The impact of variation in grain number and individual grain weight on winter wheat yield in the high yield potential environment of Ireland

European Journal of Agronomy 87 (2017) 40–49

Contents lists available at ScienceDirect

European Journal of Agronomy journal homepage: www.elsevier.com/locate/eja

The impact of variation in grain number and individual grain weight on winter wheat yield in the high yield potential environment of Ireland

MARK

Joseph P. Lyncha, Deirdre Doylea, Shauna McAuleyb, Fiona McHardyb, Quentin Danneelsa, ⁎ Lisa C. Blackb, Ethel M. Whiteb, John Spinka, a b

CELUP, Teagasc, Oak Park, Co. Carlow, Ireland AFBI, Plant Testing Station, Crossnacreevy, Belfast BT6 9SH, United Kingdom

A R T I C L E I N F O

A B S T R A C T

Keywords: Winter wheat Yield Sink Source Tillering Grain weight

Previous studies from regions that produce high proportions of global winter wheat have highlighted that difference in sink size influences the majority of variations in winter wheat yield. However, the potential for source limitation due to environmental differences in regions that consistently produce a large sink capacity (i.e. > 20,000 grains/m2), such as Ireland, have not been widely studied. The aim of this study was to characterise the variation in growth pattern and yield components that contribute to variations in grain yield in regions of high yield potential, and to identify the periods of development that are most likely to influence yield in these regions. Monitor crops of winter wheat were grown at three sites with contrasting latitudes on the island of Ireland, during three growing seasons (2013–2015). Crops were assessed regularly for measurements of crop growth and development, including biomass accumulation, canopy development and light interception. Grain yield ranged between 10.7–15.8 t/ha at 15% moisture content, with a grand mean of 12.7 t/ha. Results indicated that variations in grains/m2 had a larger effect on winter wheat yield than variations in individual grain weight. Variability in grains/m2 was influenced by changes in spikes/m2 more than the number of grains/spike. While spikes/m2 at harvest was significantly related to the number of shoots/m2 at GS59, no significant relationship was observed between the shoots/m2 at the time of maximum tillers/plant and spikes/m2 at harvest. Furthermore, a significantly negative linear relationship was observed between shoots/m2 at the time of maximum tillers/plant and grains/spike. Therefore, high rates of tillering were not beneficial to yield formation in the majority of crops monitored. A strong effect of individual grain weight was observed at one site of the nine evaluated in the study, indicating that a partial source limitation of yield is possible in certain Irish environmental conditions. However, variations in grain yield of crops of winter wheat grown at different locations in Ireland in different seasons were primarily driven by variations in grain number, and therefore were generally sink-limited.

1. Introduction The pressure to maximise yields in Irish wheat production systems has increased recently, due to increased demand for grain caused by the large predicted increases in population by 2050 (Alexandrator and Briunsma, 2012), and the small profit margins available above high inputs costs required to protect crops against disease and weeds, especially during periods of low grain prices (O’Donovan, 2016). Despite substantial increases in the national average wheat yield from the 1960’s through to the late 1990’s, this trend has slowed substantially in the last decade (Fischer and Edmeades, 2010; CSO, 2016). The limiting factors causing this plateauing of yield can result from either a limitation of the sink size available to store yield during grain

⁎

Corresponding author. E-mail address: [email protected] (J. Spink).

http://dx.doi.org/10.1016/j.eja.2017.05.001 Received 1 November 2016; Received in revised form 3 May 2017; Accepted 8 May 2017 1161-0301/ © 2017 Published by Elsevier B.V.

filling or a limitation of the source available to produce assimilate for grain filling (Fischer, 1985). As the determination of sink size and potential source generally occurs at different periods of crop development with only a small degree of overlap (Slafer et al., 2014), the identification of which factor has a greater influence on the variation in yield for a certain region is important for targeting efforts in breeding, the development of new technologies and crop management, and to achieve a greater proportion of yield potential. During the past 30 years, many studies have investigated which factors limit the yield of cereal crops through alterations in the sink:source ratio, either through the manual reduction in sink size (Slafer and Savin, 1994; Acreche and Slafer, 2006; Calderini et al., 2006; Zhang et al., 2010; González et al., 2014), alterations in the

European Journal of Agronomy 87 (2017) 40–49

J.P. Lynch et al.

2.1. Experimental design

potential radiation interception (Fischer, 1985; Savin and Slafer, 1991; Beed et al., 2007; Ahmadi et al., 2009; Serrago et al., 2013; Asseng et al., 2017), the analysis of contrasting genotypes (Abbate et al., 1998; Shearman et al., 2005; Acreche and Slafer, 2009) or the analysis of crops grown in different environments (Bingham et al., 2007; PeltonenSainio et al., 2007; Kennedy et al., 2017). As summarised by Borrás et al. (2004) and Slafer et al. (2014), the majority of previous studies have indicated a major sink limitation (generally through grain number per m2) of winter wheat yield for the majority of global cereal crops, with reductions in the radiation source available per unit of sink postanthesis adequately compensated for by a utilisation of potential assimilates stored in the stem (Gent, 1994; Ehdaie et al., 2008; Serrago et al., 2013). Studies by Calderini et al. (2001), Xie et al. (2016) and Asseng et al. (2017) have highlighted a strong relationship between crop growth during the pre-anthesis period of spike development and a crops’ sink size, with unfavourable conditions potentially reducing grain number through increased shoot mortality (Sparkes et al., 2006) and a reduced fruiting efficiency for grain on the spike (González et al., 2011; Gonzalez-Navarro et al., 2016). However, many previous studies have also observed a source-sink “co-limitation” of yield, indicating a partial effect of a low source availability-high sink capacity environment on crop yield (Calderini et al., 2006; Peltonen-Sainio et al., 2007; Acreche and Slafer, 2009; Serrago et al., 2013; González et al., 2014). Winter wheat yields in Ireland are generally amongst the highest globally, facilitated by a climate that provides abundant rainfall without temperature extremes that allow for slow development and a long duration to capture resources (Burke et al., 2011). However, the Irish climate is also characterised by variability in the incidence of radiation during summer months due to cloud cover (Sweeney, 2014), and thus, the natural occurrence of a high sink-low source environment post-anthesis is plausible. Furthermore, it is important to determine the typical variation of winter wheat yield in a high yielding climate like Ireland, as Slafer et al. (2014) reported that while large changes in yield are mostly due to grains/m2, small changes may be due to either grains/m2 or individual grain weight. A field evaluation of the potential source or sink limitations of high-yielding winter wheat, based solely on site and seasonal variations in climate, without manual alterations of source or sink capacity has not previously been published, to the authors’ knowledge. Despite this, previous physiological studies conducted for winter (Bingham et al., 2007) and spring barley (Kennedy et al., 2017) in the UK and Ireland, respectively, have indicated a strong sink limitation for these crops despite relatively high sink capacities when compared to other international regions. These articles highlighted the importance of pre-anthesis growth to obtaining a high yield, and therefore identified a critical period in development during which crop management should be optimised. However, the greater potential for creating sinks and yield in crops of winter wheat, compared to barley, (Ugarte et al., 2007) likely results in a lesser source: sink ratio and a greater potential for source limitation. Furthermore, little information is available on the growing dynamics of wheat grown at different latitudes across the island of Ireland, as the climatic conditions can differ with region (Holden and Brereton, 2004), and thus, crop development and growth may differ somewhat. The aim of this study was to characterise the variation in growth pattern and the yield components that contribute to variations in grain yield at contrasting sites with high yield potential environments for wheat, and to identify the periods of development that are most likely to influence yield in these regions.

The experiments were located at three sites across the island of Ireland between 2012 and 2015, Crossnacreevy, Co. Down (54°33′N, 5°51′W; sowing dates: 8 November 2012, 29 October 2013, 4 December 2014), Oak Park, Co. Carlow (52°51′N, 6°54′W; sowing dates: 25 October 2012, 14 October 2013, 14 October 2014) and Killeagh, Co. Cork (51°56′N, 8°1′W; sowing dates: 23 October 2012, 15 October 2013, 6 November 2014), that represented the range in latitude of the typical wheat growing regions on the island. At Oak Park and Killeagh, each experiment consisted of six 2.5 × 24 m plots per site, which were arranged into three blocks of two plots, each containing one plot designated to observations/non-destructive analysis and the other to analysis that required destructive analysis throughout the season. At Crossnacreevy, the experiments each comprised three rows of eight plots, each 2.0 × 18 m, bordered by two plots, alternate plots being designated to observations/non-destructive analysis/harvest and to destructive analysis throughout the season. The variety sown in each experiment (Triticum aestivum L. cv. JB Diego, Senova, Cambridge, UK) was selected due to its commercial popularity and high-yielding performance in recommended list trials. Varieties were sown into continuous arable fields with good soil conditions following inversion ploughing and harrowing. Seeding rates were selected to achieve a target plant population of 260–280 plants/m2. Crops were treated with a three spray fungicide program to minimise disease incidence, and all other crop management, including herbicides, insecticides and nutrient management, were applied at rates to minimise crop stress and facilitate a high yielding crop. Specifically, inorganic nitrogen applications where applied in three splits (0.3:0.4:0.3 at GS25/30, GS30/31 and GS37, respectively) at the Oak Park and Killeagh sites, and two splits (0.5:0.5, GS30 and GS37, respectively) at Crossnacreevy in all three years. The total rates applied were 200, 180 and 200 kg N/ha at Oak Park and 200, 180 and 220 kg N/ha at Killeagh for 2013, 2014 and 2015, respectively, while 220 kg N/ha was applied at Crossnacreevy for each year. Other nutrients (P, K and S) were applied at rates that would be non-limiting to crop growth and development, as described by Wall and Plunkett (2016). All sites had been in continuous tillage crop production for at least ten years. Protection against disease, weeds and pests was successful at all sites. Lodging assessments were conducted every three days during the grain filling period, however no significant amounts of lodging was observed in any of the experiments. Weather data, including indices of temperature, radiation and rainfall, were obtained using the instruments and standards described by Fitzgerald and Fitzgerald (2004) from weather stations within 2 km of the experiments at Crossnacreevy (Skye MiniMet Automatic Weather Station, 2013-14 and courtesy of Agrii, Cheltenham, UK, 2014-15) and Oak Park (courtesy of Met Éireann, Dublin, Ireland), with weather data for the Killeagh site obtained from Roches Point, Co. Cork, 15 km southwest of the experimental site (courtesy of Met Éireann). Thermal time (°C days) was calculated using a base value of 0 °C by the method reported (method 1) by McMaster and Wilhelm (1997). Photosynthetically active radiation (PAR) was estimated as 0.5 the value of solar radiation (MJ/m2), similar to Bingham et al. (2007). Meteorological data were summarised for three periods during each growing season and presented in Table 1. 2.2. Growth stage assessment and leaf/tiller emergence Soon after crop emergence, 10 plants were selected at random and tagged in each observation plot. In addition, a tag was placed on the most recent emerged leaf and the number of leaves present was recorded. Observations were conducted on a monthly basis from emergence until GS30, with assessments of growth stage (Zadoks et al., 1974) conducted in addition to a count of the newly emerged leaves on the main stem since the previous observation, and the total number of tillers. Subsequent to GS30 crops were assessed on a weekly

2. Materials and methods The methodology described below is based on previous work on the growth and development of barley in the UK (Bingham et al., 2007) and Ireland (Kennedy et al., 2017). 41

European Journal of Agronomy 87 (2017) 40–49

J.P. Lynch et al.

Table 1 Meteorological data for sites used in this study. Incident PAR (MJ/m2)

Mean Temperature (°C)

Total Rainfall (mm)

Season 2013

Site Crossnacreevy Oak Park Killeagh

1 Oct–28 Feb 5.5 6.1 7.8

1 Mar–31 May 6.1 7.2 7.9

1 Jun–31 Aug 14.8 16.1 15.7

1 Oct–28 Feb naa 270 295

1 Mar–31 May 481 569 612

1 Jun–31 Aug 806 794 891

1 Oct–28 Feb 565 323 449

1 Mar–31 May 281 136 209

1 Jun–31 Aug 180 157 215

2014

Crossnacreevy Oak Park Killeagh

6.3 6.6 8.4

8.9 9.7 9.8

14.3 15.9 15.6

naa 268 308

610 545 627

746 747 845

705 552 673

172 196 213

275 173 140

2015

Crossnacreevy Oak Park Killeagh

6.5 6.5 8.5

7.6 8.4 9.2

13.2 14.2 14.0

265 287 326

659 654 661

725 753 791

623 453 527

284 169 207

256 192 235

1981–2010 averageb

Crossnacreevy Oak Park Killeagh

5.8 6.7 8.3

7.9 8.8 9.4

14.0 14.9 14.9

naa 262 296

naa 559 635

naa 687 789

547 400 479

243 179 200

254 191 212

Weather data from Republic of Ireland courtesy of Met Éireann. Weather data from Northern Ireland courtesy of UK Met Office and Agrii. a Long term photosynthetically active radiation (PAR) data in MJ/m2 for Northern Ireland are not available (na). b PAR long term average based on 10 year data (2006–2015) as no longer data available.

2.4. Radiation interception and radiation use efficiency (RUE)

basis until harvest. Estimates of phyllochron at each site were calculated as the gradient of the best fit line when data for the number of leaves emerged were linearly regressed against accumulated thermal time, as described by Kennedy et al. (2017).

An assessment of radiation interception by the crop canopy in each plot was conducted on sampling dates using a Sunscan Canopy Analysis System (Delta-T Devices, Cambridge, UK). Ten measurements of incident radiation and radiation below the canopy were taken per plot at a 45° angle to the crop row. Measurements were conducted between 10:00 and 14:00, when any form of precipitation was absent. The calculation of an extinction coefficient (kpar) for canopy PAR transmission was estimated at each site from light interception recordings taken on the date closest to anthesis (GS61) using the formula:

2.3. Canopy size, biomass accumulation and N content On the same recording dates described in Section 2.2, a single quadrat (0.7 × 1.0 m) was obtained from each destructive sampling plot. Quadrats were not sampled within 0.3 m of plot edges or previously sampled quadrats. Prior to GS30, complete plants including roots were placed in a sealed plastic bag to avoid moisture loss while in transit to the laboratory. After GS30, samples were cut at ground level. Samples were stored at 4 °C for between 0 and 4 h prior to processing. Subsequently, plants (prior to GS30 only) and shoots were counted, weighed and two subsamples (one representing 0.1 and another representing 0.2 of the original number of shoots) was taken, with the smaller sub-sample used for separation into either green or nongreen leaf, stem and spike (after GS55), along with dead/dying shoots. The absolute area of green leaf, stem and spike was determined using a WD3–WinDIAS Leaf Image Analysis System (Delta-T Devices, Cambridge, UK) and green area index (GAI) was calculated as the area of green material relative to the area of ground it was sampled from. Afterwards, all samples, including the second sub-sample, were dried in an oven with forced-air circulation at 70 °C for 48 h and weighed for dry matter (DM) determination, with the components of the separated sample analysed for N content using a FP 328 auto-analyser (LECO Corporation, St. Joseph, MI, USA). Prior to harvest, a quadrat (0.7 × 1.0 m) was sampled for the determination of yield components. The total sample was weighed and two sub-samples (0.2 and 0.4 of the original number of shoots) taken and the smaller sub-sample separated into straw and spikes. All samples were subsequently dried at 70 °C for 48 h, with the separated components analysed for N content using a FP 328 auto-analyser (LECO Corporation, St. Joseph, MI, USA). Spikes were subsequently hand threshed between two foam boards and separated into grain and chaff using a 1 mm slotted sieve (Glasblaserei, Institute of Fermentation and Biotechnology, Berlin, Germany). Grain number was determined using an automated grain counter (Pfeuffer GmbH, Kitzingen, Germany), after which grain weight was determined. Yield values were adjusted to 15% moisture, while grain number per m2 was calculated from the final spike numbers and grains per spike data.

kpar =

ln(I / IO ) GAI

where IO is the incident PAR and I is the PAR transmitted below the canopy. This value was subsequently used to estimate the proportion of PAR intercepted (PARint) by the canopy on sampling dates during senescence using the formula:

Proportion of PARint = 1 − e(GAI ).(−Kpar ) Daily estimates of the proportion of PARint were calculated by interpolating values between measured values and accumulated PARint was determined by applying these values to the recorded daily PAR. Radiation use efficiency (RUE; g DM/MJ) at each site was estimated for either the period between GS30 and GS92 or post GS59 as the gradient of the best fit line when biomass DM accumulation at each sampling date was regressed against the corresponding estimation of accumulated PARint.

2.5. Grain filling indices The duration of the period from GS59 to the end of grain filling was estimated by split-line regression between the recorded spike biomass (t DM/ha) and the accumulated thermal time (°Cdays), with the point of intersection considered the end of grain filling. The duration in terms of days and the PARint during grain filling was subsequently determined based on the estimated thermal time value. Rate of grain filling (RGF; g/°Cday) was calculated using the gradient of the initial line in the split line regression and the grains/m2 value for GS92, as no data on individual grain weight increases during grain filling was available. The stem biomass reduction (SBR; t DM/ha) was determined from the difference between stem biomass at GS92 and GS59.

42

European Journal of Agronomy 87 (2017) 40–49

J.P. Lynch et al.

grains/m2 appeared to have a closer relationship (Fig. 1). Indeed, data from the Crossnacreevy site in 2014 had the greatest residual to the estimated relationship between grains/m2 and yield, and when this data was excluded from analysis a significant relationship between these variables was observed across the other site-seasons (P = 0.001, R2 = 0.81, n = 8), while the relationship between individual grain weight and yield was much weaker (P = 0.307; R2 = 0.03). Results from the multiple linear regression analysis (Table 4) indicated that when grains/m2 and individual grain weight were considered together, a highly significant effect (P < 0.001; R2 = 0.89) on yield was observed across all evaluated site-seasons. Grain N content had a significant negative relationship with yield (P = 0.035; R2 = 0.48), while harvest index did not relate to yield (P = 0.439). The number of spikes/m2 tended to have a closer relationship with grains/m2 (P = 0.076; R2 = 0.34) than the number of grains per spike (P = 0.286; R2 = 0.05), although neither factor alone had a significant relationship with grain number. When both variables were considered together in a multiple linear regression, a highly significant relationship was observed (P = 0.002; R2 = 0.83, Table 4).

2.6. Statistical analysis Mean data from the three replicate plots of each site were calculated prior to subsequent analysis on the site-season mean values (n = 9). Standard deviations of the mean and the coefficient of variation were determined to investigate variability between site-seasons using Microsoft Excel 10 (Microsoft Corporation, New Mexico, USA). Simple linear and quadratic regressions were conducted to identify relationships between a range of measured variables (i.e. between yield and yield components, and between yield components and indices of growth and development), while multiple linear regressions were conducted to evaluate the influence of a combination of variables on indices of interest. In situations where relationships may be compromised by autocorrelation (i.e. grain number and spike number), separate calculations were made to determine variables based on the alternative sub-sample. All statistical analysis was conducted using the statistical software package Genstat version 14.1 (VSN Internation Ltd., Hemel Hempstead, UK). 3. Results

3.3. Factors contributing to grain numbers 3.1. Weather data Mean data for establishment and tillering indices are presented in Table 5. Variation between the observed crops was high (CV > 20%) for plants/m2, phyllochron, maximum shoots during the season per m2, thermal duration from sowing to the date of maximum shoots/m2, proportion of shoot survival and the number of spikes per plant at harvest. Variation for tiller number per m2 reduced as the crop progressed through development, from plants per m2 (CV = 29%), maximum shoot number per m2 (CV = 21%) and shoots per m2 at GS59 (CV = 12%). Large differentials between Crossnacreevy 2014 and Oak Park 2015 sites, compared to the other 7 sites, contributed considerably to the variation for plants/m2, maximum shoots/m2, proportion of shoot survival and spikes/plant. Across the nine site-seasons, the relationship between the number of spikes/m2 at harvest to either the number of plants/m2 at GS30 (P = 0.447) or the maximum number of shoots/m2 during the growing season (P = 0.382) was not significant, whereas the number of shoots/ m2 at GS59 had a strong relationship to final spike number (P = 0.008, R2 = 0.61; Fig. 2). No significant linear relationship (P > 0.16; Table 6) was observed between the number shoots/m2 at GS59 and either the maximum number of shoots/m2 during the growing season, the number of days or °C days accumulated during the sowing to GS30, or the GS30 to GS59 periods, the amount of PARint by the crop between GS30 and GS59, the biomass or GAI amount per shoot at GS30 or GS59, or N uptake during stem extension. A positive trend was observed (P = 0.087) across the nine site-seasons for a relationship between the number of plants/m2 and the number of shoots/m2 at GS59. There was a relatively strong negative relationship between shoot survival at GS59 and the maximum shoots/m2 (P = 0.015; R2 = 0.54), while the number of grains/spike did not significantly relate (P > 0.117) to the any of the indices of growth and development evaluated, with the exception of a significant negative linear relationship with the maximum number of shoots per m2 (P = 0.046; R2 = 0.25). No significant direct relationship was observed between any evaluated factor and grains/m2, although the positive relationship between shoot survival and grains/m2 was trending towards significance (P = 0.10).

Throughout the three seasons, Crossnacreevy consistently incurred the coolest average temperatures, while the Killeagh site consistently had the warmest mean temperatures for the 1 October-31 May period (Table 1). When compared to long-term averages, the 1 March-31 May period of the 2013 season was cooler (between 1.5 and 1.8 °C decrease) than average and the 1 June-31 August period of 2013 was warmer (between 0.8 and 1.2 °C increase) than the typical mean temperatures for that period. Variations in temperature from long term average during these periods did not exceed 1 °C in either 2014 or 2015. The incident PAR was highest during the 1 March-31 May period of 2015 than the other years, while the incident radiation between 1 June–31 August was greatest during 2013. The incident radiation at the Killeagh site was higher than the other sites throughout the three seasons. The incident PAR between 1 June-31 August was higher (between 56 and 102 MJ/m2 greater) than the 10-year average for all three seasons of the study at Oak Park, and the 2013 and 2014 season at Killeagh. No long term radiation data was available for comparison at Crossnacreevy. Annual rainfall from 1 September–31 August ranged from 616 to 1163 mm between site-seasons. The 1 October–28 February and 1 June–31 August periods of 2013 had less rainfall than the other years included in the study for all three sites. Across the three years, there were no consistent differences between sites for the total rainfall during the evaluated periods. During the 2013 season at Oak Park, the total rainfall between 1 October and 31 August was 154 mm less than the long-term average, with rainfall between 1 May–31 July 2013 being 84 mm (41%) less than the long-term average. When crop development stages were considered, no one site or season had consistently higher or lower mean daily temperature, total incident PAR, or total rainfall than others in the study throughout all phases of a growing season (Table 2), with considerable variation occurring between sites observed for some phases of development in each season. 3.2. Yield components

3.4. Factors contributing to individual grain weight Across the nine site-seasons, grain yield ranged between 10.7–15.8 t/ha at 15% moisture content, with a grand mean of 12.7 t/ha (Table 3). No consistent effect was observed for site or season for either yield or any of the yield components in the study. When regressed individually against observed yield, neither grains/m2 (P = 0.069; R2 = 0.31) or individual grain weight (P = 0.118; R2 = 0.21) had a significant relationship with grain yield, however

Crops grown at Crossnacreevy consistently had the lowest aboveground biomass at GS59, when compared to the other sites (Table 7). The duration of the period from GS59 to the end of grain filling was more variable when reported as days (CV = 12%) than when reported as thermal time (CV = 7%). The mean amount of PARint by the crops from GS59 to senescence was 350 MJ/m2, while the mean rate of grain 43

European Journal of Agronomy 87 (2017) 40–49

J.P. Lynch et al.

Table 2 Meteorological data during various stages of development of winter wheat crops grown at three locations between 2013 and 2015. 2013

2014

2015

C’creevy

Oak Park

Killeagh

C’creevy

Oak Park

Killeagh

C’creevy

Oak Park

Killeagh

Mean daily temperature (°C) Sowing − GS31 GS31-39 GS39-59 GS59-GS59+10d GS59+10d −GS59+20d GS59+20d −GS59+30d GS59+30d −GS59+40d GS59+40d − Harvest

5.1 11.8 12.5 13.3 17.1 17.8 16.0 14.6

5.8 10.9 13.2 14.3 16.5 19.7 18.5 16.4

7.2 10.1 12.6 13.8 15.4 18.3 18.3 16.1

6.8 10.7 13.4 14.7 13.4 14.8 17.2 14.1

6.8 10.8 12.6 13.3 15.6 14.8 14.5 18.0

8.1 11.2 12.2 13.3 16.3 15.6 15.3 17.8

5.5 9.7 13.5 14.3 13.1 12.2 13.6 13.9

6.7 9.5 11.1 14.0 16.4 15.2 12.5 13.8

7.5 10.1 11.8 14.6 15.3 14.5 14.1 13.5

Total incident PAR (MJ/m2) Sowing − GS31 GS31-39 GS39-59 GS59-GS59+10d GS59+10d −GS59+20d GS59+20d −GS59+30d GS59+30d −GS59+40d GS59+40d − Harvest

na 401 182 121 209 230 229 426

1244 275 417 172 206 220 166 213

1192 260 584 197 223 256 200 271

na 415 351 186 191 158 177 343

820 510 228 166 217 190 159 269

1109 456 263 227 253 188 162 274

1133 617 413 174 139 124 139 435

952 687 403 215 183 139 155 171

813 622 481 213 189 153 176 106

Total rainfall (mm) Sowing − GS31 GS31-39 GS39-59 GS59-GS59+10d GS59+10d −GS59+20d GS59+20d −GS59+30d GS59+30d −GS59+40d GS59+40d − Harvest

677 27 53 29 0 8 43 43

392 13 34 4 5 0 19 61

563 26 98 4 3 0 41 62

604 65 38 27 16 16 5 154

617 76 30 54 0 6 18 6

781 58 12 26 0 19 9 9

491 82 17 29 38 39 31 111

454 91 31 0 15 30 24 29

397 147 36 8 32 52 41 24

GS59+xd represents the number of days after GS59. na = not available.

fill (RGF) was 6.9 (x10−2) mg/°C day, with no consistent effect of site or season and a relatively high degree of variation (CV = 12%, 14% respectively). The RUE of crops from GS30–GS92 had a low variability (CV = 5%; mean = 2.9 g/MJ); however, when only the period from GS59 to the end of grain filling was considered, the RUE between crops was more variable (CV = 10%), with a mean value of 2.7 g/MJ. Fig. 3 and Table 8 report the relationship between the cumulative PAR intercepted after GS59 and the accumulation of biomass in the spike portion of the crop. The quadratic relationship between these factors had a significant fit (P < 0.05; R2 > 0.95) at all sites evaluated with the exception Killeagh 2013, which was close to significant (P = 0.06). While the estimates of coordinates were similar between Oak Park and Killeagh in all 3 seasons evaluated, the quadratic term for Crossnacreevy was lower than the other sites for all seasons,

and to a significant degree in 2013 and 2014. No significant individual linear relationship was observed between the individual grain weight and either the GAI or above-ground biomass at GS59, the thermal duration or total PAR intercepted between GS59 and the end of grain filling or the number of grains/spike or grains/m2 (Table 9).

4. Discussion All crops included in this study achieved yields greater than 10.5 t/ ha, and therefore represent high-yielding crops in the context of previous studies on limiting factors for winter wheat yield (Slafer and Savin, 1994; Calderini et al., 2006; Reynolds et al., 2008; Zhang et al., 2010; Lollato and Edwards, 2015), including studies in the more

Table 3 Yield components of nine winter wheat crops grown across Ireland between 2013 and 2015. Season

Site

Yield (t/ha)a

Grains/mb

IGW (g)b

Spikes/mb

Spike weight (g)

Grains/spike

Harvest Indexc

Grain N (g/kg)

2013

Crossnacreevy Oak Park Killeagh

15.8 10.7 15.0

30182 23635 29288

51.4 46.7 52.3

666 473 552

2.8 2.6 2.7

45.3 49.7 53.3

0.59 0.51 0.58

8.3 12.7 9.9

2014

Crossnacreevy Oak Park Killeagh

10.7 12.1 13.4

28587 21935 26269

41.7 55.0 53.6

547 525 600

2.3 2.9 2.5

52.2 42.1 45.6

0.57 0.49 0.46

11.3 10.9 na

2015

Crossnacreevy Oak Park Killeagh

11.9 11.6 13.4

23608 21869 25136

50.7 50.3 51.0

488 526 638

2.5 2.7 2.3

48.4 41.8 39.6

0.53 0.48 0.43

9.9 10.0 8.0

12.7 1.70 13

25612 2969.5 12

50.3 3.74 7

557 61.7 11

2.6 0.19 8

46.4 4.52 10

0.52 0.052 10

10.1 1.44 14

Grand Mean Std. Dev. CV% a b c

Yield represents the grain yield at 85% moisture from pre-harvest quadrat samples. IGW = individual grain weight at 85% DM. Harvest Index = g grain/g total biomass. Std. Dev. = standard deviation of the mean, CV% = coefficient of variation. na = not available.

44

European Journal of Agronomy 87 (2017) 40–49

J.P. Lynch et al.

Fig. 1. Relationship between grain yield (t/ha) and (A) grains/m2 and (B) individual grain weight for nine winter wheat crops grown across Ireland between 2013 and 2015. Table 4 Multiple linear regression coordinates for yield and yield components for data from nine winter wheat crops grown across Ireland from 2013 to 2015 (n = 9)/. Yield

Estimate

SEM

P-value

Intercept Grains/m2 TGW Regression

−15.92 0.000457 0.337 P < 0.001, R2 = 0.89, residual d.f. = 6

3.64 0.000070 0.0555

0.005 < 0.001 < 0.001

Intercept Spikes/m2 Grains/spike Regression

Grains/m2a Estimate −15303 42.9 359.6 P = 0.002, R2 = 0.83, residual d.f. = 6

SEM 6950.0 7.02 96.0

P-value 0.070 < 0.001 0.010 Fig. 2. Relationship between spikes/m2 at GS92 and shoots/m2 at GS59 for nine winter wheat crops grown across Ireland between 2013 and 2015.

a

For regression analysis using grains/spike as a response variable, grains/spike was calculated using yield and TGW data to obtain independent variables.

consistent interactions between climatic effects and crop development. Although variations in sowing dates may have reduced the ability to observe effects of climate, results for the study generally concurs with Holden and Brereton (2004), who reported that the regions included in the present study had relatively similar agroclimatic attributes. All crops monitored in the present study were above the average national yields achieved for each growing season (48% greater on average; CSO, 2016; DAERA, 2016), which reflected a low degree of

temperate climates of North Western or Northern Europe (Shearman et al., 2005; Peltonen-Sainio et al., 2007). This supports the findings of Burke et al. (2011) that the Irish climate is conducive to achieving amongst the highest yields of winter wheat globally. No consistent major effects of crop location were observed in the present study on indices of growth and development, despite contrasting latitudes in relation to the island of Ireland, which was likely due to a lack of

Table 5 Establishment and tillering indices of nine winter wheat crops grown across Ireland between 2013 and 2015. Season

Site

Plants/m2,a

Phyllochron (°C days)

Max Shoots/ m2,b

Sowing to max shoot/ m2,b Days

°C days

Shoots/m2 at GS59

Shoot survival (%)c

GS30 to GS59

Days

°C daysd

Spikes/plant at GS92

2013

Crossnacreevy Oak Park Killeagh

267 130 292

81 92 110

1172 788 878

190 189 118

998 1071 726

682 423 632

58 54 72

49 47 49

553 526 550

2.5 3.6 1.9

2014

Crossnacreevy Oak Park Killeagh

154 249 204

110 131 161

582 930 904

218 181 181

1657 1291 1271

570 551 613

98 59 68

63 57 50

707 640 560

3.8 2.1 2.9

2015

Crossnacreevy Oak Park Killeagh

167 334 207

88 121 127

920 1314 1016

158 154 172

913 1000 1128

573 615 659

62 47 65

62 62 62

669 616 620

2.9 1.6 3.1

221 65.1 29

131 23.5 21

945 198.9 21

173 26.5 15

1117 251.7 23

591 71.9 12

65 13.7 21

56 6.4 11

604 57.7 10

3 0.7 26

Grand Mean Std. Dev. CV% a b c d

Plants/m2 determined at GS30. Max shoots = the number of shoots/m2 at the time of highest shoots per plant. Shoot survival = 100 x shoots per m2/max shoots. °C days calculated using a base temperature of 0 °C. Std. Dev. = standard deviation of the mean, CV% = coefficient of variation.

45

European Journal of Agronomy 87 (2017) 40–49

J.P. Lynch et al.

Table 6 Relationship between indices of growth and development and the number of shoots/m2 at GS59, the percentage of shoots present at the stage of maximum tillers/plant that remain at GS59,number of grains per spike and grains/m2 for nine winter wheat crops grown across Ireland between 2013 and 2015. Shoots/m2 at GS59

Plants/m2 Maximum shoots/m2,a Shoot survival% at GS59b Duration from sowing − GS30 (days) Duration from GS30 − GS59 (days) °C days − sowing to GS30c °C days − GS30 to GS59c PARint/m2 − GS30-GS59d GAI/shoot − GS30 GAI/shoot − GS59 Biomass/shoot (g) − GS30 Biomass/shoot (g) − GS59 N uptake − GS30 to GS59 (kg/ha) a b c d

Shoot survival% at GS59b

Grains/m2

Grains/spike

P-value

R2

P-value

R2

P-value

R2

P-value

R2

0.087 0.159 0.823 0.489 0.652 0.612 0.894 0.786 0.694 0.334 0.854 0.495 0.924

0.27 0.16 < 0.01 < 0.01 < 0.01 < 0.01 < 0.01 < 0.01 < 0.01 < 0.01 < 0.01 < 0.01 < 0.01

0.253 0.015 – 0.601 0.547 0.911 0.175 0.475 0.847 0.663 0.998 0.224 0.459

0.06 0.54 – < 0.01 < 0.01 < 0.01 0.14 < 0.01 < 0.01 < 0.01 < 0.01 < 0.01 < 0.01

0.348 0.046 0.117 0.522 0.380 0.859 0.889 0.576 0.613 0.571 0.980 0.722 0.572

< 0.01 0.38 0.22 < 0.01 < 0.01 < 0.01 < 0.01 < 0.01 < 0.01 < 0.01 < 0.01 < 0.01 < 0.01

0.999 0.521 0.103 0.786 0.327 0.520 0.602 0.317 0.947 0.723 0.884 0.202 0.496

< 0.01 < 0.01 0.24 < 0.01 0.02 < 0.01 < 0.01 0.02 < 0.01 < 0.01 < 0.01 0.11 < 0.01

Maximum shoots = the number of shoots/m2 at the time of highest shoots per plant. Shoot survival = 100 x shoots per m2/max shoots. °C days calculated using a base temperature of 0 °C. PARint = photosynthetically active radiation intercepted by the canopy. ne = not estimated.

reflecting a sink limitation of yield for the majority of crops studied. Grain number is a product of the number of shoots that remain at crop maturity with grain-bearing spikes, and the number of grains per spike. Results from the present study indicated that spikes/m2 had a greater influence on grain number than grains/spike, with neither component having a strong direct effect when considered alone. Shearman et al. (2005) reported that increases in the grains/m2 of wheat varieties grown in the UK between 1972 and 1995 were primarily attributed to increases in spikes/m2 with little variation in grains/spike observed, while similar studies in warmer regions such as Argentina have reported significant variation in grains/spike among genotypes (Abbate et al., 1998; González et al., 2014). Analysis of a large dataset by Slafer et al. (2014) indicated that variations in grains/ m2 due to environmental conditions were primarily due to changes in spike/m2. The number of spikes at harvest depends on many developmental factors, including plant establishment, tillering dynamics and shoot survival to anthesis (Assuero and Tognetti, 2010). None of these factors alone had a significant relationship with final spike numbers in the current study, reflecting the impact that tillering and shoot survival

limitations from nutrient deficiencies, disease, weeds or pests on crop growth and development. When the large values of grains/m2 and individual grain weight reported in the present study are considered in the context of the relatively low incident PAR (as low as 437 MJ/m2 during the 30 day period after GS59), it can be concluded that environmental conditions of the study reflect a low source: sink ratio post-anthesis that has rarely been studied in field conditions previously, without direct manipulation of plant anatomy. As summarised by Borrás et al. (2004) and Slafer et al. (2014), previous studies have largely concluded that grain number is the dominant driver of cereal yield globally. However, some of these studies have also reported a potential co-limitation of yield by both sink and source in crops that have relatively low source: sink ratios (Calderini et al., 2006; Acreche and Slafer, 2009; González et al., 2014; Serrago et al., 2013; Asseng et al., 2017). Despite this, the results from the present study, conducted in a field environment with a relatively low source: sink ratio, concur with the consensus that variations in the yield of winter wheat are dictated by changes in grain number to a greater extent than changes in individual grain weight, therefore

Table 7 Indices of radiation interception and grain filling of nine winter wheat crops grown across Ireland between 2013 and 2015. Season

Site

GS59 to end of grain filling

GS30-GS92

GS59

RUEa

GAIb

Crop Biomassc

Days

°C daysd

PARinte

RUEa

RGFf

SBRg

2013

Crossnacreevy Oak Park Killeagh

3.0 2.8 3.0

8.2 5.1 7.4

10.6 11.4 17.2

45 37 40

712 616 631

403 359 304

2.6 2.5 2.5

5.8 6.1 6.4

1.9 2.7 5.4

2014

Crossnacreevy Oak Park Killeagh

3.0 2.8 3.0

6.3 6.8 6.6

9.2 12.4 14.4

51 49 48

770 753 728

388 444 321

2.7 2.9 3.1

5.6 8.0 7.6

0.6 1.6 1.7

2015

Crossnacreevy Oak Park Killeagh

2.7 2.9 3.3

6.1 6.5 7.0

12.2 15.6 17.1

54 51 54

724 757 764

340 326 323

3.1 2.2 2.6

7.3 8.4 6.6

2.6 4.8 2.0

2.9 0.16 5

6.7 0.81 12

13.4 2.72 20

48 5.8 12

717 53.6 7

356 43.7 12

2.7 0.28 10

6.9 0.94 14

2.6 1.5 57

Grand Mean Std. Dev. CV% a

RUE = average radiation use efficiency, g DM/MJ. GAI = green area index, m2/m2. Above-ground biomass, t DM/ha. d °C days calculated using a base temperature of 0 °C. e PARint = photosynthetically active radiation intercepted by the canopy. f RGF = rate of individual grain filling, x 10−2 mg/°Cday, estimated from total spike biomass accumulation. g SBR = stem biomass reduction, t DM/ha reduction in stem biomass between GS59 and the end of grain filling GS92. b c

46

European Journal of Agronomy 87 (2017) 40–49

J.P. Lynch et al.

Fig. 3. The spike biomass and photosynthetically active radiation intercepted after GS59, and the estimates of the quadratic relationship between them for winter wheat crops grown in ), Oak Park ( ) and Killeagh ( ) between 2013 and 2015. PARint = photosynthetically active radiation intercepted by the canopy. Crossnacreevy ( For Crossnacreevy 2014, the last data point collected was omitted from regression analysis as ear biomass reduced from the previous value, and therefore was likely recorded after grain filling had ceased. It is represented by . Table 8 Estimates of the quadratic term of the 2nd order polynomial fitted to data for spike biomass accumulation and photosynthetically active radiation intercepted (PARint)/m2 after GS59. Season

Site

Quadratic term

Table 9 Linear regression of individual grain weight and indices for crop, radiation interception and yield components for nine winter wheat crops grown across Ireland from 2013 to 2015 (n = 9). Individual grain weight (mg)

Quadratic relationship

Estimate (x106)

P-value

P-value

R2

SEM

2013

Crossnacreevy Oak Park Killeagh

1.7 17.9 15.3

0.275 0.002 0.149

< 0.001 < 0.001 0.06

0.97 0.99 0.99

0.81 0.10 0.49

2014

Crossnacreevy Oak Park Killeagh

4.8 28.9 17.8

0.076 0.007 0.001

< 0.001 0.001 < 0.001

0.95 0.95 0.98

0.94 1.02 0.71

2015

Crossnacreevy Oak Park Killeagh

7.3 13.5 13.2

0.001 0.011 0.057

< 0.001 < 0.001 0.010

0.99 0.99 0.98

0.47 0.52 0.58

GAI at GS59a Total Biomass DM at GS59b PARint GS59 to end of grain fillingc °C days − GS59 to end of grain fillingd Grains/spike Grains/m2 a b c d

P-value

R2

0.225 0.136 0.899 0.963 0.168 0.558

0.09 0.19 < 0.01 < 0.01 0.15 < 0.01

GAI = green area index, m2/m2. t/ha. PARint = photosynthetically active radiation intercepted by the canopy. °C days calculated using a base temperature of 0 °C.

SEM = standard error of the mean.

of the present study. However, the negative relationship observed between the maximum number of shoots/m2 during the growing season and shoot survival indicates that excessively high tillering did not benefit final spike number. This may be a consequence of density increasing competition per shoot, as light quality at the base of the crop affects tillering dynamics (Sparkes et al., 2006). However, tiller production and senescence is affected by a complex arrangement of many factors, including assimilate, temperature, water and nutrient availability (Assuero and Tognetti, 2010) and therefore a definitive explanation is beyond the scope of this study. The contribution that increases in fruiting efficiency (grains per g of spike) can make to grain number in wheat has previously been highlighted by Slafer et al. (2015) and Elia et al. (2016) as a potential breeding method to increase yields. However, despite considerable variation in the number of grains per spike between crops in the present

can have to compensate for reduced plant establishment in the NorthWest European climate. A similar lack of effect of changes in either plant number or the number of stems at GS30 has previously been reported by Whaley et al. (2000) and Berry et al. (2003), respectively, for winter wheat grown in the UK. This contrasts with findings for spring barley in Ireland reported by Kennedy et al. (2017), who found a strong requirement for good conditions and growth pre-GS31 to achieve high yields in Ireland, indicating that the sink-limitation of winter wheat may occur at a later stage of development than spring barley in a similar environment. Dreccer et al. (2013), reported that the growth rate per plant from GS10 to GS30 was related to the number of tillers present at GS30, which concurs with the increased tillering observed in some crops of low plant density in the relatively consistent environmental conditions

47

European Journal of Agronomy 87 (2017) 40–49

J.P. Lynch et al.

average daily photothermal quotient), while the 2014 crop incurred a 6% reduction in the average daily incident PAR (a 77% reduction in the average daily photothermal quotient). The period between spike emergence and anthesis has previously been shown to have a significant impact on potential grain weight (Calderini et al., 2001), through a determination of carpel size and endosperm cell number. Studies that have manually altered the source: sink ratio for wheat crops postanthesis, either through increases in source from the removal of grain sites (Acreche and Slafer, 2009; Ahmadi et al., 2009) or source reductions through shading (Asseng et al., 2017) generally have not observed significant changes in RUE or the rate of grain fill. Therefore, it can be hypothesised that the large reduction in the photothermal quotient at Crossnacreevy 2014, together with a relatively low RUE determined during a period of higher photothermal quotient and a low amount of stem reserves at anthesis, resulted in a deficit of post-anthesis assimilate for adequate grain-fill, and a source-sink co-limitation. While this effect is a consequence of a quite variable growing season, such climatic variation is generally not rare in some regions of North Western Europe, particularly Ireland (Sweeney, 2014).

study (40–53 grains/spike), a weaker relationship was observed between this variable and grains/m2 than spikes/m2. The high shoot densities and variability in shoot survival in the present study may have contributed to this effect, as a negative relationship between grains/ spike and the maximum number of shoots/m2 during the growing season was observed. Lázaro et al. (2010) and González et al. (2011) previously highlighted a positive relationship between growth of the developing spike and the number of grains per spike at harvest time, which indicates that the high shoot densities of some crops in the present study may have negatively impacted on grains/spike through a reduction in the growth rate per spike. Furthermore, previous studies by Thorne and Wood (1987) and Berry et al. (2003) have reported only negligible contributions of biomass from senesced shoots pre-anthesis to final crop biomass. Thus, excessive tillering followed by a high degree of shoot mortality would likely negatively affect both components of grain number in conditions similar to the Irish climate. Despite results indicating a general sink limitation of wheat yield across most of the evaluated sites, the crop evaluated at Crossnacreevy in 2014 contrasts this effect, which achieved only the joint lowest yield of the study despite a relatively high grain number. Although high rainfall towards the end of a relatively long grain filling period may have partially contributed to a lower individual grain weight through post-filling yield loss (Biddulph et al., 2008), it is clear from the results that this effect is due to a lower degree of grain filling (Fig. 3). However, this could reflect either a sink limitation, through a lower potential grain weight, or an assimilate shortage resulting in a lower degree of grain fill which would reflect a partial source limitation. Potential grain weight is influenced by the development phases of the spike, primarily pre-anthesis (Fischer, 2007; Xie et al., 2016), through environmental factors during carpel growth, (Calderni et al., 2001; Foulkes et al., 2011), or through the composition of grains on a smaller spike, with a higher proportion of distal grains reducing the mean grain weight (Acreche and Slafer, 2006; Beed et al., 2007). Therefore, a reduced rate of growth during this phase of development may promote further sink limitation of the crop by reducing the capacity for grain fill post-anthesis. However, this explanation seems unlikely for the Crossnacreevy 2014 site, as incident PAR during this phase was greater than the other evaluated sites in the same year, GAI at GS59 indicated the crop canopy was large enough to facilitate the interception of the majority of PAR, and the above-ground biomass per shoot at GS59, whilst lower than most other sites in the study, was similar to Crossnacreevy 2013, which achieved a 23% greater individual grain weight at harvest. To evaluate the potential of source limitation at Crossnacreevy 2014, a comparison with the 2013 season at the same site seems appropriate, due to a similar grain number and incident PAR for 30 d after GS59 (and therefore, similar source: sink ratios), but contrasting individual grain weight and yield. Both these crops had a considerably reduced rate of grain fill, in relation to both thermal time and PARint, when compared to the other sites. This may be a consequence of a reduced availability of stored carbohydrate reserves in the stem components of these crops, as indicated by a lower reduction in stem DM weight between GS59 and GS92 (between 1.0 and 3.5 t/ha less), in comparison to the other sites. Numerous previous studies have highlighted the potential contribution of stem reserves to grain filling (Gent, 1994; Asseng and van Herwaarden, 2003; Beed et al., 2007; Zhang et al., 2010; Serrago et al., 2013; Asseng et al., 2017), and the reduced availability of stem reserves likely increased the requirements on postanthesis assimilate production to sustain grain fill. However, the 2013 and 2014 seasons at Crossnacreevy differ in the dynamics of PAR incidence during the first 30 days post GS59. During the first 10 d post GS59 in 2014, 53% more PAR was incident for the crop than the same phase of development in 2013 (an 83% increase in the average daily photothermal quotient). Subsequently, between 10 d and 30 d post-GS59, the Crossnacreevy 2013 site incurred an 81% increase in the average daily incident PAR (a 26% increase in the

5. Conclusion Variations in grain yield of crops of winter wheat grown at different locations in Ireland in different seasons were primarily driven by variations in grain number, and therefore were generally sink-limited. A lack of relationships between sink size and plant establishment or tillering suggests that the period of ear development and anthesis is likely critical to maximising winter wheat yield in environments favourable to cereal crop growth such as North Western Europe. However, the potential for a sink-source co-limitation was observed in a situation of high sink size with low assimilate production during grain filling and insufficient stem reserves. These results indicate that in a similar high winter wheat yield climate, crops potentially could be somewhat susceptible to source limitation in particularly variable seasons. Acknowledgements The authors would like to acknowledge the technical assistance of Jim Grace, John Hogan, Stephen Collins, Teagasc farm staff, and Colin Garrett and the farm team at Crossnacreevy. Funding for this study was provided by the Department of Agriculture, Food and the Marine, Ireland. References Abbate, P.E., Andrade, F.H., Lázaro, L., Bariffi, J.H., Berardocco, H.G., Inza, V.H., Marturano, F., 1998. Grain yield increase in recent argentine wheat cultivars. Crop Sci. 38, 1203–1209. Acreche, M.M., Slafer, G.A., 2006. Grain weight response to increases in number of grains in wheat in a Mediterranean area. Field Crops Res. 98, 52–59. Acreche, M.M., Slafer, G.A., 2009. Grain weight: radiation interception and use efficiency as affected by sink-strength in Mediterranean wheats released from 1940 to 2005. Field Crops Res. 110, 98–105. Ahmadi, A., Joudi, M., Janmohammadi, M., 2009. Late defoliation and wheat yield: little evidence of post-anthesis source limitation. Field Crops Res. 113, 90–93. Alexandrator, N., Briunsma, J., 2012. World Agriculture Towards 2030/2050: The 2012 Revision. Food and Agriculture Organization of the United Nations, Rome. Asseng, S., van Herwaarden, A.F., 2003. Analysis of the benefits to wheat yield from assimilates stored prior to grain filling in a range of environments. Plant Soil 256, 217–229. Asseng, S., Kassie, B.T., Labra, M.H., Amador, C., Calderini, D.F., 2017. Simulating the impact of source-sink manipulations in wheat. Field Crops Res. 202, 47–56. Assuero, S.G., Tognetti, J.A., 2010. Tillering regulation by endogenous and environmental factors and its agricultural management. Am. J. Plant Sci. Biotechnol. 4, 35–48. Beed, F.D., Paveley, N.D., Sylvester-Bradley, R., 2007. Predictability of wheat growth and yield in light-limited conditions. J. Agric. Sci. 145, 63–79. Berry, P.M., Spink, J.H., Foulkes, M.J., Wade, A., 2003. Quantifying the contributions and losses of dry matter from non-surviving shoots in four cultivars of winter wheat. Field Crops Res. 80, 111–121.

48

European Journal of Agronomy 87 (2017) 40–49

J.P. Lynch et al.

Holden, N.M., Brereton, A.J., 2004. Definition of agroclimatic regions in Ireland using hydro-thermal and crop yield data. Agric. For. Meteorol. 122, 175–191. Kennedy, S.P., Bingham, I.J., Spink, J.H., 2017. Determinants of spring barley yield in a high-yield potential environment. J. Agric. Sci. 155, 60–80. Lázaro, L., Abbate, P.E., Cogliatti, D.H., Andrade, F.H., 2010. Relationship between yield, growth and spike weight in wheat under phosphorus deficiency and shading. J. Agric. Sci. 148, 83–93. Lollato, R.P., Edwards, J.T., 2015. Maximum attainable wheat yield and resource-use efficiency in the southern great plains. Crop Sci. 55, 2863–2876. McMaster, G.S., Wilhelm, W.W., 1997. Growing degree-days: one equation, two interpretations. Agr. For. Meteorol. 87, 291–300. O'Donovan, T., 2016. Crop Costs and Returns 2016. Teagasc. Peltonen-Sainio, P., Kangas, A., Salo, Y., Jauhiainen, L., 2007. Grain number dominates grain weight in temperate cereal yield determination: evidence based on 30 years of multi-location trials. Field Crops Res. 100, 179–188. Reynolds, M.P., Pietragalla, J., Setter, T.L., Condon, A.G., 2008. Source and sink traits that impact on wheat yield and biomass in high production environments. In: Proceedings of International Symposium on Wheat Yield Potential: Challenges to International Wheat Breeding. Mexico. pp. 136–147. Savin, R., Slafer, G.A., 1991. Shading effects on the yield of an Argentinian wheat cultivar. J. Agric. Sci. 116, 1–7. Serrago, R.A., Alzueta, I., Savin, R., Slafer, G.A., 2013. Understanding grain yield responses to source-sink ratios during grain filling in wheat and barley under contrasting environments. Field Crops Res. 150, 42–51. Shearman, V.J., Sylvester-Bradley, R., Scott, R.K., Foulkes, M.J., 2005. Physiological processes associated with wheat yield progress in the UK. Crop Sci. 45, 175–185. Slafer, G.A., Savin, R., 1994. Source—sink relationships and grain mass at different positions within the spike in wheat. Field Crops Res. 37, 39–49. Slafer, G.A., Savin, R., Sadras, V.O., 2014. Coarse and fine regulation of wheat yield components in response to genotype and environment. Field Crops Res. 157, 71–83. Slafer, G.A., Elia, M., Savin, R., Garcia, G.A., Terrile, I.I., Ferrante, A., Miralles, D.J., Gonzalez, F.G., 2015. Fruiting efficiency: an alternative trait to further rise wheat yield. Food Energy Secur. 4, 92–109. Sparkes, D.L., Holme, S.J., Gaju, O., 2006. Does light quality initiate tiller death in wheat? Europ. J. Agron. 24, 212–217. Sweeney, J., 2014. Regional weather and climates of the british Isles—Part 6: Ireland. Weather 69, 20–27. Thorne, G.N., Wood, D.W., 1987. The fate of carbon in dying tillers of winter wheat. J. Agric. Sci. 108, 515–522. Ugarte, C., Calderini, D.F., Slafer, G.A., 2007. Grain weight and grain number responsiveness to pre-anthesis temperature in wheat barley and triticale. Field Crops Res. 100, 240–248. Wall, D.P., Plunkett, M., 2016. Major and micro nutrient advice for productive agricultural crops, fourth ed. Teagasc, Wexford. Whaley, J.M., Sparkes, D.L., Foulkes, M.J., Spink, J.H., Semere, T., Scott, R.K., 2000. The physiological response of winter wheat to reductions in plant density. Ann. Appl. Biol. 137, 165–177. Xie, Q., Mayes, S., Sparkes, D.L., 2016. Preanthesis biomass accumulation of plant and plant organs defines yield components in wheat. Eur. J. Agron. 81, 15–26. Zadoks, J.C., Chang, T.T., Konzak, C.F., 1974. A decimal code for the growth stages of cereals. Weed Res. 14, 415–421. Zhang, H.P., Turner, N.C., Poole, M.L., 2010. Source-sink balance and manipulating sinksource relations of wheat indicate that the yield potential of wheat is sink-limited in high-rainfall zones. Crop Pasture Sci. 61, 852–861.

Biddulph, T.B., Plummer, J.A., Setter, T.L., Mares, D.J., 2008. Seasonal conditions influence dormancy and preharvest sprouting tolerance of wheat (Triticum aestivum L.) in the field. Field Crops Res. 107, 116–128. Bingham, I.J., Blake, J., Foulkes, M.J., Spink, J., 2007. Is barley yield in the UK sink limited?: I. Post-anthesis radiation interception, radiation-use efficiency and sourcesink balance. Field Crops Res. 101, 198–211. Borrás, L., Slafer, G.A., Otegui, M.E., 2004. Seed dry weight response to source-sink manipulations in wheat, maize and soybean: a quantitative reappraisal. Field Crops Res. 86, 131–146. Burke, J., Spink, J.H., Hackett, R., 2011. Wheat in the Republic of Ireland. In: Bonjean, A., Angus, W., van Ginkel, M. (Eds.), In The World Wheat Book: A History of Wheat Breeding. Lavoisier Publishing, Paris, France. CSO, 2016. Agriculture Area Used and Crop Production by Region, Type of Land Use and Year Central Statistics Office 2016. Available from http://www.cso.ie/px/pxeirestat/ Statire/SelectVarVal/Define.asp?maintable=AQA01&PLanguage=0 Calderini, D.F., Savin, R., Abeledo, L.G., Reynolds, M.P., Slafer, G.A., 2001. The importance of the period immediately preceding anthesis for grain weight determination in wheat. Euphytica 119, 199–204. Calderini, D.F., Reynolds, M.P., Slafer, G.A., 2006. Sourcesink effects on grain weight of bread wheat, durum wheat, and triticale at different locations. Aust. J. Agric. Res. 57, 227–233. DAERA, 2016. Farming Statistics: Provisional 2016 Cereal and Oilseed Rape Production Estimates. Department for Environment, Food and Rural Affairs, United Kingdom. London, UK. Dreccer, M.F., Chapman, S.C., Rattey, A.R., Neal, J., Song, Y., Christopher, J.T., Reynolds, M., 2013. Developmental and growth controls of tillering and water-soluble carbohydrate accumulation in contrasting wheat (Triticum aestivum L.) genotypes: can we dissect them? J. Exp. Bot. 64, 143–160. Ehdaie, B., Alloush, G.A., Waines, J.G., 2008. Genotypic variation in linear rate of grain growth and contribution of stem reserves to grain yield in wheat. Field Crops Res. 106, 34–43. Elia, M., Savin, R., Slafer, G.A., 2016. Fruiting efficiency in wheat: physiological aspects and genetic variation among modern cultivars. Field Crops Res. 191, 83–90. Fischer, R.A., Edmeades, G.O., 2010. Breeding and cereal yield progress. Crop Sci. 50, 85–98. Fischer, R.A., 1985. Number of kernels in wheat crops and the influence of solar radiation and temperature. J. Agric. Sci. 105, 447–461. Fischer, R.A., 2007. Understanding the physiological basis of yield potential in wheat. J. Agric. Sci. 145, 99–113. Fitzgerald, W.J., Fitzgerald, D.L., 2004. Meteorological measurements. In: Keane, T., Collins, J.F. (Eds.), In Climate, Weather and Irish Agriculture. Agmet, Dublin. Foulkes, M.J., Slafer, G.A., Davies, W.J., Berry, P.M., Sylvester-Bradley, R., Martre, P., Calderini, D.F., Griffiths, S., Reynolds, M.P., 2011. Raising yield potential of wheat. III. Optimizing partitioning to grain while maintaining lodging resistance. J. Exp. Bot. 62, 469–486. Gent, M.P.N., 1994. Photosynthate reserves during grain filling in winter wheat. Agron. J. 86, 159–167. González, F.G., Miralles, D.J., Slafer, G.A., 2011. Wheat floret survival as related to preanthesis spike growth. J. Exp.Bot. 62, 4889–4901. González, F.G., Aldabe, M.L., Terrile, I.I., Rondanini, D.P., 2014. Grain weight response to different postflowering. source:sink ratios in modern high-yielding argentinean wheats differing in spike fruiting efficiency. Crop Sci. 54, 297–309. Gonzalez-Navarro, O.E., Griffiths, S., Molero, G., Reynolds, M.P., Slafer, G.A., 2016. Variation in developmental patterns among elite wheat lines and relationships with yield: yield components and spike fertility. Field Crops Res. 196, 294–304.

49