ACO301 - Defra Final Project Report March 2008 (ANNEX 4)

ACO301 - Defra Final Project Report – March 2008 (ANNEX 4)

ANNEX 4

Impacts of future weather changes on the national sugar beet production and some aspects of sugar beet agronomy

The impact was assessed for the national total sugar production and the national unit land sugar yield using the UK 2007 beet crop size and its spatial distribution and with the daily weather generated for 150 years under weather scenarios of Baseline, 2020HI, 2050HI and 2080HI at ten representative weather stations. However, the impact was assessed for some important aspects of sugar beet agronomy using daily weather generated for weather station at Broom’s Barn. The main findings are:

• the national unit land sugar yield will increase due to beneficial effects of warmer spring and increased CO2 concentration, but the difference between the extreme low and the extreme high sugar yield will also increase in the future, mostly due to drought. This variance will pose a serious production chain problem
• drought stress related yield reductions will increase until 2050, but then remain at about the same level thereafter
• both seed drilling and seedling emergence date will be progressively earlier as the temperature gets warmer in the future
• the occurrence of freezing damage to roots in field will be less likely in the future and almost disappear by 2080
• powdery mildew and virus yellows diseases will have the potential be more severe in future, making sugar beet cultivation even more reliant on effective control measures
• beet cyst nematode will pose a serious problem as temperature gets warmer, due to faster completion of developmental life cycles
• the national unit land sugar yield is likely to increase but this will rely much more on effective means of disease control.

Introduction

Growth and yield of sugar beet depends primarily on the amount of solar radiation that the canopy can intercept and on the efficiency with which it is converted into dry matter and sugar (Werker and Jaggard, 1998). At present sugar beet is spring-sown. Unlike wheat, it is a vegetative crop and it accumulates yield (i.e. sugar) from an early stage. There is no particular growth stage that is especially sensitive to an extreme weather event. Again, this is in contrast to reproductive crops where, for example, high temperature at flowering might have a catastrophic impact on yield. The longer the beet crop growsand the faster it grows, the bigger the yield; adverse events at any time will cause yield losses proportional to the radiation resources that are not efficiently used. The risk of a complete crop failure is small and maintenance of plant health is needed throughout the season to achieve the potential yield.

Factors such as sowing date (or emergence date) and harvesting date that directly determine the duration of growth also affect sugar yield directly: the longer the growing season, the larger the yield. The one extreme event that can cause a catastrophic yield loss is freezing of the beet. This can happen in the field, prior to harvest, or in the storage clamp on the farm, where the beet should be insulated. The major diseases for sugar beet production in England include rhizomania, powdery mildew and virus yellows: the severity of the last two diseases is related to weather in winter. Beet cyst nematode and Cercospora are potentially serious diseases that might become more serious due to warming climates in the future.

This part of the project aims to assess the impacts of weather under the high CO2 emission scenarios in the of 2020 (2020HI), 2050 (2050HI) and 2080 (2080HI) time segments on the extremes of UK sugar production and sugar yield per hectare using ten representative sites and the sugar beet crop size and soil type distribution in 2007 and to assess the impact of these weather scenarios on the extremes of crop sowing and emergence dates. Weather data from Broom’s Barn has been used to assess likely changes in significance of the major crop diseases.

Methods

We chose 10 weather stations that are scattered in the current sugar beet growing areas in England (Fig. 1) For each weather station, 150 years of daily maximum and minimum temperature, rainfall and global radiation were generated for high CO2 climate scenarios (Semenov , 2007), representing four time slots from the baseline (1961-1990) to 2080. Daily potential evapotranspiration was calculated with the Priestly-Taylor (1972) method.. The CO2 concentrations were assigned values at 350, 440, 590 and 810ppm for the four time slots.

The latest version of the Broom’s Barn sugar beet growth model was used to simulate sugar yield (Qi et al., 2005). The model now distinguishes the effects of soil type on canopy development, intercepted radiation efficiency and partition of dry matter to sugar. Sugar yield under rain-fed conditions were simulated each year with ten sowing dates and a harvesting date fixed on 31 October, and for five soils differing in available water capacity (AWC). The five soils were sandy (AWC=11%), sandy loam (AWC=15%), clay loam (AWC=18%), silt loam (AWC=21%) and peat/organic loam (AWC=24%).

We used a database representing the 2007 beet crop, comprising the location, area and soil texture of every field. The impact of different future weather scenarios on the extremes of sugar yield was undertaken by linking each individual field to the nearest weather station. The sugar yields were therefore based on the weights of different weather stations and of the proportion of crop areas on different soil types

The effects of CO2 and temperature were incorporated as below (Richter et al., 2006):

RUE(T*CO2) = RUE * f[CO2] * fT(1)

where RUE is radiation use efficiency, which increases linearly by 30 % when the [CO2] is doubled:

f[CO2] = 0.3*([CO2] - 350) / 350 )+1(2)

and temperature effect on RUE is expressed as follows:

fT = 1 - 0.0025*(T - 18)2(3)

where T is the mean air temperature. We assumed that fT is equal to 1 when T is cooler than 18°C and equal to zero when T is at or above 38°C.

The impact of the future weather on the national sugar production and yield per unit area

Sugar yield is affected by various physical and biotic factors. The model simulations assume that crops are grown with sufficient plants and these have adequate nutrition and are free from diseases, pests, weeds. Sugar yield is strongly influenced by sowing date and harvesting date. At each weather station and for each of the 150 years, we simulated sugar yield for the 10 sowing dates after 1 January when the soil conditions and temperature thresholds. The harvesting date was fixed at 31 October, by which time most weather variations cease to affect yield. Then the mean sugar yield was calculated over the 10 sowing dates for each year for each individual field associated with one of the five soil texture types.

Analysis of variance showed that the median sugar yield per hectare on different soil types differed significantly among different weather stations for all climate scenarios (Table 1a-d). Therefore it was necessary to weight the effects of spatial variations in weather and soil to estimate average yields and total production by linking individual fields to their nearest weather station.

Table A4.1a The median sugar yield associated with different weather stations and the result for analysis of variance among different weather stations under different soil texture types for the baseline scenario.

Table A4.1b Median sugar yield associated with different weather station and the result for analysis of variance among different weather stations under different soil texture types for the 2020HI climate scenario.

Table A4.1c Median sugar yield associated with different weather stations and the results for analysis of variance among different weather stations under different soil texture types for the 2050HI climate scenario.

Table A4.1d Median sugar yield associated with different weather stationsand the results of analysis of variance among different weather stations under different soil texture types for the 2080HI climate scenario.

Table A4.2 shows the national total sugar production in the ranges of 5-, 50- and 95-percentiles if the sugar beet crop area of 118000 ha and its distribution in 2007 were maintained the same under different weather scenarios. Total sugar production will be increased by 15, 35 and 53% by 2020, 2050 and 2080, respectively if the crop area remains the same. However, the difference between the extreme low (5-percentile) and extreme high (95-percentile) production would increase to the extent that good seasons may produce twice as much sugar as poor seasons by 2080.

Table A4. 2 The 5-, 50- and 95-percentiles of total sugar production (Million tonnes Mt) and national unit land yield (t/ha) under different weather scenarios using sugar beet crop area of 118000 ha and spatial distribution in 2007 in England.

T

he increase in the national total sugar production has resulted from the national unit land sugar yield increase (Table 2). The sugar yield increases on all soil types in all future weather scenarios. However, the size of increases depends on soil type (Fig. 2a, Table 3). The yield should be expected to increase most on soils that have the greatest water holding capacity (e.g. silt loam). Yield variability increases similarly in all soils (Fig. 2b).

Figure A4.2 Impact of future climate change on sugar yield (t/ha) (a) and on the variation in yield for different soil types (b) in England.

Table A4.3 Changes in 5-, 50- and 95-percentiles of simulated national unit land sugar yield (t/ha) for rain-fed crops harvested on 31 Oct. under different weather scenarios

The impact of future weather on the yield loss due to drought.

The impact of future weather changes on sugar yield loss was assessed with weather data generated for Broom’s Barn. Rain-fed sugar yield was expressed as a percentage of the water-stress free yield in sandy, sandy loam and silt loam soils. The 5-, 50- and 95-percentiles of these yield losses are shown in Table 4. The median and 95-percentile losses will increase until 2050, when the median loss on all soil textures will exceed 20%. On sandy soils in hot dry years the 95-percentile losses might reach 60% by 2050 (Table 4).

Table A4.4 Changes in 5-, 50- and 95-percentiles of relative sugar yield reductions (%) attributed to drought stress under recent past and future weather scenarios

The impact of future weather on the sugar beet sowing and emergence dates.

The impacts of future weather changes on sugar beet drilling dates and seedling emergence dates were examined with data generated for the Broom’s Barn weather station. The base temperature for germination of sugar beet seeds is 3°C (Gummerson, 1986). Because this is so cool, the sowing date of the beet crop is mainly determined by soil conditions. Observations show that when soils used for beet production reach a moisture deficit of 3.5mm in the top 5 cm of soil, shallow cultivation and sowing can take place (Thomasson, 1982). We defined a potential seed drilling date when a machine workable date is followed by the air minimum temperature above 0°C for four or more consecutive days. Starting from 1 January each year, we used this model to select the first 10 dates when sowing was possible, and assumed that this was the number of work days every farmer would need to complete their sowing. The crop emergence date is defined in the model as the date after sowing when the accumulated air temperature above 3°C reaches 120°C days. From these ten seed drilling and crop emergence dates, the annual mean sowing and crop emergence dates were computed for each of the 150 years. Then the 5-, 50- and 95-percentiles were calculated among the 150 years for each climate scenario. The impact of future scenarios will increase the likelihood that sugar beet drilling date can be carried out earlier and seeds will also emerge progressively earlier as the time slices move further into the future (Table 4). However, the variability in the seed drilling and crop emergence dates would not be affected by the future weather scenarios (Table 5).

Table A4.5 Changes in 5-, 50- and 95-percentiles of sugar beet crop sowing and emergence dates under different weather scenarios using the weather station at Brooms’ Barn Crop Research Centre.

The impact of future weather on the likelihood of freezing damage to roots

The impact of the future weather changes on the occurrence of freezing damage to beet roots was analysed with data generated for the weather station at Broom’s Barn. It was reported that air minimum temperature below -5°C is the threshold for freeze damage to beet roots left in fields in December, January and February (Jaggard et al., 2006). Cold-spells of air minimum temperature below -5°C for two consecutive nights or more are computed in December to February for 150 years in each of four weather scenarios and have then been translated into probability of occurrence (Fig. 3). In the recent past, damaging cold spells at Broom’s Barn have occurred with a probability of about 45% for these months. As soon as the 2020’s, this probability will drop to 25% for 2-day spells and only 10% for 3 days. Strangely, the probabilities for the longer cold spells are the same today as they are predicted to be in the 2020’s. The reduced probability of 2-day freezes in future will improve prospects for leaving the roots in fields until required by the factory. This could drastically reduce storage losses and costs compared with current beet storing practices, using insulated clamps.

The impact of future weather on the extremes of diseases

Two of the important sugar beet diseases, powdery mildew and virus yellows, were assessed using the 150 years of daily weather generated at Broom’s Barn for all four periods. Models for powdery mildew (Asher and Williams, 1991) and virus yellows (Qi et al., 2004); both use temperatures in January-March at Broom’s Barn. The powdery mildew model also uses data on temperature and rainy days in April-August. At present, powdery mildew is usually a problem (Table 6) but is controlled well. Under future weather scenarios, powdery mildew incidence will be even more of a problem (Table 6) unless plant breeders can make the varieties much more resistant. In the absence of resistance, fungicide use on beet will become universal. The virus yellows model used here assumes the use of effective pesticides to control the aphid vectors of the virus. The disease is well controlled under present weather conditions (Table 6), but future weather changes will favour: over winter survival and reproduction of the aphids, the host plants and their viruses, and early flight of the aphids to new hosts in spring. Provided the efficacy of the insecticide systems can be maintained, this will not be too serious (Table 6). However, the spring in 2007 was so dry after sowing that the seed borne insecticides did not move into soil solution and thus were not taken up by the seedlings until well after the aphid infestations had arrived. In future, any increase in long dry spells could make these environmentally-friendly pesticide delivery systems redundant.

Table A4.6 Changes in 5-, 50- and 95-percentiles of disease incidences (%) for powdery mildew and virus yellows in sugar beet under recent past and future weather scenarios

Table A4.7 Changes in 5-, 50- and 95-percentiles of accumulated thermal time (°Cd) above a base temperature of 10°C after sugar beet crop emergence under recent past and future weather scenarios