Managing precipitation use in sustainable dryland agroecosystems

Transcription

1 Ann. ppl. Biol. (24), 144: Printed in UK 127 Mnging precipittion use in sustinble drylnd groecosystems By GARY A PETERSON* nd DWAYNE G WESTFALL Deprtment of Soil nd Crop Sciences, Colordo Stte University, Fort Collins, CO 8523 USA (Accepted 8 July 23; Received 24 Mrch 23) Summry In the Gret Plins of North Americ potentil evportion exceeds precipittion during most months of the yer. About 75% of the nnul precipittion is received from April through September, nd is ccompnied by high tempertures nd low reltive humidity. Drylnd griculture in the Gret Plins hs depended on whet production in whet-fllow groecosystem (one crop yer followed by fllow yer). Historiclly this system hs used mechnicl weed control prctices during the fllow period, which leves essentilly no crop residue cover for protection ginst soil erosion nd gretly ccelertes soil orgnic crbon oxidtion. This pper reviews the progress mde in precipittion mngement in the North Americn Gret Plins nd synthesises dt from n existing long-term experiment to demonstrte the mngement principles involved. The long-term experiment ws estblished in 1985 to identify drylnd crop nd soil mngement systems tht would mximize precipittion use efficiency (mximiztion of biomss production per unit of precipittion received), improve soil productivity, nd increse economic return to the frmers in the West Centrl portion of the Gret Plins. Embedded within the primry objective re subobjectives tht focus on reducing the mount of summer fllow time nd reversing the soil degrdtion tht hs occurred in the whet-fllow cropping system. The experiment consists of four vribles: 1) Climte regime; 2) Soils; 3) Mngement systems; nd 4) Time. The climte vrible is bsed on three levels of potentil evpotrnspirtion (ET), which re represented by three sites in estern Colordo. All sites hve nnul long-term precipittion verges of pproximtely 4-45 mm, but vry in growing seson open pn evportion from 16 mm in the north to 1975 mm in the south. The soil vrible is represented by ctenry sequence of soils t ech site. Mngement systems, the third vrible, differ in the mount of summer fllow time nd emphsize incresed crop diversity. All systems re mnged with no-till techniques. The fourth vrible is time, nd the results presented in this pper re for the first 12 yr (3 cycles of the 4-yr system). Compring yields of cropping systems tht differ in cycle length nd systems tht contin fllow periods, when no crop is produced, is done with technique clled nnulistion. Yields re nnulised by summing yields for ll crops in the system nd dividing by the totl number of yers in the system cycle. For exmple in whet-fllow system the whet yield is divided by two becuse it tkes 2 yr to produce one crop. Cropping system intensifiction incresed nnulised grin nd crop residue yields by 75 to 1% compred to whet-fllow. Net return to frmers incresed by 25% to 45% compred to whet-fllow. Intensified cropping systems incresed soil orgnic C content by 875 nd 14 kg h -1, respectively, fter 12 yr compred to the whet-fllow system. All cropping system effects were independent of climte nd soil grdients, mening tht the potentil for C sequestrtion exists in ll combintions of climtes nd soils. Soil C gins were directly correlted to the mount of crop residue C returned to the soil. Improved mcroggregtion ws lso ssocited with increses in the C content of the ggregtes. Soil bulk density ws reduced by.1g cm -3 for ech 1 kg h -1 of residue ddition over the 12-yr period, nd ech 1 kg h -1 of residue ddition incresed effective porosity by.3%. No-till prctices hve mde it possible to increse cropping intensifiction beyond the trditionl whet-fllow system nd in turn wter-use efficiency hs incresed by 3% in West Centrl Gret Plins groecosystems. Cropping intensifiction hs lso provided positive feedbcks to soil productivity vi the incresed mounts of crop residue being returned to the soil. Key words: No-till, soil wter storge, summer fllow, cropping systems, whet, mize, millet Introduction Efficient precipittion mngement is the key issue in mintining sustinble drylnd griculture in the North Americn Gret Plins. Soil tillge nd crop selection re the principl tools vilble to mnge *Corresponding Author E-mil: gry.peterson@colostte.edu 24 Assocition of Applied Biologists

2 128 GARY A PETERSON & DWAYNE G WESTFALL the precipittion. Efficient precipittion mngement involves mximiztion of precipittion cpture in the soil, minimising loss of the stored soil wter, nd mximizing wter-use efficiency (WUE) of the plnt. WUE = (Crop Yield)/(Precipittion received + soil wter depletion during the cropping seson). The primry mngement tool for mximizing wter cpture nd minimising soil wter loss is reduced tillge or the complete voidnce thereof. Tillge choices, including type nd timing, ffect the mount of crop residue cover mintined on soil surfce nd the soil pore size exposed to the tmosphere. Recognising the reltionships between climte nd tillge is criticl for efficient precipittion mngement. Efficient use of wter is lso governed by crop choice, plnting dte, fertiliser use, etc. This pper discusses the principles of drylnd groecosystem mngement nd explins how our understnding of these principles hs chnged over the lst 15 yr in the Gret Plins region. Precipittion nd Temperture Ptterns In the North Americn Gret Plins, potentil evportion exceeds precipittion during most months of the yer, which significntly ffects wter cpture, wter retention, nd wter use efficiency by crop plnts. About 75% of the nnul precipittion is received from April through September, nd is ccompnied by high tempertures (Fig. 1) nd low reltive humidity. Note tht the open pn evportion trcks closely with the verge ir temperture (Fig. 2), nd tht this potentil to lose the precipittion from the soil is high. To illustrte the lrge differences tht exist mong drylnd groecosystems of the world, contrsts between the Gret Plins temperte climte nd two types of Mediterrnen climte (Morocco nd Oregon USA) re provided in Figs. 3 nd 4. Morocco nd Oregon both hve Mediterrnen climtes nd receive the bulk of their precipittion in the coolest months of the yer, which mens they hve smller evportive losses during the time they receive their precipittion reltive to the Gret Plins environment. Note, however, tht ir tempertures during the winter precipittion period in the Moroccn climte re 8-1ºC wrmer thn in the Oregon Mediterrnen climte. Therefore the wter cpture nd retention issues differ, even though they both hve Mediterrnen climtes. For exmple, snow mngement becomes n importnt issue in the Oregon environment. Obviously precipittion distribution in reltion to the growing seson nd evportion potentil dicttes crop choices, crop sequences, nd effectiveness of wter conservtion prctices in ny given climte. Precipittion (mm) Temperture ( C) Precipittion (mm) Temperture ( C) Fig. 1. Long-term ( ) precipittion nd temperture distributions in estern Colordo, USA.! = precipittion; " = temperture Jn Feb Mr Apr My Jun Jul Aug Sep Oct Nov Dec Month Jn Feb Mr Apr My Jun Jul Aug Sep Oct Nov Dec Month Fig. 2. Long-term ( ) temperture nd open pn evportion distributions for the summer growing seson in estern Colordo, USA.! = evportion; " = temperture Jn Feb Mr Apr My Jun Jul Aug Sep Oct Nov Dec Month Fig. 3. Long-term precipittion distributions in western Morocco ( ;!), estern Colordo ( ; "), nd estern Oregon (194-22; #), USA. Jn Feb Mr Apr My Jun Jul Aug Sep Oct Nov Dec Month Fig. 4. Typicl temperture distributions in western Morocco ( ; )!, estern Colordo ( ; "), nd estern Oregon (194-22; #), USA. Temperture ( C) Gret Plins historic sitution Drylnd griculture in the Gret Plins hs developed round whet production in whetfllow groecosystem. The whet-fllow system ws developed to decrese risk of crop filure (Peterson Evportion (mm)

3 Mnging precipittion use in sustinble drylnd groecosystems 129 et l., 1996). The soil wter stored during the fllow period increses the probbility of successful whet crop. Spring whet is grown in the northern Gret Plins nd is plnted in lte spring with lte summer hrvest. The fllow period tht follows vries in length from 18 to 21 months depending on exct plnting nd hrvest dtes. The ctul time the whet plnts re growing in the field is only 3 to 6 months out of the 24 month cycle. In the centrl nd southern Gret Plins winter whet is grown nd the plnts re present for bout 1 months out of the 24 month cycle. Weed control during the fllow period is criticl becuse mximum wter storge cn only occur if the fields re weed free. Prior to the dvent of herbicides the only fesible weed control ws tillge, which usully left the soil surfce brren of residue cover. Summer fllow becme wy of life for frmers in the Gret Plins following the 193s dust bowl. Higher wrtime prices nd much improved trctor power systems nd implements fcilitted tillge weed control during the fllow period (Greb, 1979). According to Hs et l. (1974) there re more thn 6.1 million hectres of summer fllowed lnd in the US. Gret Plins lone, nd since the Cndin provinces lso use the summer fllow technique, the totl re is even lrger. Greb (1979) chronicled progress in winter whetfllow systems from the erly 19s through to 1977 nd then projected progress through to 199. Chnges since 1916 in fllow tillge systems hve improved wter storge, fllow efficiency (% of fllow precipittion stored s soil wter), winter whet yield, nd precipittion-use efficiency (PUE) (Tble 1). As tillge type chnged, the number of tillge opertions per fllow period decresed, nd the mount of crop residue remining on the soil surfce incresed. PUE doubled from 1916 to 1975, incresing from 1.22 to 2.78 kg of whet h -1 mm -1. This ws lrgely due to improved fllow-period soil wter storge efficiencies, which incresed from 19 to 33% over the sme time period. Greb (1979) predicted tht fllow efficiency would increse to 4% by 199, resulting in PUE of 3.25 kg h -1 mm -1. The mechnisms tht llowed the improvements in fllow efficiency nd PUE re complex rry of intercting fctors. They include: (1) Residue interception of rindrop impct, which helps mintin infiltrtion rte; (2) Decresed first stge evportion rtes due to cooler soil tempertures under the residue; (3) Less opportunity for stimulted evportion due to fewer tillge events; (4) Decresed wind speed t the soil surfce becuse of residue protection; (5) Improved fertilistion prctices; (6) More opportunistic weed control; (7) Improved semi-dwrf whet cultivrs; nd (8) More timely fllow opertions becuse of more trctor horsepower nd better equipment (Peterson et l., 1996). No single fctor hs chnged the system, but ll hve worked in concert to crete net positive outcome. Modern soil wter storge potentil in fllow As noted bove, Greb (1979) predicted incresed fllow storge efficiency bsed on improved residue mngement nd more economic no-till methods. Unfortuntely, we hve not been ble to improve upon the 35% storge efficiency Greb chieved in the erly 197s. Fllow efficiency reports from the 198s nd into the 199s re generlly less thn 42%, regrdless of the climtic zone where the dt were collected (Tble 2). The rnge of efficiencies reported by McGee et l. (1997) in Colordo under no-till conditions, 17-28% were disturbingly low. They represent wide rnge of climte nd soil combintions, nd it ppers it will be difficult to improve upon them with current fllow technology. Tble 1. Progress in fllow systems nd winter whet yields, United Sttes Centrl Gret Plins Reserch Sttion, Akron, Colordo, USA Yers Chnges in fllow systems Number of Tillges Fllow wter storge Fllow efficiency Winter Whet Yield Precipittion use efficiency b mm % kg h -1 kg h -1 mm Mximum tillge; plow hrrow (dust mulch) Conventionl tillge; shllow disk, rod weeder Improved conventionl tillge; begin stubble mulch Stubble mulch; begin minimum tillge with herbicides (1969) Projected estimte. Minimum tillge; begin no-till 1983 Bsed on 14 months fllow, mid-july to second mid-september b Assuming 2 yers precipittion per crop in whet-fllow system Source: Adpted from Greb (1979)

4 13 GARY A PETERSON & DWAYNE G WESTFALL Fllow-storge efficiency ws eqully low in the northern climtes of Cnd nd North Dkot, despite their lower evportion potentils. The much longer fllow period, 21 months for spring whetfllow systems, compred to 14 months for winter whet-fllow systems, lso contributes to the low efficiencies in the lower evportion climtes. Greb et l. (1967) nd Unger (1978) demonstrted tht surfce residue gretly increses fllow wter storge, but tht residue mounts in excess of 6.7 t h -1 re required to chieve fllow efficiencies greter thn 35-4%. Unfortuntely in the Gret Plins residue mounts t whet hrvest, the mximum residue ccumultion point in the system cycle, re commonly in the 2.2 to 5.6 t h -1 rnge. With fvourble precipittion they cn rech 7.8 t h -1, nd on soils tht receive runoff wter from surrounding hills they my even rech 1.1 t h -1, but the ltter cses re rrity. Thus becuse residue levels in the Gret Plins re usully do not rech 6.7 t h -1 threshold, it is quite likely tht mximum fllow storge efficiencies were reched in the 197s. Looking for chnges in tillge systems tht will improve wter storge efficiency lso seems futile. Ultimtely surfce residue quntity is reduced by tillge, nd therefore tillge decreses residue effectiveness. In ddition, tillge directly ffects wter storge vi its effect on mcropore spce nd ccelerted evportion. Tillge my crete surfce soil mcropores tht improve cpture but it simultneously increses evportive wter loss. Furthermore the whet-fllow system with summer fllow tillge for weed control leves the soil highly vulnerble to wind nd wter erosion for Tble 2. Modern soil wter storge efficiencies of no-till nd reduced till summer fllow systems over rnge of environments in the Gret Plins of the USA Wter Storge Efficiency (%) Men Rnge Stte or Province Reference Ssktchewn, Cmpbell et l Cnd North Dkot Deibert et l Montn Tnk, Colordo Smik, Colordo McGee et l., b Knss Schlegel, Knss C A Norwood c Texs Jones & Johnson, 1993 Spring Whet (21-month fllow) ll other dt winter whet (14-month fllow) b Reduced Tillge c Personl communiction - Southwest Knss Reserch Center, Knss Stte University, Grden City, Knss USA long periods of time becuse of lck of cover. The frequent tillge coupled with smll returns of crop residue to the soil lso hs ccelerted soil orgnic C oxidtion. Since mny Gret Plins soils hd low initil mounts of orgnic C, soil structurl strength nd stbility lso hve declined. The net result is decresed wter infiltrtion potentil, nd thus lower wter storge efficiencies. Problems nd Chllenges Mngement nd wter storge interctions It ppers tht fllow mngement techniques cnnot be substntilly improved beyond the levels reported in Tble 2, even with no-till, becuse residue mounts produced in the Gret Plins re too smll to llow further reductions in evportion rtes nd totls. Therefore nother pproch must be found for incresing PUE. Our chllenge is to simultneously increse wter cpture, decrese the wter storge time in the soil, nd to choose plnt species nd rottions tht use the wter more efficiently. Wter cpture nd storge efficiency re usully highest when the soil surfce is dry nd in receptive condition for rinfll. At whet hrvest time in the Gret Plins (July), soils often re t zero plnt vilble wter content nd cn bsorb wter rpidly. At this point in the crop cycle, there is mximum residue cover on the soil; thus soil conditions re very receptive to wter infiltrtion. Smik & Wicks (1968) reported herbicidl weed control nd decresed tillge gretly improved wter storge during the erly portion of the fllow period in winter whet system. Conventionl plough tillge tretments stored no wter in the erly fllow period, while minimum tillge tretments stored 12% of the precipittion nd complete no-till incresed wter storge efficiency to 24%. By spring of the following yer, which is only 8 months into the 14 month fllow period, the plough tillge hd stored only 16% of the precipittion (56 mm of wter), while the minimum till nd no-till systems hd stored 4% (14 mm of wter) nd 6% (21 mm of wter), respectively. Obviously when wter is stored erly in the fllow seson, storge becomes less efficient in the lter prt of the fllow period. For spring whet-fllow systems, Hs & Willis (1962) lso found tht little or no soil wter ws stored in summer fllow fter 1 July. These dt point to the possibility of terminting the fllow period in no-till nd reduced tillge systems before July nd plnting summer crop to use the wter vi trnspirtion rther thn lose it to evportion. Mngement techniques tht foster erly wter cpture nd retention fter whet hrvest nd during the winter nd spring periods usully result in moist

5 Mnging precipittion use in sustinble drylnd groecosystems 131 surfce soils tht re ner field cpcity by My. When rin event occurs in the summer period, the gretly reduced infiltrtion rtes llow wter to remin on the no-till soil surfce for longer periods compred to tilled conditions. The high tempertures in the lte fllow period from July to September, in concert with high vpour pressure deficits in the ir, ccelerte evportion nd keep precipittion cpture to minimum. On sloping lnd, runoff is incresed by the compct soil surfce nd wter cpture is decresed even more thn on level lnd (Jones et. l., 1994). In conventionl winter whet-fllow systems, where weeds re not controlled fter hrvest nd/or tillge is used for weed control, soil wter contents in spring re much lower nd wter storge potentil during the My-September period is greter thn in no-till systems. Since these soils re tilled multiple times for weed control during the fllow period, they hve more mcropore spce t the surfce nd wter infiltrtion is not impeded. Unfortuntely, ny wter stored in the tillge lyer is usully rpidly lost to evportion becuse the fllow weed control tillge hstens evportion by exposing moist soil during the hottest period of the yer. Reports by Blck & Power (1965), Deibert et l. (1986), nd Norwood (1994) hve ll substntited the inefficiency of wter storge in no-till systems during the lte portion of the summer fllow period, whether it is winter or spring whet sitution. In erly no-till reserch, Blck & Power (1965) working within spring whet fllow system found tht the fllow storge efficiency from hrvest to the following My ws 66%, from My to September 9%, nd from September to seeding the next spring 19%. Deibert et l. (1986) working in spring whetfllow systems reported erly fllow storge efficiencies of 56-59% [9 to 125 mm of wter stored], but efficiencies of only 26-36% fter full 21-month fllow period [112 to 117 mm of wter stored]. Norwood (1994) in Knss reported storge efficiency of 46% for the 11 month period from winter whet hrvest to spring sorghum plnting [175 mm of wter stored]. Norwood s whet-fllow system, for the entire 14-month fllow period, only hd n efficiency of 23% [137 mm of wter stored]. There ws wter loss of 38 mm during the lte fllow period. A long fllow period ppers to decrese stored wter in most instnces. Potentil for improving precipittion use efficiency Most dt indicte tht there cn be s much or more stored wter in no-till mnged soils in the spring fter whet hrvest s there will be if fllow is continued until fll whet plnting. It ppers tht intensifying the cropping pttern, by shortening the summer fllow period nd using the precipittion nerer to the time it is received, would increse the overll system PUE nd ultimtely increse soil productivity vi the incresed nnul mounts of residue dded to the soil. Mterils nd Methods We estblished our long-term groecosystem project in 1985 to identify drylnd crop nd soil mngement systems tht would mximise PUE, improve soil productivity, nd increse economic return to the frmers in the West Centrl portion of the Gret Plins. Specificlly we were looking for wys to reduce the mount of summer fllow time nd to reverse the soil degrdtion tht hs occurred becuse of the tilled whet-fllow cropping system. The experiment hs four vribles: 1) Climte regime; 2) Soils; 3) Mngement systems; nd 4) Time (Fig. 5). All phses of ech cropping system re present ech yer. The cropping systems were rndomly imposed cross the summit position of the lndscpe nd then continued downslope without further rndomistion. The experimentl unit is therefore specific soil series (slope) within site nd within cropping system phse. The experimentl design is split-split-block design tht includes loction, slope position, nd cropping systems vribles within two replicted blocks (Peterson et l. 1993). Cropping systems were rndomly ssigned in strips within ech block t ech loction. Slope positions run within the strips cross cropping systems. Although slope positions were not rndomized, split block nlysis t ech loction ws used s if they hd been (Steel & Torrie, 1997). The vrince ws prtitioned ppropritely to test the min effects nd ny interctions. A complete discussion regrding the justifiction for the experimentl design, the design detils nd overll mngement re reported in Peterson et l. (1993). Climte grdient The climte vrible is bsed on three levels of potentil evpotrnspirtion (PET), which re represented by three sites in estern Colordo. All sites hve long-term precipittion verges of pproximtely 4-45 mm, but vry in growing seson open pn evportion. They re locted ner Sterling (4.37 o N, o W), Strtton (39.18 o N, o W), nd Wlsh (37.23 o N, o W) in estern Colordo. Sterling hs n open pn evportion of 16 mm during the cropping seson, while Strtton hs 1725 mm, nd Wlsh hs 1975 mm (Peterson et l., 1993, 21). Ech site hd been mnged with tillge in either whet-fllow (Sterling nd Strtton) or sorghum-fllow (Wlsh) system for more thn 5 yers prior to our initition of no-till mngement in 1985.

6 132 GARY A PETERSON & DWAYNE G WESTFALL Soil grdient Ech site hs soil cten with three distinct soils identified s summit slope, side slope, nd toe slope positions. The soils t the three sites re either Argiustolls or Ustochrepts (Peterson et l., 1993). Cropping systems Mngement systems with vrious cropping intensities were exmined cross the soil sequences t ech loction in strips 6.1 m wide by 185 to 3 m long, depending on site. These systems included: winter whet (Triticum estivum L.)-fllow (WF); winter whet-mize (Ze mys L.)-fllow (WMF); winter whet-mize-proso millet (Pnicum miliceum L.)-fllow (WMPF); continuous cropping (CC) (crops grown over the yers included mize, sorghum, winter whet, forge millet, nd sunflower); nd perennil grss (G). Grin sorghum [Sorghum bicolor (L.) Moench] replced mize in the cropping systems t Wlsh becuse it is more suited to the high ET nd longer growing seson thn mize. All phses of ech system were present ech yer. All detils regrding herbicide choices, rtes, nd ppliction times for the no-till systems were reported by Peterson et l. (21). The systems represent grdient of cropping intensity (crops divided by yers in the rottion), thus WF hs n intensity fctor of.5. Intensity fctors for WMF, WMPF, nd CC re.67,.75, nd 1., respectively. The grss does not hve n intensity fctor since it is perennil system. Grss stnds were estblished in the spring of 1986 nd contin mixture of perennil species including both wrm nd cool seson grsses (Peterson et l. 1993). Time The fourth vrible is time, nd the results presented in this pper re for the first 12 yers (3 cycles of the 4-yer system). A minimum time horizon of 2 yers (5 cycles of the 4-yr rottion) is plnned (Peterson et l., 1993). Mesurements System responses were ssessed vi totl boveground plnt productivity, wter use efficiency, chnges in soil chemicl, physicl, microbiologicl properties, nd economic evlutions. All soil smples, dry mtter yields, soil wter mesurements, etc. were collected from benchmrk res within ech experimentl unit. Detils hve been reported in Peterson et l. (1993). Complete climtic records Fig. 5. Experiment design for the drylnd groecosystem project in Colordo, USA.

7 Mnging precipittion use in sustinble drylnd groecosystems 133 hve been mintined t ech site, nd soil smples hve been rchived on regulr bsis. Sttisticl nlyses Anlyses of Vrince (ANOVA) were done using the procedure generl liner model (GLM) of the Sttisticl Anlysis System (SAS, Anon., 1999) for test of ll min effects nd interctions. The option mens in the GLM procedure ws used to obtin ll min effect men seprtions using Fisher s Protected Lest Significnt Difference (LSD) using the pproprite error term when the ANOVA showed significnce (P =.5). Loction ws tested with repliction (Loction) term. Slope nd site by slope interction ws tested using slope*repliction (Loction) term. Cropping intensity nd loction by cropping ws tested using the cropping*repliction (Loction) term. When interctions were significnt, LSDs were clculted by compring slopes within sites (site slope), cropping system within site (site cropping), cropping system within slope (slope cropping system), nd cropping system within site nd slope (site slope cropping system) using the pproprite stndrd error term. Results nd Discussion Grin nd totl bove-ground biomss yield Intensifying the cropping systems hs incresed nnulised grin yield by more thn 75% reltive to the yield of the WF system (Fig. 6). These yield increses hve trnslted into 25-4% gins in net income for frmers (Peterson et l., 1993b; Kn et l., 22). The lrgest step gin in nnulized yield ws chieved with the ddition of mize or sorghum to the system (two crops in 3-yr system). Incresing cropping intensity to the 4-yr system only resulted in smll yield increses reltive to the 3-yr system. We were not ble to compre CC with the other systems in terms of grin yield becuse the CC system included forge crops tht do not hve grin component. Yields of totl bove-ground biomss do, however, llow us to compre CC with the other systems, which re entirely grin-bsed. Note in Fig. 7 tht CC produced nnulized totl biomss yields superior to ll systems contining fllow. Avoiding fllow yer, even in rid climtes, hs mximised totl bove-round biomss production. Unfortuntely the monetry vlue of the forge crop in the continuous systems is so low tht the CC, s we hve prcticed it, hs not been s profitble s the 3- or 4-yr systems tht still include summer fllow period. Plnt wter use efficiency (WUE) Systems with fewer summer fllow periods hve incresed WUE. The 3-yr system hd 27% increse in grin WUE reltive to WF, nd the 4-yr system hd WUE tht is 37% greter thn WF (Fig. 8). Incresed grin nd biomss yields were produced becuse wter, tht usully is lost to evportion in the WF system, ws vilble for trnspirtion vi crop production in the no-till systems. Soil physicl properties One of the mjor gols in precipittion mngement is to hve surfce soil conditions tht re receptive to wter when it is received. The soils were ntively low in orgnic mtter becuse they were formed in n rid environment. Despite this condition, they hd good ggregte strength when first plced under cultivtion. The ntive grss vegettion, with its highly developed surfce root Grin yield (kg h -1 ) Totl biomss (kg h -1 ) 8 7 LSD = yr 3-yr 4-yr Cropping systems Fig. 8. Grin wter use efficiency s ffected by three cropping systems verged over climte nd soil grdients nd yers ( ) in the drylnd groecosystem project in Colordo, USA. WUE (kg h -1 mm -1 ) LSD = =.5 2-yr 3-yr 4-yr Cropping systems Fig. 6. Annulised grin yield s ffected by three cropping systems verged over climte nd soil grdients nd yers ( ) in the drylnd groecosystem project in Colordo, USA LSD = =.5 2-yr 3-yr 4-yr Cont. crop Cropping systems Fig. 7. Annulised totl biomss yield s ffected by four cropping systems verged over climte nd soil grdients nd yers ( ) in the drylnd groecosystem project in Colordo, USA.

8 134 GARY A PETERSON & DWAYNE G WESTFALL system, provided well structured soil tht llowed pioneer frmers to do good job of wter cpture within the limits of the technology of their er. Since ll of the cropping systems relied on frequent tillge events, the orgnic soil C depleted rpidly nd ggregte size nd strength diminished quickly. Surfce crusting begn to occur fter rinfll events nd the frmer s only lterntive ws more tillge to crete mcropores to improve cpture potentil of the next precipittion event. Thus, both wter cpture nd retention cpbilities hve declined with time. The im ws to determine if, by the use of no-till prctices coupled with cropping intensifiction, the negtive effects of pst mngement could be reversed to improve soil structure nd ultimtely the wter cpture potentil of our soils. Properties of the immedite surfce soil lyer, such s bulk density, porosity nd mcroggregtion, ffect pore spce nd pore sizes, nd re useful indictors of chnges in wter cpture potentil. Cropping system intensifiction under no-till mngement hs decresed the bulk density of the surfce soil lyer (Shver et l., 22). For exmple, CC reltive to WF decresed soil bulk density from 1.32 to 1.22 g cm -1 (Tble 3). This reduction in bulk density resulted in n increse of.4 m 3 m -3 in totl porosity; mening tht there is 4% more spce to infiltrte wter from rinfll event. Furthermore there ws n bsolute increse of 5% in effective pore spce, which mens tht, even if the surfce soil is t field cpcity wter content, there is 5% more spce to ccommodte rin event. The improvements in porosity re relted to the shift in proportion of mcroggregtes reltive to microggregtes tht occurred (Tble 3) (Shver et l., 22). In the Gret Plins environment rpid wter intke during nd fter rin event is criticl becuse wter tht ponds on the soil surfce for even short time is quickly evported in rid environments. The improvements in mcroggregtion nd effective porosity thus increse the opportunity for more efficient wter cpture. Soil orgnic crbon The cusl gent for the improvement in physicl properties hs been the ddition of more crop residue biomss to the soil reltive to the WF system (Shver et l., 23). Coupled with the lck of soil disturbnce in no-till environment, the dditionl residue C hs promoted ggregtion nd hs incresed ggregte stbility. Soil orgnic C levels hve incresed fter only 12 yr of intensively cropped no-till mngement (Tble 4) (Sherrod et l., 23). Ech step of incresing cropping intensity tended to increse surfce soil orgnic C t ll soil depths, but the increses were sttisticlly significnt in the surfce -2.5 nd cm soil lyers. The 3-yr system, WMF, tended to increse soil orgnic C levels, reltive to WF, in the two soil lyers nerest the surfce, but significnt differences were not found until intensifiction reched the 4-yr system. Continuous cropping, with no summer fllow period, incresed the orgnic C content of the surfce 2.5 cm of soil by 39% reltive to WF. Soil orgnic C increses were closely ssocited with the chnges in physicl chnges reported bove (Shver et l., 23). Furthermore the increses in soil C were directly linked to incresed crop residue biomss returned to the soil over the 12-yr life of the Tble 3. Soil physicl properties of the surfce 2.5 cm of soil s ffected by cropping intensity verged over climte nd soil grdients in Colordo fter 12 yr of no-till mngement. (Shver et l., 22) Bulk Density Totl Porosity Effective Porosity Mcroggregtes Microggregtes Cropping System g cm -1 m 3 m -3 m 3 m -3 kg kg -1 kg kg -1 WF WMF CC α = WF = whet-fllow; WMF = whet-mize-fllow; CC = continuous crop Tble 4. Soil orgnic C by soil depth s ffected by cropping system intensifiction fter 12 yr of no-till mngement (Sherrod et l. 23) Cropping System Soil Depth (cm) WF (kg h -1 ) WMF (kg h -1 ) WMPF (kg h -1 ) CC (kg h -1 ) α =.1 (kg h -1 ) NS NS WF = whet-fllow; WMF = whet-mize-fllow; WMPF = whet-mize-proso millet-fllow; CC = continuous crop

9 Mnging precipittion use in sustinble drylnd groecosystems 135 experiment (Shver et l., 23). Discussion No-till technology hs gretly ltered our bility to mnge precipittion in drylnd systems. Specificlly it hs improved the potentil for precipittion cpture nd for soil wter retention in our overll wter conservtion strtegy. In turn this hs permitted incresed cropping intensity tht hs proven to be both gronomiclly nd economiclly sound in the Gret Plins of North Americ. Furthermore these mngement strtegies hve provided positive feedbcks to the soil system tht should improve long-term productivity. Positive feedbcks include incresed soil orgnic C levels tht hve improved surfce soil structure, which in turn hve incresed the wter infiltrtion rtes of the soils. Incresed wter cpture becuse of greter infiltrtion hs the potentil to increse PUE by providing more wter for plnt production nd dditionl crop residue return to the soil. More seson-long cover, whether it be crop cnopy or crop residue, nd higher wter infiltrtion rtes ll hve mjor impcts in decresing soil erosion by wind nd wter. How do we fctor these findings into fundmentl principles tht cn be used to solve problems under other climte nd soil environments? Principles nd Inferences Fllow stges An nlysis of the fllow period in no-till WF system is instructive nd will llow us to better understnd how to mximize PUE. Frhni et l. (1998), using soil wter storge dt from the longterm experiment, divided the 14-month fllow period into three stges (Fig. 9). Definitions re: Stge I, the 2.5 month period from whet hrvest until mid- September; Stge II, the 7.5 month over-winter period from fll to erly My; nd Stge III the 4.5 month lte fllow period from spring until to whet plnting in mid-september. Note tht the vrious crop nd non-crop phses in 3- or 4-yr system fit well with these periods. The prtitioning ws not rbitrry; ech stge coincided with combintion of soil wter nd ir temperture conditions (Frhni et l., 1998). Stges I, II, nd III re coded s ornge, blue, nd red zones to depict the wter storge potentils of the stges. Wter-storge efficiencies re gretest, (5-85%), in the blue zone becuse ir tempertures re lowest nd precipittion is received s snow nd/or low intensity storms (Fig. 9). Storge efficiency in the ornge zone is much lower, 1% to 35%, becuse ir tempertures re much higher during this stge. However, surfce soil conditions re dry (ner wilting point), nd thus some wter infiltrtes before it is lost to evportion. Rinstorm intensity governs whether one is t the low or high end of the rnge. The red zone hs high ir tempertures, surfce soil nd profile wter contents ner field cpcity, evportion is high, nd no wter is stored during this time period (-4-5%). These fllow period stges cn be used to determine wht hs chnged s cropping systems re intensified nd cn give us July September (Hrvest) Stge I [Ornge] (2.5 months) Fllow Stges Stge II [Blue] (7.5 months) My September (Spring Plnt) (Autumn Plnt) Stge III [Red] (4.5 months) High Air Temperture Dry soil Surfce t wilting point Low Air Temperture Soil surfce t field cpcity Lower soil profile t wilting point High Air Temperture Wet soil profile t field cpcity 1-35% Precipittion Stored 5-85% Precipittion Stored -4-5% Precipittion Stored Fig. 9. Fllow stges nd their chrcteristics for drylnd cropping systems in estern Colordo, USA.

10 136 GARY A PETERSON & DWAYNE G WESTFALL insight s to wht might be fesible for future improvements. An nlysis of the cropping systems in our experiment reveled tht switching from 1 crop in 2-yr system, like WF, to 2 crops in 3-yr or 3 crops in 4-yr system did not pprecibly chnge the proportion of time in fllow (Tble 5). In fct, the proportion of time in fllow ctully incresed s cropping ws intensified, while the proportion of time in crop decresed. Note, however, tht the proportion of time in Stge I fllow decresed from 1% of the totl system time in 2-yr system to 5% in 4-yr system with summer crops. Furthermore the proportion of time in Stge III fllow decresed from 19% to 11% for the sme comprison. Since precipittion storge efficiency in Stge I ws only 1% to 35%, nd ws essentilly % in Stge III, hving smller proportion of the fllow time in these stges nd more in Stge II, which is highly efficient storge time, benefits the overll system. These shifts in when fllow is occurring prtilly explin the dvntge of the more intensive systems. An nlysis of when precipittion is received reltive to the fllow stges nd the cropping seson will complete the explntion. It ws demonstrted bove tht during fllow stges I nd II one cn expect to sve 1-35% nd 5-85% of the precipittion, respectively, which is in strk contrst to the % storge efficiency expected in Stge III. Dt in Tble 6 show tht cropping intensifiction did not pprecibly lter the proportion of the precipittion received during fllow Stges I & II, which re the best wter storge periods. However, intensifiction decresed the proportion of the precipittion occurring during fllow Stge III from 34% in 2-yr WF system to 17% for 4-yr system like WMPF. Since no wter cn usully be stored in Stge III, this represents gin for the system. In ddition it drmticlly incresed the proportion of the precipittion tht flls when summer crop cn be grown. This period of the yer is especilly fvorble for summer crops like mize, sorghum, nd millet. Switching from 2-yr (WF) to 4-yr system like WMPF tht includes summer crops incresed the proportion of the precipittion tht fell while crops were present in the field from 32% in 2-yr system (WF) to 47% for 4-yr system like WMPF. Receiving precipittion during the cropping period resulted in improved system PUE becuse the crop cnopy nd underlying residues left by the no-till system bsorb the rindrop impct, keep the soil surfce cooler, which decreses evportion, nd the overll result is net increse in soil wter storge. Runoff lso is diminished nd opportunity time for wter infiltrtion is incresed. An dditionl fctor tht contributes to improved wter conservtion during the cropping seson is tht the plnts continuously exhust the vilble wter from the surfce soil lyers, which improves the infiltrtion rte becuse dry soil is more receptive to wter thn is wet soil. Appliction of fllow nlysis to other climtes Cn we pply the principle of fllow stges to other climtic regions? In the Gret Plins sitution, most of the precipittion flls during the wrmest time of yer, which is within the growing seson for severl well dpted plnt species. An nlysis of WF system in Mediterrnen climte in the stte of Oregon, USA indicted tht the proportion of time in ech fllow stge ws the sme s for WF in the Gret Plins becuse the fll plnting nd summer hrvest dtes of the winter whet crop were essentilly identicl in both environments (Tble 7). In the wrmer Mediterrnen climte of Morocco, however, there were substntil differences in the proportion of time in ech fllow stge nd the time in crop. Whet ws plnted lter nd hrvested erlier in Morocco, which decresed the proportion of time in crop to 29% in contrst to 4% in the two cooler environments, nd incresed the proportion of time Tble 6. System nlysis Proportion of precipittion received in the vrious fllow stges nd during the cropping seson for three cropping systems Precipittion in stges I & II Precipittion in stge III Precipittion in cropping seson System % of totl precipittion for given system 2-Yer (WF) Yer (WMF) Yer (WMPF) WF = whet-fllow; WMF = whet-mize-fllow; WMPF = whet-mize-proso millet-fllow Tble 5. System nlysis Time in crop nd time in vrious fllow stges for three cropping systems in the Gret Plins System Totl time in crop Totl time in fllow Time in Stge I Time in Stge II Time in Stge III % of totl rottion time 2-Yer (WF) Yer (WMF) Yer (WMPF) WF = whet-fllow; WMF = whet-mize-fllow; WMPF = whet-mize-proso millet-fllow

11 Mnging precipittion use in sustinble drylnd groecosystems 137 Tble 7. System nlysis Proportion of time in fllow stges nd proportion of precipittion received in whet-fllow system in Colordo nd Oregon USA nd Morocco Fllow & Precipittion Distributions Climte Stge I Stge II Stge III Crop % of Rottion Time or % of Precipittion Proportion of system time in crop or fllow Colordo Oregon Morocco Proportion of system precipittion in ech stge Colordo Oregon Morocco System = 24 months for complete crop-fllow cycle in Stges I nd III fllow. In the Gret Plins environment such shift would hve negtive effect on soil wter storge becuse of the low precipittion storge efficiency possible in Stges I nd III, but this ws not the cse in the wrmer Moroccn environment s will be seen below. Much lrger contrsts ppered when the environments were compred on the bsis of precipittion distribution. In the Gret Plins only 32% of the precipittion fell while whet crop ws in the field, which contrsted shrply with 46% for the Oregon environment nd 5% for the Moroccn environment (Tble 7). One of the mjor wter svings in the Gret Plins environment ws voiding Stge III fllow becuse wter storge efficiency ws % in tht stge. In Oregon, however, only 12% of the precipittion fell during Stge III, nd in Morocco only 3%, mking voidnce of Stge III fllow much less criticl in the two Mediterrnen environments (Tble 7). Furthermore in Oregon nd Morocco 37% nd 47% of the precipittion ws received in Stge II fllow, respectively, nd becuse of the cool tempertures, this ws n excellent period in which to store wter. In the Gret Plins environment, which lso hs the potentil for mximum storge in Stge II, there is much less precipittion to be sved. Note, however, in the Oregon sitution extr ttention to snow nd snow melt retention would be required. In Morocco, where there re no frozen soil nd snow melt runoff problems, this is n even better period for soil wter storge. Bsed on this nlysis it is plin tht cropping intensifiction with wrm seson plnts such s mize or sorghum would hve little chnce for success in either the Oregon or Moroccn climte. Whet nd other cool seson species obviously fit the Mediterrnen environment very well reltive to the Gret Plins. Scientists who wish to find wys to improve PUE in other environments cn use this pproch to nlyse precipittion nd temperture distributions in reltion to vrious crop species nd thus identify the best intervention points for improved mngement. If fllow is to be used in cropping system, one should mke every ttempt to void the periods when precipittion storge is likely to be grossly inefficient. The gretest PUE will be chieved if non-crop periods cn be decresed in length nd if one cn grow crop during the time when the precipittion is being received. References Anon SAS User s Guide. Cry, North Crolin, USA: SAS Institute. Blck A L, Power J F Effect of chemicl nd mechnicl fllow methods on moisture storge, whet yields, nd soil erodibility. Soil Science Society Americ Proceedings 29: Cmpbell C A, Zentner R P, Steppuhn H Effect of crop rottion nd fertilizers on moisture conserved nd moisture use by spring whet in southwestern Ssktchewn. Cndin Journl of Soil Science 67: Deibert E J, French E, Hog B Wter storge nd use by spring whet under conventionl tillge nd no-till in continuous nd lternte crop-fllow systems in the northern Gret Plins. Journl of Soil nd Wter Conservtion 41: Frhni H J, Peterson G A, Westfll D G Drylnd cropping intensifiction: A fundmentl solution to efficient use of precipittion. Advnces in Agronomy 64: Greb B W Reducing drought effects on croplnds in the west-centrl Gret Plins. USDA Informtion Bulletin. No. 42. Wshington, DC 242: US Government Printing Office. 31 pp. Greb B W, Smik D E, Blck A L Effects of strwmulch rtes on soil wter storge during summer fllow in the Gret Plins. Soil Science Society Americ Proceedings 31: Hs H J, Willis W O Moisture storge nd use by drylnd spring whet cropping systems. Soil Science Society Americ Proceedings 26: Hs H J, Willis W O, Bond J J Introduction. In Summer Fllow in the Western United Sttes. USDA Conservtion Reserch Report No. 17, p Wshington, DC 242: US Goverment Printing Office. 16 pp. Jones O R, Smith S J, Southwick L M Tillge system effects on wter conservtion nd runoff wter qulity - Southern Gret Plins Drylnds, pp In Proceedings of the Gret Plins Residue Mngement Conference, August Amrillo, Texs. Gret Plins Agriculturl Council Bulletin No. 54. Lincoln, Nebrsk: Gret Plins Agriculturl Council. Jones R, Johnson G L Cropping nd tillge systems for drylnd grin production. USDA-ARS. Report No Conservtion nd Production Reserch Lbortory, Bushlnd, TX. Wshington, DC: USDA

12 138 GARY A PETERSON & DWAYNE G WESTFALL Kn D A, O Brien D M, Burgener P A, Peterson G A, Westfll D G. 22. An economic evlution of lterntive crop rottions compred to whet-fllow in Northestern Colordo. Technicl Bulletin. TB2-1. Fort Collins, CO: Agriculturl Experiment Sttion Colordo Stte University. McGee E A, Peterson G A, Westfll D G Wter storge efficiency in no-till drylnd cropping systems. Journl of Soil nd Wter Conservtion 52: Norwood C A Profile wter distribution nd grin yield s ffected by cropping system nd tillge. Agronomy Journl 86: Peterson G A, Westfll D G, Cole C V Agroecosystem pproch to soil nd crop mngement reserch. Soil Science Society of Americ Journl 57: Peterson G A, Schlegel A J, Tnk D L, Jones O R Precipittion use efficiency s ffected by cropping nd tillge systems. Journl of Production Agriculture 9: Peterson G A, Westfll D G, Tomn N E, Anderson R L. 1993b. Sustinble drylnd cropping systems: Economic nlyses. Technicl Bulletin TB93-3. Fort Collins, CO: Agriculturl Experiment Sttion Colordo Stte University. Peterson G A, Westfll D G, Peirs F B, Sherrod L, Poss D, Gngloff W, Lrson K, Thompson D L, Ahuj L R, Koch M D, Wlker C B. 21. Sustinble drylnd groecosystem mngement. Technicl Bulletin TB1-2. Fort Collins, CO: Agriculturl Experiment Sttion Colordo Stte University. Schlegel A J Effect of cropping system nd tillge prctices on grin yield nd soil wter ccumultion nd use. In Conservtion Tillge Reserch. Knss Agriculturl Experiment Sttion Report of Progress 598, p Eds J S Hickmn nd E Schofield. Shver T M, Peterson G A, Sherrod L A, Ahuj L R. 23. Cropping intensifiction in drylnd systems improves soil physicl properties: Regression reltions. Geoderm 116: Shver T M, Peterson G A, Ahuj L R, Westfll D G, Sherrod L A, Dunn G. 22. Surfce soil properties fter twelve yers of drylnd no-till mngement. Soil Science Society of Americ Journl 66: Sherrod, L A, Peterson G A, Westfll D G, Ahuj L R. 23. Cropping intensity enhnces soil orgnic crbon nd nitrogen in no-till groecosystem. Soil Science Society of Americ Journl 67: Smik D E Fllow mngement prctices for whet production in the Centrl Gret Plins. Agronomy Journl 82: Smik, D E, Wicks G A Soil wter during fllow in the Centrl Gret Plins s influenced by tillge nd herbicide tretments. Soil Science Society Americ Proceedings 32: Steel R G D, Torrie J H Principles nd Procedures of Sttistics. 3 rd Edn. New York, USA: McGrw-Hill Book Co. 666 pp. Tnk D L Spring whet plnt prmeters s ffected by fllow methods in the Northern Gret Plins. Soil Science Society of Americ Journl 53: Unger P W Strw-mulch rte effect on soil wter storge nd sorghum yield. Soil Science Society of Americ Journl 42: