View of Assessing the risks and uncertainties of regional crop potential under a changing climate in Finland

Assessing the risks and uncertainties of regional crop potential ^{under a} changing ^{climate in Finland}

Timothy R. Carter

1

^,RiittaA.Saarikko' and Kai J. Niemi

AgriculturalResearch Centreof^Finland, 'Officeaddress: FinnishMeteorological Institute, Box503,FIN-00101 Helsinki,Finland

Results arepresented ofamodelling study toestimate theregional suitabilityand potential produc- tivityof selected cropsinFinland underachanging climate. Model simulationswereconductedacross aregular 10 km gridoverFinland for various cultivars of thefollowingcrops: spring wheat, barley, oats,potato andmaize,and for two nematode pests andafungaldisease of potato. Modelswere run for both thepresent-day(1961-1990) climateand scenarios of future climate. Results are presented as maps. The main findings of the study are:(1) A warming ofthe climate induces ^{shifts in the}

northernlimitof cerealsuitabilityofsome 100-150 km per °C.⁽²⁾Changes inclimate and carbon

dioxide concentrationby 2050 are estimated to enhance average grain yields ofpresent-day barley cultivars inallregions. (3) Underprojected warming,the potential distribution of nematode species expands northwards and additional generationsof some species are likely.The risk^{of late} blight occurrence ^{increasesinallregions. (4)} By 2050 grainmaize could becultivatedreliably infavoura-

bleregions of^{southernFinland,and} satisfactory yieldsobtained. (5)Uncertaintiessurround all esti- mates,including uncertainties inprojections of future climate, model errors andassumptions and observational errors.

Keywords:suitability,cropyield, spring cereals, maize,potato, lateblight, nematodes, 2050scenario

ntroduction

It has been demonstrated in numerous studies that global climate change canhave significant impacts_on crop production in differentparts of the world (e.g. Parry 1990, Rosenzweig and Par- ry 1994,Reilly etal. 1996).In Finland ^too,re- search suggeststhat increasingtemperature and higher concentrations of carbon dioxide in the

atmospherecan lengthen the growingseasonand enhance the productive potential of field crops (e.g. Kettunenetal. 1988,Hakala and Mela ^1996, Kleemola and Karvonen 1996). These conclu-

sions _are largely based_{on two sources}of infor- mation: (i) experiments with plants grown un- der controlled conditions and (ii) simulations with mathematical crop growth models. Such findings tendtobe .site specific, and usually fail adequately to describe geographical variations

©Agricultural and Food ScienceinFinland ManuscriptreceivedFebruary 1996

in crop response. However, knowledge about possible changes in the regional _pattern of crop potential may be ofgreat valueto agricultural decision makers.

Several previous climate change studies in Europe have focusedon changes in crop suita- bility, using simple modelstodelimit regions in which future agroclimatic conditions would be appropriate for successful crop cultivation. Stud- ies have been reported of wheat (Kenny et al.

1993,Brignall et al. 1994, Bindi ^etal. 1993), maize(Carter etal. 1991

a,

Kenny and Harrison 1992

a,

Brignalletal. 1994);sunflower and soy- bean (Carteret

al.

1991b), grapevine (Kenny and Harrison

1992

b) and cauliflower (Kenny etal.

1993).Fewer studies have attemptedtoestimate regional crop productivity using mechanistic cropgrowth models. Some model site yields and attempttoscale up^tothe surrounding regionor country using qualitative criteria (e.g. Wolf 1993, Wolf and_vanDiepen 1995;and,inaglobal study, Rosenzweig and Iglesias 1994) orusing objec- tive interpolation and extrapolation (e.g. Wil- liamsetal. 1988 in Canada; Rötter and vanDi- epen 1994, Davies etal. 1993).A small number have_run simple crop growth modelsover areg- ular grid, using interpolated long-termmeancli- matological data as input (e.g. Leemans and Solomon 1993at globalscale, Jones and Carter

1993,Harrison ^etal. 1995).To our knowledge, no grid-based modelling studies have yet con- sidered both themeanand the interannual varia- bility of regional crop response to a changing climate. Variability of crop potential can be as importantasthe meanbecause it is_{a measure} of the risk and reliability of crop production (Mearns etal. 1992, Semenov and Porter 1995).

In this paper,we present results ofa studyto estimate the regional suitability and potential productivity of selected crops in Finland under achanging climate. We demonstrate,for the first time,the application of crop modelsover areg- ular gridtoevaluate changesnotonly in themean but also the variability of regional crop poten- tial. Inaddition, weprovideanillustration of the changingpatternof damage potential frompests and diseases. Some of the major uncertainties

of the approach are quantified and _a number of future research needs identified.

Methods

Geographical analysis system

The approach adopted in the study employs a national-scale geographical analysis system in conjunction with crop-climate models (Carter and Saarikko 1996). The models, which range from simple agroclimatic indices to complex crop growth simulation models, are run using data organised in anetwork of3827 grid boxes at 10 x 10 km resolution acrossFinland.

Environmental data at this resolution have been obtained forclimate,landcoverand_topog- raphy, but gridded soils data for Finlandare not available. Regional agricultural data^oncropped area and yields were also obtainedto map the current production pattern and to validate the indices and models. A geographical information system, IDRISI, is usedtocombine the data and display the results.

Climatological dataare for the baseline pe- riod 1961-1990, the standard reference used in the Finnish Research Programme on Climate Change(SILMU). Station data for each year of the period and for 30-yearmeanshave been in- terpolated^to the 10km grid by akriging method (Henttonen 1991) atthe Finnish Meteorological Institute (FMI). The data include monthly pre- cipitation and monthlymeansof^maximum,min- imum and mean temperature and global solar radiation. Observed values of radiation _were supplemented with values derived from obser- vations of cloudiness and sunshine duration.

In addition, daily data _were obtained from FMI for various climatic variablesatindividual meteorological stations in Finland. These data were used for model validation.An alternative source of daily climatological time series (for precipitation,meantemperature andcloudiness) was a stochastic weather generator, CLIGEN

(Posch 1992, 1994, Carteretal. 1995).The pa- rameters of CLIGENwerederived for all mete- orological stations reporting daily data in Fin- land during the period 1961-1990. Daily time series canbe generated for any point inFinland, by interpolation of the stationparameters.Daily solar radiation values are derived from cloudi- ness ^amounts using empirical relationships es- tablished from site data.

Model testing and application to the grid

The models are described in the next section.

Each modelwassubjected ^toconventionalsen- sitivity testing, in order to evaluate possible sources of model uncertainty. Where possible, model outputs were validated against observa- tions from experimentalsites,varietytrials, pest and disease observation networks and regional statisticalsources. Particular attentionwas paid to the performance and applicability of models acrossthe fullrange ofenvironmental conditions currently found in Finland, to justify applying the modelsto the grid. Moreover, models were also tested for climatic conditions outside the range experienced in Finland, to assess their applicability for simulating effects of climate change.

The approach used toconduct model simu- lationsoverthe 10km grid varied between mod- els. All models operate on adaily timestep, but their input datarequirements vary widely. Most of the models of crop and pest/disease develop- mentarebasedon measuresof accumulatedtem- perature, whichcan be evaluatedtoan accepta- ble accuracy using daily temperatures interpo- lated frommeanmonthly temperatures. Howev- er, several models are highly sensitive to the

within-month daily variability ofclimate, espe- cially precipitation. For these models, realistic daily datawererequiredateach grid box. Where soilparameters wererequired, in the absence of gridded soil data the_parameters of generic soil types were defined, and modelswere runfor each

soiltype across the entire grid.

Simulating ^a changed ^climate

Climate changes were simulatedby altering the baseline climatological data according to the SILMU climatic scenarios (Carter et al. 1995, Carter 1996). The basic scenario employed in all model simulations is SILMU policy scenario 1 (central“best guess”) for 2050.However, some model runs have also been conducted for the

range of scenario uncertainty, using SILMU pol- icy scenarios 2 (low) and 3 (high), and for time horizons of 2020 and 2100. These scenarios specify seasonal changes whichareuniformover the whole of Finland. In addition, severalruns have also explored variants of scenario 1, using the SILMU scientific scenarios la, lb and Ic.

These specify the regional patternof changeon amonthlybasis,basedonestimates from differ-

ent global climate models. All SILMU climatic scenarios _areaccompanied by consistentscenar- ios of carbon dioxideconcentration, which are required as inputs tosome yield models (Carter

1996).

In the simulationsreportedhere, the baseline climate ^at a future date has been adjusted ac- cording ^toeach scenario by applying thesame changes toall years of the baseline period. This approachtreatsthe future climateasifitwerein equilibrium. Inreality, ofcourse,the future cli-

mateis unlikelytobe in equilibrium, but will be changing continually.However,exploratory crop model simulations indicate that removing the trend from the scenario climate doesnotsignif- icantly alter the crop response.

Models

Five different modelsorsetsofmodels have been employed in the study. Theyaredescribed briefly here,but allaredocumented in detail elsewhere.

Their selection should^not imply that they are any more appropriate for this application than other models^- it merely indicates that amodel was available for this study, and sufficient data

could be obtained for model testing. Other mod- els will be applied ^tothe gridatalater date.

Cereal development models

Phenological models have been constructed to estimate the growing time and phasic develop- ment of spring-sown cereal cultivars ^(Saarikko and Carter 1996). Relationships wereexamined between environmental factors (temperature, precipitation and photoperiod) and phenological observations after sowing (heading and yellow ripening) ^at experimental stations in Finland during the period 1970-1990. Temperature alone wasfound toexplain thecourse of phenological development, and a linear relationship was es- tablished between daily mean air temperature and daily development rate for all phases and cultivars (Figure 1).Modelswereconstructed for three cultivars of wheat ⁽Triticumaestivum), of barley (Hordeum vulgare⁾ and ofoats (Arena saliva)for the phases sowing toheading, head- ing^toyellow ripening and the entire phase sow- ing toyellow ripening.

In orderto apply the modelto the grid, the beginning and end ofa“favourable growing pe- riod” need^tobe defined. Sowing of spring^cere-

als was assumed^totake place on the day when smoothed dailymean temperatureexceedsB°C, based on sowing date information from theex- perimental sites. The end of the favourable grow- ing period is defined by the dateon which mean daily temperature falls below 12°C. This date approximately correspondstothe25% probabil- ity of first autumn frost occurrence in Finland ascomputed by Solantie (1987).

When estimating crop development across the Finnish grid, smoothed daily mean air ^tem- peratures were obtained from monthly means using the Brooks sine _curve interpolation meth- od (Brooks 1943). The development ofeach cul- tivar_wascomputed for individual years during the 30 year baseline period and during 30 year periods corresponding^toarange of SILMUsce- narios.

Barley yield simulation model

Barley is grown widely inFinland, accounting for about 20% of the cultivated area in 1993 (Agricultural Information Centre 1994).Barley yields have been simulated using_aprocess-based cropgrowthmodel,CropWatN,developedatthe Department of Plant Production, University of Helsinki (Karvonen and Kleemola 1995). The model calculates crop growth and development on a daily basis as a function of global radia- tion, mean air temperature and precipitation.

Simulations canbe conducted either for poten-

tial productionorfor soilwaterand nitrogen lim- ited production. The direct effects of increasing atmospheric C0₂concentrationon plant photo- synthesisare also computed. The main features of the model are described by Kleemola and Karvonen(1996),the only differencebeing aless detailed soil temperature scheme in the version used here than in their updated model.

The modelwascalibrated using detailed ob- servations ofclimate, soils and the growth and development of barley (cv. Porno) from experi- ments at Jokioinen during 1982-1987 (Ilola et al. 1988). Values of more than 30 parameters wererequired, and many of theseweresite spe- Fig. 1.Relationshipbetweenmeandevelopmentrate(day'1⁾

and meantemperature (°C) inspringwheat (cv. Ruso) for thephase sowingtoyellow ripening. Broken linesare95%

confidence limits. (Datasource:Officialvariety trials,^Ag- ricultural Research Centre of Finland)

cific, since they related tosoil properties. A pri- or sensitivity analysis indicated that modelled yields could be highly sensitive to parameter values relating tothe movementofwaterin the soil.

To testwhether the model could be applied toother conditions inFinland, dataon phenolo- gy, fertilizer application, soil type and yields were obtained for official variety trialsatfour experimental stations in differentparts of Fin- land during the period 1970-1990. Sincemeas- ured soil characteristics werenot available, pa-

rametervalues for the fine sand, sand loam and silty clay typesreported from the siteswere es- timated basedonpublished material(J.Kleemo- la,personalcommunication). The modelwas run using weather data from each site.

A comparison of simulated against observed grain yields^atthe four sites is given in Figure 2, classified by soiltype.There is a positive rela- tionship between the^two(r⁼0.48)although the scatteris quite large. Some of the discrepancies between the modelled and observed valuesare clearly systematic and related to soil type.This isnotsurprising, given that the precisenatureof the soils ateach site _{was not}known. These _re- sults_wereencouraging enough, however, tojus- tify applying the modelatnational scale.

In orderto apply the model_to the grid, two

soiltypes were selected,representing _arange of

waterholding characteristics: aheavy claytype,

using the measuredparameters fromJokioinen, and the fine sandtype, using the_sameparame- ter estimates used in the validation exercise.

Parameters for crop phenology were thosere- ported for the Porno cultivar by Saarikko and Carter(1996).

Daily temperature and radiation data were obtained from monthly data using the Brooks interpolation technique (Brooks 1943).The ef- fect of using smoothedtemperature and radia- tion datawas evaluated in anearlier sensitivity analysis, and found toproduce anoverestimate of final yield compared with using observed dai- ly values of about 19%. This is ^tobe expected since daily extremes oftemperature and radia- tion, which inhibit growth, are notrepresented in the smoothed data.

Gridded monthly precipitation was allocat- ed between days according^tothe frequency dis- tribution of precipitation observed atJokioinen during 1961-1990. Given the high sensitivity of the modelled yield tosoil water, this technique produced themostrealistic precipitation distri- bution of several methods tested.

Modelruns were conducted for the30 indi- vidual years of the 1961-1990 baseline and for a30-year scenario climate representing 2050,us- ing SILMU scenario 1. The modelwas runatall grid boxes for unstressed (potential) conditions and for rainfed conditions on both soil types.

Simulations were run for scenariotemperature and precipitation changes^alone,for elevated C0₂ concentration(523 ppm) alone,and for these_two conditions combined.

Maize yield simulation model

Maize(Zeamays) is atropical crop occasional- ly grown for silage and sweet corn in Finland.

Trials conducted in the 1970

s

in southern Fin- land suggested that with very early hybrid vari- eties _a mature grain yield could be harvested twice in ten years,good quality silage material obtained six years intenand asatisfactory crop eight years in ^ten(Pullietal. 1979).Later work, atEuropeanscale,indicated that climaticwarm- Fig. 2.^Simulatedvsobservedgrain yield in barley for three

soil typesat toursitesinselected yearsduring 1970-1990.

Modelled phenology is fixed at observed dates (Datasource:

Officialvarietytrials.AgriculturalResearch Centre ofFin-

land)

ing of 3-4°C would have the potential to shift the northern limit for economic yields of grain maize from northern Germany and central Po- land, where it lies today, into southern Finland (Carter etal. 1991

a,

Kenny and Harrison

1992

^a).

The CERES-Maize model(Jonesand Kiniry 1986)has been adopted in this study. It has been used in a number of previous climate change studies in differentpartsof the world (e.g. Rosen- zweig and Iglesias 1994). Like CropWatN, CERES-Maize attempts to simulate the main processes of crop development and growth. It requires daily minimum and maximumtemper- ature, global radiation and precipitation as in- puts. A key requirement for running the model is the selection ofparameterscharacterising the cropphenology and yield capacity. Thesewere estimated for _a short-season hybrid, based on information given by Pullietal.(1979).

Model simulations with the CERES-Maize modelwere exploratory, and the only runs con- ducted to date have been for potential yields.

They arethus independent of soil type and as- sume no wateror nutrient ^stress. Gridded data onmonthly climatewereconverted todaily val- uesin thesamewayasfor the barley model. Sim- ulations were conducted for each year of the

1961-1990 baseline climate and for the climate estimated for 2050 under SILMU scenario I.

Effects of C0₂increaseoncropgrowthwere^not modelled.

Nematode models

Simple models of^twopestsofpotato⁽Solanum tuberosum) have been applied tothe grid. One, thepotato cyst nematode,is _atroublesomepest atthe present-day. Theother,the Columbiaroot- knotnematode, couldrepresent aserious threat in the future (Tiilikkala etal. 1995).

Thepotato cystnematode{Globodera rosto- chiensis), is the most noxious pest in present-

day Finnish potato production. This parasite is restricted in its range by temperature to south- ern and central Finland (Tiilikkala 1991). Its absence from northern Finland is avaluableas-

setfor the production of high quality seedpota- toes.The northern limit of the nematode approx- imately coincides withaneffectivetemperature sum(ETS) of 800 day-degrees aboveabasetem- perature of5°C, cumulated during the period with air temperatures exceeding 9°C (Tiilikka- la,personal communication).

The Columbiaroot-knot nematode(Meloido- gyne chitwoodi^), is a newly discoverednema- tode pest in Europe (Tiilikkala et al. 1995). It hasawide host range and is well adapted ^tolow temperatures. The nematode is aseriouspestof potato in the Pacific Northwest of theUSA, but has only recently been discovered in Europe, in the Netherlands. M. Chitwoodi can overwinter as eggs and juveniles. Thispest, unlike G. _ros- tochiensis, canproduce multiple generations, the number dependent upon the soiltemperature.An extra generation, especially late in the season, will result in atremendous increase in popula- tion densities.

The model used here is again basedoneffec- tivetemperature sum above abase temperature of 5°C,but cumulated in this case during the period with airtemperaturesexceeding 5°C. The approximate day-degree requirements for differ- entgenerations have been defined in a^pest risk assessment by Tiilikkala etal. (1995).

Both nematode indices have been computed acrossthe Finnish grid, using dailytemperatures interpolated from the monthly means (as de- scribed above) for each of the 30 years of the baseline climate and for the baseline values adjust- ed according^to anumber of SILMU scenarios.

Potato late blight model

Potato late blight ⁽Phytophthora infestans) is a damaging fungal pathogen that grows in the leaves and stemsof thepotatoplant. It progres- sively kills the tissues and reduces the effective photosyntheticareaof the plant. As a result, lit- tle assimilate is passedtothe tubers and yield is reduced. The spores aredispersed by rain, and can initiatenewinfections onleavesor,if washed into thesoil, onthe developing tubers.

The onset of late blight has been closely monitored inrecent years ^at seven sites in Fin- land, as part of_a Nordic observation network established in 1992. Results from this monitor- ing network have been used^todevelopasimple index of blightonset basedon a daily tempera- ture sum cumulatedon wet days (Kaukoranta

1996).This index has been coupled to a tuber growth model simplified from themoredetailed model of MacKerron and Waister (1985). The model computes potential daily tuber growth based on radiation and temperature, assuming thatwater and nutrients are not limiting. The coupled model is firstrun through toharvest to compute the potential tuber yield assuming no blightsymptoms.Subsequently, the model isrun again, this time calculating the date ofonset of blightsymptoms,whereupon the leafareaindex is reduced to zero over atwo week period, thus curtailing tuber growth. The difference between the accumulated tuber growth from the date of first _symptoms until harvest and the blight-af- fected tuber growthrepresents the potential tu- ber loss duetothe disease. More details of the site validation of this model for Finnish condi- tionsaregiven by Kaukoranta (1996).

A key requirement in applying the modelto the grid was for daily temperature and precipi- tationdata,todefine theonsetof blight. Realis- tic daily values could not be obtained by inter- polating the gridded mean monthly values, so instead the stochastic weather generator, CLI- GEN, was used to produce daily climatic time series.

To test the validity of using CLIGEN, the modelwas first tested for individual _meteoro- logical stations in Finland to compare results based on the stochastically generated weather with those based _on observed weather during 1961-1990. The analysis revealed that the weath- er generatorunderestimates the inter-annualvar- iability ofclimate_{on a}seasonal basis because it fails to replicate accurately the persistence of weather^eventssuchas warmspells and droughts.

As a result, while themean of modelled tuber loss_was similar in bothcases,the variance based ongenerated weatherwaslower than that simu-

lated for the observed conditions.Thus, results from the grid should be interpretedas conserva- tive estimates of tuber loss^-the actual variabil- ity of losses is likely_tobegreaterthan depicted.

Theparameters of CLIGENwereinterpolat- edto each grid box location, and 30-year time series of baseline and scenario climateweregen- erated for all boxes across the grid. All three SILMU policy scenarios were simulated for 2020,2050 and 2100.

Results

Given ^the numerousmodels applied^tothe grid system,and the multiplicity of scenariosexam- ined,it is only possible in this papertoprovide a short synthesis of the major results obtained.

More detailed findingsarepublished elsewhere (e.g. Carter and Saarikko 1995, 1996).

At the outset, it should be stressed that the mapped results obtained in this study arepre- liminary and aredesignedtobe illustrative. The mapsdepict _patterns of crop potential thatare climaticallydetermined, and donot account for factors such as the local soils, land cover and terrain. Over large regions these factors would preclude crop production altogether. Methods of identifying and excluding such regions from analysis_arecurrently being addressed.

Regional suitability of spring cereals

Three aspects of the regional suitability force- reals under changing climatearesummarised in this section: changes in crop risk and reliability, shifts in ^patterns of suitability and changes in rates of phenological development. Uncertain- ties associated with all of these are considered in alater section.

Changesin croprisk and reliability. As cli- matewarms, sothe probability of modelled crops reaching the yellow ripeningstage increases. In

Fig. 3.(a) Actual cultivatedareaofspringwheatin 1990as a percentageof total arable (source: Finnish Board of Agricul- ture); (b and c) estimatedprobabilityof successfulripening (percent) forcv.Ruso under thebaseline, 1961-1990climate^(b) and SILMU scenario 1(best guess)by 2050^(c).

cereals, the importance of reaching this stage dependson the finaluse of the crop. Forexam- ple, reliable attainment of full maturity is more important on farms where good quality cereal productionrepresents amajorsourceof_income, thanon farms where it merely supplements oth- eractivities.

The estimated regional probability of ripen- ing for an early-maturing spring wheat cultivar (cv. Ruso) is shown in Figure 3b for the base- line climate. As might be expected, the reliabil- ity of wheat cropping is greatest in the warmer south of Finland and declines northwards. This mapcan be compared ^to the actual pattern of cultivation for all spring wheat varieties in 1990 (Figure 3a), which is representative of theaver- age present-day distribution. The northernmost cultivation areas are approximately coincident with the 60% probability limit, while the zone ofmostintensive cultivation (>lO%of totalar- able ^area) coincides with areliability well in excess of^80%. For barley and ^oats, the corre- sponding probabilities are lower, probably re- flecting the lower dependence _on ripening for these crops.

Figure 3c illustrates the corresponding prob- ability zones mapped for the climate in 2050 under SILMU scenario 1 (mean annual _warm- ing of 2.4°C). Here, the zone of high (>80%) reliability extends into central Finland and there is approximately an even chance of successful ripening in southern Lapland.

Shifts

^{in thepattern}

of

suitability. We have adopted the 80% probability level todefine the limit of suitability for all cereal species inour subsequent analyses, assuming that farmers would be willing ^{to accept}acrop "failure" in 2 yearsoutof10. Shifts in suitability for each crop have been mapped for 2020, 2050 and 2100un- der SILMU scenario 1, relative^to the baseline.

This is illustrated forbarley (cv. Porno) in Fig- ure4. The geographical extent of these shifts is summarised in Table 1 for acultivarofeach crop.

The shiftscan also be interpretedas rates, ^which are shownasaveragesatthe bottom of Table

1.

These imply that under the best guess SILMU scenario the northern limit of reliable cereal cul- tivation in Finland would shift northwards by, on average, about 45-60 km per decade up to 2100.

Changes in the ^rate

of

phenological devel-

opment. Higher temperatures increase therate of phenological development ofcereal crops ^(cf.

Figure I) so that plants can complete a given phase in_ashorter period. Figure 4 demonstrates how the phase heading ^to yellow ripening in barley^(cv. Porno) progressively shortens_asthe climate warms under SILMU scenario 1. The duration of this phase is of particularinterest, as it contains the grain filling period. A marked foreshortening of the phase would be expected to reduce grain yields, an effect confirmed in simulations with the barley growth model, CropWatN(see below).

TableI.Estimated shiftsinthe northern limit ofsuitabilityrelative to thebaseline for cereal cultivars under SILMUscenario 1(best guess) by2020,2050and2100,and themean^rateof shift up to2100(km/decade).

Shiftsunder SILMU Springwheat Spring barley Oats

scenario 1 ^{(cv. Ruso) (cv. Aira) (cv. Veli)}

24°E 29°E 24°E 29°E 24°E 29°E

2020(km) 150 150 90 290 130 280

2050(km) 270 460 230 340 280 500

2100(km) 550 640 490 620 550 630

Mean rate to2100

(km/decade) 50 58 45 56 50 57

Fig. 4.^Simulatedchange induration of thephaseheading to yellowripening inbarley (cv. Pomo) relative to the 1961-1990 baseline for SILMU scenario 1 (best-guess) by ^(a)2020,(b) 2050and (c)2100. Unitsare days and areas ofexpanded suitability^arealso shown.

Changes ⁱⁿ regional ^cereal yields: ^the

case of barley

The results presented in this sectionarefor bar- ley. However, given the similarities of growth habit between spring-sown barley, wheat and

oatsin ^Finland,the regionalpattern of response toclimate change shown here can be regarded asbroadly indicative of responses for all three cereals ^- only the specific location and details of changes differ.

Simulations with the CropWatN model used the phenological development model for barley (cv. Porno) mapped above and similar methods tocomputesowing and harvest dates.^Here, the modelled estimates of dry matter yield within thezones of suitability(i.e. 80% reliability) are reported for both stressed and unstressed crops.

Estimates of the30-yearmeanpotential grain yield for the baseline periodareabout twice those reported from agricultural districts in annualsta- tistical yearbooks, which is notsurprising given that nutrients andwaterwereunlimiting in these runs. Simulated baseline yields for afine sand soil _were close to the potential values, an unre- alistic result which is probably attributable to

inappropriate selection of soilparameters asde- scribed earlier. In^contrast,themean yields esti- mated for _aheavy clay soil (Figure sa), includ- ing ^water limitations and assuming an annual nitrogen fertilizer application of 100 kg/ha, are much closer in magnitude tothose reported op- erationally in Finland. Note, however,that since a uniform heavy clay soil is assumed over the wholecountry,the regionalpatternof modelled yieldscannotbe compared with the actual yields, on arange ofsoils,reported from differentparts

of thecountry.Thepattern depicted indicates the highest yields in east-central Finland, declining to the north, mainly as a function of low tem- perature, and towards the western coast, where the mean growing season soil waterdeficit is

greatest.

To compare the relative effects of increasing C0₂concentration and of changing climateac- cordingtoSILMU scenario I by 2050, bothwere analysed separately and then in combination.

Figure 5b shows the modelled mean yield for temperature and precipitation changes alone.

Climate change alone hasa negative effect on the yield of the Porno barley cultivar over the whole of southern Finland (compare with Fig- ure sa).This is primarily due to the shortening of the grain filling phase under the increasedtem- peratures (cf.Figure 4),although increased pre- cipitation (by 6% in the_summer months under this scenario) servesto modify this effect. Fur- thernorth, however,the highertemperatures are beneficial for this variety of barley, enhancing yields and shifting the northern limit of reliabil- ity some 300 km northwards.

The yield response ^to elevated CO, (523 ppm) alone,assuming nochange ⁱⁿthe^climate, is positive ^(lessthan 1 t/ha) with little variation across thecountry ^(not shown here). The com- bined effects of increased CO, and climate change are shown in Figure

sc.

Yield responses arepositive across the whole country, with the beneficial effect of increasingCO, slightlyout- weighing the negative effect of higher_tempera- turesin southern Finland. The pattern of chang- es for both the potential and fine sand simula- tionsare very similar ^tothose for clay, but the magnitude of changes is proportionally greater

since crops experience littleor nowaterandnu- trientstress in these simulations.

Estimates of the inter-annual variation of yields were also obtained. The standard devia- tion of modelled annual yields on heavy clay under the present-day climateranges from about 1 t/ha in the south of Finlandto approximately 2 t/haatthe northern limit ofreliability. This de- creases under the 2050 scenario over much of southern and central Finland and increases slight- ly in theextreme south_aswell_asin_some north- ern regions whichwere unsuitable under the baseline climate.

Changes in _pest and disease risk of potato

Potato cultivation in Finland extends north of the Arctic Circle, but yields areconstrained by the short growing season as well as occasional drought during the period of growth. Potato sel-

Fig. 5.Modelled grain yield of barley (cv. Pomo)on a heavy claysoil (t/ha) under (a) the baselineclimate, 1961-1990, (b) SILMU scenario 1(best guess)by 2050for climate (temperature andprecipitation) change only,and (c) SILMU scenario

1for climate and CO, changes combined.

dom reaches physiological maturity anywhere in Finland under the present-day climate,its growth usually curtailed in the autumnby frost or sur- plus soil water. Results from the coupled late blight/potential tuber growth model indicate that for the warming by 2050 under SILMU scenario 1 (central), the crop would reliably reach full maturity in the south of Finland before lowtem- peraturesbecame arisk for harvesting.For this reason, in spite of _an earlier sowing date, the growing period would actually shorten by about 10 days for present-day cultivars of potato. In contrast,the growing period would extend by 3 4 weeks in the north of thecountry. Under this scenario, potential dry ^matter tuber yields in- crease by some 3-4 t/ha (about 20-30%) in southern Finland and by morethan 5 t/ha insome central and northernareas, where little growth is possible under the baseline climate. Howev- er,soil moisture and C0₂ effects still needtobe incorporated in the model beforea morerealis- ticpattern of response can be obtained.

While the prospects forpotato yields appear favourable under a warming, the potential for

pestand disease damage of the crop also increas- es. Under SILMU scenario 1 ^(2050), the north- ern limit of thepotato cyst nematode(themedi- an limit during the 30 years ^simulated) extends on average by about 250 km northwards from its location under the baseline climate, occupy- ing many of theareasofcurrentseed production in Lapland (Figure 6). Under SILMU scenarios 2 and 3(lowand high), northward shifts by 2050 average about 50km and 400km,respectively.

The effect of future warming on the poten-

tial development of the Columbian root-knot nematode, were it tobe introduced in Finland, is no less significant. The model results ^suggest that under the baseline climate this species would be capable of producingonefull generation and hatching a second over much of southern and central Finland (Figure 7a).In_warmer years such as 1975 and 1988, it could complete a second generation and hatchathird in the ^mostfavour- Fig. 6.Potential distribution of the potato cyst nematode (Globodera rostochiensis) based ontemperature (median limit) under the baseline climate, 1961-1990,and its extension by 2050 under three alternative future climates: SILMU low, central (best guess) and high policy scenarios.

able regions. Under the scenario I warming by 2050, the nematode would be capable inan av- erage year of producingtwofull generations and hatchingathird in southern Finland (Figure 7b), with the possibility in warmeryears ofcomplet- inga third and hatchingafourth generationata few locations. The damage implications of such changesareself-evident.

Finally, the results of the simulations of po- tato late blight also indicateanincreased risk of occurrence and damage potential forunprotect- ed crops. Figure 8a shows the estimated tuber loss under the baseline climate. This isgreatest

in the south ofFinland, where yields arehighest

and where blight symptoms, and hence ^losses, are estimatedto occur in mostor all years(not shown). Losses arenegligible north of the Arc- tic Circle.

Under SILMU scenario 1for 2050, the date ofonset of blight symptoms is earlier than for the baseline by some 20 days in the south and 30 days in the north.^In the absence of crop pro- tection, losses wouldoccur in every year over all of central and southernFinland, and thean- nual risk of blight is^greaterthan50%overmost of the remainder of the country. Average tuber losses increase overthe whole countryrelative tothebaseline,with thegreatestincrease in loss Fig. 7.Potential number ofgenerationsof the Columbia root-knot nematode(Meloidogyne chitwoodi)basedontemperature under:^(a)the baseline climate 1961-1990and (b) SILMU scenario 1(best guess) by2050.

occurring in centralFinland, where the increas- es in potential yield areestimatedtobe largest (Figure 8b). The estimated increase in loss due to late blight is similar in magnitude and may evenexceed the simulated increase in tuberyield over much of central and southernFinland, but yield gains outweigh losses in the north.

Changes in the potential for cultivating

new crops: the ^case of maize

All of the above model simulations pertain to cropvarieties thatare cultivatedatthepresent-

day in Finland. One important conclusion is that as the climate changes, these varieties will be- come progressively maladapted to the new en- vironment. An obvious farming response is to substitute these with different varietiesor even with new species that can properly exploit the changed conditions. One candidateas a substi- tutecrop fora warmerclimate is maize,and the model simulations with CERES-Maize_were in- tended^toassessthe potential for this crop under the changed conditions.

The results of these simulations_{are summa-} rised in Figure 9. Even in the mostfavourable areas of thecountry, a measurable (non-zero) Fig.8. Simulatedpotentialtuber loss inpotato (t/ha) due to lateblight (Phytophthora infeshms)under:^(a)baselineclimate,

1961-1990and (b) SILMU scenario I(best guess)by 2050.

grain yieldcanonly be expected in about 6 years in 10 (Figure 9a).Over^mostof thecountry,dur- ing mostyears, the crop would^noteven reach the grain filling stage of development. 30-year average baseline yields for ^an unstressed crop areestimatedtobe about2 t/ha^atisolated loca- tions in southernFinland, although in favoura- ble years yields as high_as7 t/ha are simulated.

Except for theserare instances, however, it is unlikely that the grain yield would be ofasuffi- cient quality for any useother than green fodder orsilage.

Under thetemperature changes given by SIL

MU scenario 1 by 2050, the probability of ob- taining a non-zero grain yield is estimated to have increasedto 100% in southernFinland, and measurable yields could be obtained inover50%

of years _asfar north _as Oulu (Figure ^9b). The mean level of potential maize yields estimated under this scenario exceeds 7 t/ha in the most favourable _areas (10 t/ha in warm years), with yields averaging 4 t/haor more over a sizable

part of southern Finland. The effects of increas- ing CO,concentrationson cropphotosynthesis, which _are not modelled here, are likely to be much lower foracrop likemaize, witha

C 4

^pho-

Fig. 9.Simulatedprobability (percent)ofobtainingmeasurable^(non-zero)grain yield of maize under:^(a)baseline climate 1961-1990and^(b)SILMUscenarioI (best guess) by2050.

tosynthetic pathway, than for

C 3 ^crops

^{like bar-}

ley and _potato(e.g. see Reilly etal. 1996).C0₂ and water stress effects will be investigated in further work with the model.

Quantifying the uncertainties

We have presented results from a diverseset of models and forarange ofclimatic scenarios. The results are preliminary, and all are subject ^to uncertainties. In an attempt toquantify someof these uncertainties,four sources of uncertainty have been identified and compared using the development model for spring wheat(cv. Ruso) described earlier: cultivartype,model^error, glo- bal scenario assumptions, and regional scenario differences.

Cultivar type. The response ofa given crop species toclimate varies widely among differ- entcultivars of that crop, adiversity that forms the whole basis of plant breeding. The differenc- es in response of^two cultivars of wheat(Ruso and Kadett), chosen to represent the range of growth traits in Finnish cultivars, are shown in Figure 10a. Though this isnot strictlya measure of uncertainty, it does indicate what range ofre- sponses are implied when generalising the be- haviour of individual crop species.

Model errors. The quantification of model errors and uncertainties is perhaps the single mostdifficult and time consuming task in mod- el testing.However, the importance of suchex- ercises cannot be overstated, as the credibility of the results depends upona good knowledge of model accuracy.

Herewe use as asimple example the phono- logical models for cereals described above. The uncertainty of the linear relationship between temperature and developmentrate is delimited as95% confidence limits around the regression line in Figure 1(dashed lines).These confidence limitscan be expressed geographically as uncer- tainties in the limits of suitability (Figure 10b).

Compared ^to the differences between cultivars (Figure 10a),this sourceof uncertainty appears fairly modest.However,if the uncertainty ofany

single point plotted in Figure 1 were^tobe eval- uated, the plotted and mapped range of uncer- tainty would be muchgreater.Wider uncertain- tyranges than shown in Figure 10b would also be expected for results from the other models described in this paper, werethey ^tobe evaluated.

Global scenario assumptions. By far the

greatest source of uncertainty in the estimates of future crop potential shown here is attributa- ble to the range of SILMU temperature scenari- os.This range ^accountsfor uncertainties ^atthe global level of both greenhouse gas emissions into the atmosphere and the mean climate re- sponse ^to changing atmospheric composition (Carter 1996).The magnitude of the uncertainty range is illustrated ⁱⁿFigure 10c, which ^shows the shift in suitability of spring wheat^{(cv. Ruso)} by 2050 for each of the SILMU policy scenarios (low, central and high). The range of estimates varies froma low estimate averaging about 100 km under scenario 2,toahigh estimate of about 550 km under scenario 3. Similar results arealso obtained for the other cereal crops.

Regional scenario

differences.

^{The SILMU}

policy scenarios only express the uncertainties

at global level. There are further uncertainties

to considerat aregional level, some of which arerepresented in the SILMU scientific scenar- ios. Thesearebasedon alternative regional pat- ternsoftemperature change specified by gener- al circulation models (Carter ^1996).The effect of this source of uncertaintyon cropsuitability is shown in Figure lOd. Differences in the limits of suitability under

alternative

regional climate predictions appear to be of the same order of magnitudeascorresponding differencesassum- ing alternative crop cultivars (Figure 10a).

Discussion

The major findings of this studycan besumma- rised asfollows:

1) A warming of the climate is estimated to induce shifts in the northern limit of cereal cul-

Fig. 10.Range inestimates ofsuitability forspring wheat (cv. Ruso) attributable to: (a) cultivar type (Ruso and Kadett under the baseline climate), (b)uncertaintyof thedevelopmentmodel (baseline climate), (c) uncertaintiesinprojectionsof global climatechange(extension under the SILMUlow,central andhigh policyscenariosby2050) and (d) uncertainties in projectionsof regional climate change (extension under the SILMU scientific scenarios^-GFDL, UKTR and MPIby^2050).

Forexplanation seetext.SILMUscenariosaredescribed inCarter (1996).

tivalion of between 100 and 150 km per °C in- crease in mean annual temperature. Under the best guess SILMU^scenario, this translates into a meannorthward shift of about50 km per dec- ade up^to2100.

2)Average grain yields of present-day bar- ley cultivars in 2050 areestimatedtobegreater thanatpresent overthe wholecountry,with the greatest increases in north-central Finland un- der the best guess SILMU scenario foracombi- nation of changed climate and increased C0₂ concentration. However, increased temperature alone hasanegative effectonyields in southern Finland, due to ashortened grain filling period.

3)The risk of crop damage by pestsand dis- eases increases in all regions undera warming of the climate. Northward shifts in the distribu- tion of certain _pestscould be of_a similar mag- nitude and rate as reported above for cereal crops. Additional generations of multivoltine pest speciescan also be expected. The damage potential of diseases such as potato late blight could increase ata similarrate as the potential increase in yields of the crop host.

4) The opportunities for cultivatingnewcrops in Finland could increase under a warmer cli- mate.Estimates suggest that grain maize could be cultivated reliably by 2050 in manyparts of southern Finland under the best guess SILMU scenario.

5) Uncertainties surround all of these esti- mates.The largest uncertainty is that attributa- ble toestimates of futureclimate, but other un- certaintiestoconsider include modelerrorsand assumptions, observationerrors,and alternative modelling methods.

The resultsrepresent anearly attemptatde- picting the regional effects of climate change on Finnish agriculture. Considerable effortsarestill required _torefine the methods and to improve the models. Future research efforts shouldcon- centrateon the following areas:

Further testing and validation of existing models,toestablish their credibility when ap- plied in climate change studies. In addition

torigorous sensitivity analyses, this also re-

quires the acquisition of validation material from controlled experiments and from re- gions with a warmer climate than in Finland Refining methods of scaling-up from site models ^to the regional-scale, including im- proving methods of spatial modelvalidation, analysing and depicting uncertainty and testing and applying stochastic weathergenerators.

Improving the realism of crop potential map- ping, by accounting for local features of land

cover,soils and topography which may pre- clude arable agriculture altogether.

Widening the focus to consider climate change effects_on other crop species, includ- ing the major food and non-food crops cur- rently grown and those potentially viable under _a changed climate. Models will need

tobe tested for each crop considered.

Consideringmore aspects of farm-level ad- aptation to changing climate, for example, adjustments in fertilizer application, timing of operations, crop switches and irrigation.

The effects of all of these can be simulated with appropriate models.

Evaluating the effects of climate change on crop quality. There are many measures of quality, and most can be related toseasonal weather.CO, concentration is also known ^to affect some ^aspects of crop quality.

Clearly there is much work stilltobe done if we are toobtain_{a more} comprehensive and_ac- curatepicture of agricultural potential in Finland under the changing climate of the nextcentury.

Acknowledgements. We are gratefulto anumber of col- leaguesforsupplying models and advice: Jouko Kleemola of KemiraOy, Espoo, provided the computer code of CropWatNand much assistance incalibratingthe model.

Timo Kaukoranta and Kari Tiilikkala of theCropProtec- tionDepartment, AgriculturalResearch Centre ofFinland, Jokioinen,offered the potato lateblightand nematode mod- els,respectively. We also thank Eino Hellsten andAri Venäläinen of the Finnish Meteorological Institute for set- tingup theinterpolationroutines forclimatological data.

This workwasfundedbytheAcademyof Finlandaspart of the Finnish ResearchProgramme on Climate Change (SILMU).