and C.E. Mullins. 1990. Mechanical impedance to root growth: A review of experimental techniques and root growth responses. J. Soil Sci. 41:341–358.
• Biarnès, V., I. Chaillet, and B. Carrouée. 2004. Frost damages observed on winter grain legumes in France in 2003. p. 295. In AEP-INRA (ed.) Proc. 5th European Conf. on Grain Legumes. Dijon, France. 7–11 June 2004. AEP, Paris.
• Bourion, V., M. Duparque, I. Lejeune-Hénaut, and N.G. Munier.-Jolain. 2002. Criteria for selecting productive and stable pea cultivars. Euphytica 126:391–399.[Web of Science]
• Brisson, N., B. Seguin, and P. Bertuzzi. 1992. Agrometeorological soil water balance for crop simulation models. Agric. For. Meteorol. 59:267–287.
• Chaillet, I., and V. Biarnès. 2004. Technical proceedings: Determination of development stages. p. 58–63, In V. Biarnès et al. (ed.) Agrophysiologie du pois protéagineux. (In French.) INRA-ARVALIS, Paris.
• Crozat, Y., J.P. Gillet, and A.M. Domenach. 1992. Nodulation, N₂ fixation and incorporation of combined N from soil into pea crop: Effects of soil compaction and water regime. p. 141–142, In A. Scaife (ed.) Proc. 1st European Conf. on Grain Legumes. Angers, France. 5–7 Dec. 1990. ESA, Colmar, France.
• Crozat, Y., J.P. Gillet, and F. Tricot. 1991. Effects of the soil compaction on root distribution N, P, K uptakes and growth of pea crop. In Proc. 3rd Int. Symp. of the Soc. of Root Res. Vienna, Austria. 2–6 Sept. 1991.
• Dawkins, T.C.K., M. McGowan, P.D. Hebblethwaite, and J.C. Brereton. 1984. The effect of thorough subsoil loosening on the yield of peas (1980) and a subsequent crop of winter wheat (1981). Res. Dev. Agric. 1:53–57.
• Doré, T. 1992. Analyse, par voie d'enquête, de la variabilité des rendements et des effets précédents du pois protéagineux de printemps (Pisum sativum L.). (In French.) INA, Paris-Grignon, France.
• Doré, T., J.M. Meynard, and M. Sebillotte. 1998. The role of grain number, nitrogen nutrition and stem number in limiting pea crops (Pisum sativum L.) yields under agricultural conditions. Eur. J. Agron. 8:29–37.
• Dumoulin, V., B. Ney, and G. Eteve. 1994. Variability of seed and plant development in pea. Crop Sci. 34:992–998.[Abstract/Free Full Text]
• Egli, D.B. 1993. Cutivar maturity and potential yield of soybean. Field Crops Res. 32:147–158.[CrossRef]
• Etévé, G., M. Boilleau, J.J. Stempniak, R. Devaux, and I. Lejeune. 2004. Breeding winter peas: Toward a new backround. p. 128. In AEP-INRA (ed.) Proc. 5th European Conf. on Grain Legumes. Dijon, France. 7–11 June 2004. AEP, Paris.
• Gesch, R.W., F. Forcella, N. Barbour, B. Phillips, and W.B. Voorheees. 2002. Yield and growth response of Cuphea to sowing date. Crop Sci. 42:1959–1965.[Abstract/Free Full Text]
• Gosse, G., C. Varlet-Grancher, R. Bonhomme, M. Chartier, J.M. Allirand, and G. Lemaire. 1986. Maximum dry matter production and solar radiation intercepted by a canopy. (In French, with English abstract.) Agronomie 6:47–56.[Web of Science]
• Guilioni, L. 1997. Effet des hautes températures sur la croissance végétative et le développement des organes reproducteurs chez le pois (Pisum sativum L., cv. ‘Messire’). Conséquences sur la production de biomasse et le nombre de graines. (In French.) ENSA de Montpellier, Montpellier, France.
• Guilioni, L., and M.H. Jeuffroy. 2005. Fortes températures et fonctionnement d'un couvert de pois. (In French.) p. 164–173. In V. Biarnès et al. (ed.) Agrophysiologie du pois protéagineux. INRA-Arvalis, Paris.
• Guilioni, L., J. Wéry, and J. Lecoeur. 2003. High temperature and water deficit may reduce seed number in field pea purely by decreasing plant growth rate. Funct. Plant Biol. 30:1151–1164.
• Heenan, D.P. 1994. Effects of sowing time on growth and grain yields of lupin and field pea in south eastern New South Wales. Aust. J. Exp. Agric. 34:1137–1142.
• Jeuffroy, M.H., C. Duthion, J.M. Meynard, and A. Pigeaire. 1990. Effect of a short period of high day temperatures during flowering on the seed number per pod of pea (Pisum sativum L.). (In French, with English abstract). Agronomie 10:139–145.[Web of Science]
• Jeuffroy, M.H., and B. Ney. 1997. Crop physiology and productivity. Field Crops Res. 53:3–16.[CrossRef]
• Lecoeur, J. 1994. Réponses phénologiques et morphologiques du pois (Pisum sativum L.) à la contrainte hydrique. Conséquences sur le nombre de phytomères reproducteurs. ENSA de Montpellier, Montpellier.
• Lecoeur, J., and L. Guilioni. 1998. Rate of leaf production in response to soil water deficits in field pea. Field Crops Res. 57:319–328.
• Lecoeur, J., and L. Guilioni. 2005. Déficit hydrique et fonctionnement d'un couvert de pois. p. 164–173, In V. Biarnès et al. (ed.) Agrophysiologie du pois protéagineux. (In French.) INRA-Arvalis, Paris.
• Lecoeur, J., and B. Ney. 2003. Change with time in potential-use efficiency in field pea. Eur. J. Agron. 19:91–105.
• Lecoeur, J., and T.R. Sinclair. 1996. Field pea (Pisum sativum L.) transpiration and leaf growth in response to soil water deficits. Crop Sci. 36:331–335.[Abstract/Free Full Text]
• Lecoeur, J., and T.R. Sinclair. 2001. Harvest index increase during seed growth of field pea. Eur. J. Agron. 19:91–105.
• Le Deunff, Y. 1988. Accumulation de la matière sèche chez le pois et sa modélisation. Plant Physiol. Biochem. 26:377–382.[Web of Science]
• Lejeune-Hénaut, I., V. Bourion, G. Etévé, E. Cunot, K. Delhaye, and C. Desmyter. 1999. Floral initiation in fields-grown forage peas is delayed to a greater extent by short photoperiods, than in other types of European varieties. Euphytica 109:201–211.[Web of Science]
• Leterme, P., H. Manichon, and J. Roger-Estrade. 1994. Analyse intégrée des rendements du blé tendre et de leurs causes de variation dans un réseau de parcelles d'agriculteurs du Thymerais. (In French, with English abstract). Agronomie 14:341–361.[Web of Science]
• Lindemann, W.C., G.E. Ham, and G.W. Randall. 1982. Soil compaction effects on soybean nodulation, N₂(C₂H₄) fixation and seed yield. Agron. J. 74:307–311.[Abstract/Free Full Text]
• Manichon, H., and Y. Gautronneau. 1987. Guide méthodologique du profil cultural. (In French.) GEAREA et CEREF, Paris.
• Maurer, A.R., D.E. Jaffray, and H.F. Fletcher. 1966. Response of peas to environment. III. Assessment of the morphological development of peas. Can. J. Plant Sci. 46:285–290.
• Mokashi, D.D., J.D. Jadhav, M.R. Shewale, C.B. Gaikwas, and J.D. Path. 1997. Performance of crops and cropping systems under different dates of sowing. Indian J. Agron. 42:220–223.
• Monteith, J.L. 1977. Climate and efficiency of crop production in Britain. Philos. Trans. R. Soc. London, Ser. B 281:277–294.[Abstract/Free Full Text]
• Munier-Jolain, N.G., N.M. Munier-Jolain, R. Roche, B. Ney, and C. Duthion. 1998. Seed growth rate in grain legumes. I. Seed growth is not affected by assimilates availability on seed growth rate. J. Exp. Bot. 49:1963–1969.[Abstract/Free Full Text]
• Ney, B., C. Duthion, and O. Turc. 1994. Phenological response of pea to water stress during reproductive development. Crop Sci. 34:141–146.[Abstract/Free Full Text]
• Ney, B., and O. Turc. 1993. Heat-unit-based description of the reproductive development of pea. Crop Sci. 33:510–514.[Abstract/Free Full Text]
• Peksen, E., H. Bozoglu, A. Peksen, and A. Gülümser. 1977. Determination of the effects of different row spacings on yield and some other properties of pea (Pisum sativum L.) cultivars sown in spring and autumn. Acta Hortic. 579:313–318.
• Pigeaire, A., C. Duthion, and O. Turc. 1986. Characterisation of final stage of seed abortion in indeterminate soybean, white lupin and pea. (In French, with English abstract.) Agronomie 6:371–378.[Web of Science]
• Poggio, S.L., E.H. Satorre, S. Dethiou, and G.M. Gonzalo. 2005. Pod and seed numbers as a function of photothermal quotient during the seed set period of field pea (Pisum sativum L.) crops. Eur. J. Agron. 22:55–69.
• Ranalli, P., M. Di Candilo, and V. Faeti. 1997. Performance in northern Italy of pea breeding lines for dry seed yield. Adv. Hortic. Sci. 11:85–90.
• Rasmussen, I.A. 2004. The effect of sowing date, stale seedbed, row width and mechanical weed conrol on weeds and yields of organic winter wheat. Weed Res. 44:12–20.
• Ridge, P.E., and D.L. Pye. 1985. The effects of temperature and frost at flowering on the yield of peas grown in a Mediterranean environment. Field Crops Res. 12:339–346.
• Roche, R. 1998. Modélisation de la variabilité des profils de graines associée à la séquentialité du développement chez différents génotypes de pois protéagineux de printemps. (In French.) INA, Paris-Grignon, France.
• Roger-Estrade, J., G. Richard, H. Boizard, J. Boiffin, and J. Caneill. 2000. Modelling structural changes in tilled topsoil over time as a function of cropping systems. Eur. J. Soil Sci. 51:455–474.
• Saharia, P. 1986. Relative performance of pea varieties to sowing dates. Indian J. Agron. 31:377–379.
• SAS Institute. 1987. SAS/STAT guide for personal computer. 6th ed. SAS Institute, Cary, NC.
• Tardieu, F., and H. Manichon. 1987. Etat structural, enracinement et alimentation hydrique du maïs. II.- Croissance et disposition spatiale du système racinaire. (In French, with English abstract). Agronomie 7:201–211.[Web of Science]
• Tsegaye, T., and C.E. Mullins. 1994. Effect of mechanical impedance on root growth and morphology of two varieties of pea (Pisum sativum L.). New Phytol. 126:707–713.[CrossRef][Web of Science]
• Union Nationale Interprofessionnelle des Plantes Riches en Protéines. 2005. Statistiques plantes riches en protéines. (In French). UNIP, Paris.
• Uzun, A., and E. Açikgöz. 1998. Effect of sowing season and seeding rate on the morphological traits and yields in pea cultivars of differing leaf types. J. Agron. Crop Sci. 181:215–222.
• Yin, X., J. Goudriaan, E.A. Lantinga, J. Vos, and H.J. Spiertz. 2003. A flexible sigmoid function of determinate growth. Ann. Bot. 91:361–371.[Abstract/Free Full Text]

This Article Abstract Figures Only Full Text (PDF) Free Alert me when this article is cited Alert me if a correction is posted Services Similar articles in this journal Alert me to new issues of the journal Download to citation manager Citing Articles Citing Articles via Google Scholar Google Scholar Articles by Vocanson, A. Articles by Jeuffroy, M.-H. Search for Related Content PubMed Articles by Vocanson, A. Articles by Jeuffroy, M.-H. Agricola Articles by Vocanson, A. Articles by Jeuffroy, M.-H. Related Collections Other Legumes Crop Growth and Development Other Crop Management Other Cropping Systems Field-Scale Studies