J Plnt Growth Regul (1994) Journl of Plnt Growth Regultion Heriidl Ativity of Hydrolyzed Corn Gluten Mel on Three Grss Speies under Controlled Environments Dinn Ln-Ying Liu, 1 Nik E. Christins, 1 * nd John T. Grutt 2 1 Deprtment of Hortiulture, Iow Stte University, Ames, Iow 50011; nd 2 Grin Proessing Corportion, Mustine, Iow 52761, USA Reeived Novemer 29, 1993; epted August 19, 1994 Astrt. US Ptent No. 5,030,268 disloses tht orn gluten mel, the protein frtion of orn (Ze mys L.) grin, n e used s nturl preemergene heriide. However, orn gluten mel is insolule in wter, nd this hrteristi renders it diffiult to pply s heriide. To seek wter solule mteril with more potent heriidl tivity, the phytotoxiity of vrious smples derived from orn gluten mel nd other relted rop mterils were evluted y using three different grss speies under ontrolled environments. Greenhouse nd growth hmer iossys showed tht the smple of enzymtilly hydrolyzed orn gluten mel ws more heriidlly tive thn the orn gluten itself nd ws highly wter solule. Gluten hydrolyste prepred with teril soure proteinse hd the gretest inhiitory tivity to the root growth of germinting seeds. This wter-solule mteril derived from orn gluten mel hd growth-regulting effet on the root system nd n e used s nturl heriide. The use of heriides for weed ontrol hs een prtied to some extent sine 1927. Syntheti nd orgni pestiides hve eome mjor omponent of griulturl, forestry, nd turfgrss mngement systems (Blogh nd Anderson 1992). Although hemils with inresed effiy nd sfety hve een developed in reent yers, the grohemil industry fes ontinuing nd growing ritiism over the toxiity nd residue prolems of pestiides (Hupptz 1990). Beuse mny pestiides re toxi nd ll re iologilly tive y design, there is onern regrding their effets on humn helth nd environmentl qulity. Environmentl nd humn helth prolems ssoited with pestiides hve een doumented (Blogh nd Anderson 1992). The judiious pplition of heriides hs eome n integrl prt of griulture (Pimentel 1986). Nturlly ourring iotive ompounds, prtiulrly those shown to e nontoxi, would e useful in integrted weed-ontrol progrms. The growing wreness of the environmentl nd puli helth onsequenes of griulturl hemil use hs stimulted interest in the serh for new, environmentlly sfe heriides (Lydon nd Duke 1987). The use of nturl pestiides my help to hnge the trditionl pprohes to griulturl siene nd open the wy to the development of sustinle systems (Lovett 1991). The ojetive of this study ws to serh for heriidlly tive nd wter-solule mterils derived from orn gluten mel. US Ptent No. 5,030,268 disloses tht orn gluten mel, whih is yprodut of orn from the wet-milling proess, n e used s nturl preemergene heriide (Christins 1991). Corn gluten mel n inhiit the estlishment of germinting weeds y preventing root formtion of germinting plnts (Christins 1993), ut orn gluten mel is quite insolule in wter nd this hrteristi renders it diffiult to pply s heriide. If wter-solule, orn gluten mel relted mterils ould e developed nd used s n lterntive to onventionl heriides, they would e environmentlly nd eonomilly desirle. They ould e used s sustitute for syntheti hemil heriides or s supplement to syntheti heriides to redue the onentrtion of these pestiides in the environment. The phytotoxiity of vrious smples derived from orn gluten mel nd other relted rop mterils were evluted y using three different grss speies grown in the greenhouse nd growth hmer. Journl Pper No. J-15609 of Iow Agriulture nd Home Eonomis Experiment Sttion, Ames, IA Projet No. 3149 * Author for orrespondene
222 D. L.-Y. Liu et l. Tle 1. List of smples derived from orn gluten mel nd other rop mterils. Smple Desription 1 orn gluten mel 2 orn gluten mel treted with fungl proteinse 3 orn gluten mel treted with teril proteinse 4 orn gluten filtrte 5 lohol-solule frtion of orn gluten filtrte 6 lohol-insolule frtion of orn gluten filtrte 7 orn gluten filtrte treted with tion resin 8 orn gluten mel mixed with orn hulls nd germs 9 orn gluten mel without orn hulls nd germs 10 orn gluten mel treted with teril proteinse 11 orn gluten hydrolyste, totl ion exhnged 12 solule orn steep liquor solids 13 insolule orn steep liquor solids 14 whet gluten 15 whet gluten hydrolyste 16 orn gluten mel treted with teril proteinse 17 soyen mel Seventeen different smples derived from orn gluten mel or other grin mterils (Tle 1) were otined from Grin Proessing Corportion (Mustine, IA, USA). A numer ws ssigned to eh smple ording to the order of rrivl. Smples were tested in greenhouses nd growth hmers for their inhiitory effets on test plnts. Study I Seven orn gluten mel-relted smples with designted numers 1-7 were used to study the effets on germinting seeds under ontrolled environmentl onditions (Tle 1). Smooth rgrss (Digitri ishemum Shre.) ws the test speies in the greenhouse iossys. Plsti pots with surfe re of 58 m 2 nd depth of 5 m were filled with Niolett (fine-lomy mixed mesi Aqui Hpludolls) soil tht hd ph of 7.5. Seeds of rgrss weighing 0.11 g were spred evenly on top of the soil. Preweighed, dry smples of the test mterils were pplied uniformly on the soil surfe t rtes of 0, 0.86, 1.72, 3.44, nd 6.88 g/dm 2, exept smple 6 tht did not hve enough mteril for the 3.44 nd 6.88 g/dm 2 rtes. Three replites of the tretments nd ompletely rndomized experimentl design were used. All pots were pled rndomly on mist enh where onstnt moisture ws ville for 6 dys. After germintion, the pots were moved to greenhouse enh nd wtered every 2 to 3 dys y using fine-mist nozzle. The greenhouse temperture ws mintined within rnge of 18 to 32 C. Dt were olleted on the numer of live grss shoots t 18 dys fter seeding (DAS). Fresh lipping weights were tken t 35 DAS. These two mesurements from eh pot were verged for eh tretment. Smooth rgrss nd reeping entgrss (Agrostis plustris Huds.) were used for the Petri dish iossys. The seven test smples were pplied t rtes of 0, 0.118, 0.236, 0.355, nd 0.473 g/dm 2 to two lyers of seed-germintion lotter pper (Pkging Converters, Hudson, WI, USA) mesuring 42.25 m 2. Before the tretment, the lotter pper ws pled in 9-m-dimeter Petri dish (Fisher Sientifi Co., Pittsurgh, PA, USA) nd moistened with 7 ml of deionized distilled wter (D.D. H 2 O). Eighteen smooth rgrss seeds were pled on the lotter pper. The dishes were overed with lid nd seled with Prfilm. They were inuted in growth hmer t dy/night temperture of 25 ± 0.5 C. A 16-h photoperiod ws used, nd rdition levels were mintined t 70 µmol/s/m 2 with fluoresent lmps. After 21 dys, the germintion rte, expressed s perentge, ws derived y dividing the numer of germinted seeds y 18, the numer of seeds used in eh plte, nd multiplying tht numer y 100. Study 2 In the seond study, the effets of six orn gluten mel-derived smples numered 8-13 were iossyed y using reeping entgrss nd perennil ryegrss (Lolium perenne L.; Tle 1). Plsti pots filled with the sme type of soil s in study I were used for the iossys. Creeping entgrss ws seeded on the 64-m 2 surfe re soil t rte of 0.16 g/dm 2. Smple mterils were pplied in three replites t rtes of 0, 0.78, 1.56, 3.13, 4.69, nd 6.25 g/dm 2 for ll smples exept for smple 13, whih hd only enough mteril for three levels of pplition, 0.78, 1.56, nd 3.13 g/dm 2. All pots were rndomly pled in the greenhouse fter 6 dys on the mist enh. The environments nd irrigtion method were mintined with the sme onditions s in study 1. Dt were tken t 28 DAS on fresh lipping weight of live plnts. Perennil ryegrss ws seeded t 1.56 g/dm 2. The sme seven rtes used in the reeping entgrss iossy s desried for smples 8-11 were used. Smples 12 nd 13 were not tested euse of insuffiient mteril. The study ws onduted s desried for study 1. Dt were olleted s fresh lipping weight t 27 DAS. Perennil ryegrss seeds were lso used to test the iotivity of these six smples, 8-13, in Petri dish iossy. A volume of 6 ml of D.D. H 2 O ws pplied to two lyers of lotter pper mesuring 36 m 2. The test mterils were pplied t 0, 0.028, 0.056, 0.139, 0.222, nd 0.278 g/dm 2 to the surfe of the lotter pper in the Petri dishes. The tretments hd three replites, nd the ontrol (0 g/dm 2 ) hd six. Sixteen perennil ryegrss seeds were pled on the treted lotter pper. All 36 dishes were rndomly pled in growth hmer. Irrdine ws from fluoresent light mintined t 70 µmol/s/m 2. A dy/night temperture of 25/15 C nd 16-h photoperiod were mintined. The numer of norml seedlings, whih were defined s hving roots longer thn 0.5 m, were ounted t 14 DAS. Study 3 In the third study, perennil ryegrss ws used to investigte the effets of smples 14-17 on seedling growth in the greenhouse (Tle 1). The sme type of soil s desried in study I ws used. Perennil ryegrss ws seeded t 0.78 g/dm 2 on 64-m 2 plsti pots. The four test smples were evluted t the rtes of 0, 0.78, 1.56, 3.13, 4.69, nd 6.25 g/dm 2 in three replites. Pots were kept in the greenhouse nd wtered regulrly in the sme mnner
Nturl Heriide from Corn Gluten Mel 223 Tle 2. Effet of mterils derived from orn gluten mel on the estlishment of rgrss grown on soil in the greenhouse. Amount of smple mteril ------------------------------------------------------------- S l 0 86 /d 2 172 /d 2 344 /d 2 1 51 (2.58) 31 (1.11) 2 49 (2.69) 47 (1.78) 3 23 (1.48) 4 (0.42) 4 52 (2.73) 43 (1.36) 5 55 (3.30) 37 (3.73) 6 100 (3.95) 61 (2.49) 7 47 (1.21) 25 (1.56) MSD C 9 (0.77) 9 (0.53) Effet of eh tretment ws expressed s the numer of plnts; fresh lipping weight in grm per pot in prentheses. Numer is the men of three replites for eh of four tretment levels of smples. The ontrol hd 95 ± 12 plnts, nd fresh lipping weight ws 1.24 ± 0.54 g. MSD is the men of stndrd devitions of the vlues in eh olumn. The LSD (0.05) were 21 nd 1.28 g for numer of plnts nd fresh lipping weight, respetively. Tle 3. Effet of mterils derived from orn gluten mel on germintion of test grsses iossyed in the growth hmer. % Germintion 0.118 0.236 0.355 0.473 Smple g/dm 2 g/dm 2 g/dm 2 g/dm 2 1 89 (83) 78 (78) 56 (72) 56 (61 2 83 (89) 72 (61) 61 (61) 28 (28 3 6 (61) 6 (11) 0 (0) 0 (0) 4 61 (17) 39(56) 6(28) 0(0) 5 50 (22) 11 (44) 0 (0) 0 (0) 6 72 (61) 56 (22) 22 (17) 11 (6) 7 39 (61) 11 (6) 0 (0) 0 (0) MSDd 8 (7) 4 (4) 3 (3) 2 (2) A germinted seed ws defined s seed with protruding rdile or shoot. Crgrss nd reeping entgrss were tested; the germintion of reeping entgrss is in prentheses. Numer is the men of three replites for eh of four tret ment levels of smples. Control germintion ws 67 ± 5% for rgrss nd 61 ± 9% for reeping entgrss. d MSD is the men of stndrd devitions of vlues in eh olumn. The LSD (0.05) were 12 nd 15% for rgrss nd reep ing entgrss, respetively. s desried in study 1. Dt were olleted on fresh lipping weight of eh tretment t 28 DAS. Smples of 14-17 (Tle 1) were tested in Petri dish iossys t rtes of 0, 0.028, 0.056, 0.139, 0.222, nd 0.278 g/dm 2. The dry smples were pplied to the top of two lyers of 36-m 2 lotter pper tht were presoked with 6 ml of D.D. H 2 O. Sixteen perennil ryegrss seeds were pled on the lotter pper, nd the dishes were seled with Prfilm. All seled dishes, inluding four ontrols, were inuted for 15 dys in the growth hmer under the sme onditions s in the previous evlution. Dt were ontrolled on the numer of norml seedlings with root growth greter thn 0.5 m t 15 DAS. Sttistil Anlysis An nlysis of vrine (ANOVA) ws onduted on numer of plnts, fresh lipping weight, nd germintion vriles for eh study with the SttView omputer softwre progrm (Aus Conepts, Berkeley, CA, USA). Mens were seprted y the projeted lest-signifint differene (LSD) y using Fisher's test (Snedeor nd Cohrn 1989). Results nd Study 1 Compred with the numer of plnts in the ontrol, ll seven smples, exept the lohol-insolule frtion of gluten filtrte (smple 6), redued rgrss estlishment t the four test rtes (Tle 2). Corn gluten mel, smple 1, ws less heriidlly tive thn the teril proteinse hydrolyzed gluten, smple 3, t the two lowest rtes. The teril pro- teinse hydrolyste tht redued the estlishment of rgrss seedlings y 76, 96, 98, nd 100% t the rtes of 0.86, 1.72, 3.44, nd 6.88 g/dm 2, respetively, ws the most tive of the test smples. Seprte tests of the proteinse showed no effets of this mteril on seed germintion (dt not shown). On the sis of the fresh lipping weight dt t 35 DAS, ll seven smples hd stimultory effet on rgrss growth t the rte of 0.86 g/dm 2 (Tle 2). The stimultion ws likely due to nitrogen (N) response (Christins 1993). At rtes of 3.44 nd 6.88 g/dm 2, severl tretments redued fresh lipping weight, with the teril proteinse hydrolyzed gluten smple (3) eing the most effetive. In Petri dish tests, the teril proteinse hydrolyzed gluten smple (smple 3) nd the gluten filtrte treted with tion resin (smple 7) inhiited the germintion of rgrss t ll test rtes nd of reeping entgrss t rtes higher thn 0.118 g/dm 2 (Tle 3). The fungl proteinse hydrolyste (smple 2) inhiited germintion only t the rte of 0.473 g/dm 2. Corn gluten mel (smple 1) ws less tive thn the teril proteinse hydrolyzed gluten, smple 3, for the inhiition of oth speies. The definition of germintion in the Petri dish study inluded ny seed with protruding rdile or shoot. This my hve een misleding euse mny of the seeds ounted s germinted seed hd gretly redued root growth. Therefore, in the lter
224 D. L.-Y. Liu et l. Tle 4. Effet of mterils derived from orn gluten mel on the fresh lipping weight of reeping entgrss nd perennil ryegrss grown on soil in the greenhouse. Fresh lipping weight (g) Smple 0.78 g/dm 2 1.56 g/dm 2 3.13 g/dm 2 4.69 g/dm 2 6.25 g/dm 2 8 1.48 (0.66) 1.15 (0.73) 0.51 (0.68) 0.07 (0.63) 0 (0.76) 9 1.86 (1.01) 1.21 (0.81) 0.16 (1.07) 0 (1.56) 0 (1.05) 10 2.24 (0.81) 1.02 (1.23) 0.32 (1.08) 0 (0.76) 0 (0) 11 1.49 (1.08) 1.64 (1.25) 1.19 (1.15) 0.46 (1.82) 0.54 (1.13) 12 1.94 1.38 1.03 2.51 1.98 13 2.25 1.66 0.42 MsD 0.31 (0.08) 0.17 (0.09) 0.04 (0.08) 0.08 (0.08) 0.05 (0.06) Fresh lipping weight of perennil ryegrss is in prentheses. Smple 13 ws not evluted t the two highest rtes euse of the lk of mteril for testing. Numer is the men of three replites for eh level of smples. The ontrol hd 0.85 ± 0.13 g for reeping entgrss nd 0.78 ± 0.07 g for perennil ryegrss. MSD is the men of stndrd devitions of vlues in eh olumn. The LSD (0.05) were 0.39 nd 0.23 g for reeping entgrss nd perennil ryegrss, respetively. studies, germintion ws redefined to inlude ny rdile mesuring 0.5 m or longer. Study 2 Plnt response in the greenhouse study ws sed on the fresh lipping weight of reeping entgrss plnts t 28 DAS (Tle 4). All smples stimulted growth t rtes of 0.78 nd 1.56 g/dm 2. The inhiition effet ws seen t rtes of 3.13 g/dm 2 nd higher. The teril proteinse hydrolyste of orn gluten solution (smple 10), whih ws equivlent to smple 3 in study 1, ws the most effetive inhiitor in this study. The orn steep liquor smples 12 nd 13, were ineffetive nd stimulted growth of reeping entgrss t most rtes. Perennil ryegrss ws lso used s iossy speies euse of its rpid germintion. It generlly required greter mounts of the smples to redue plnt estlishment of the ryegrss thn of the entgrss (Tle 4). The teril proteinse hydrolyste smple (smple 10) strted showing inhiition t rtes higher thn 3.13 g/dm 2 on reeping entgrss nd 4.69 g/dm 2 on perennil ryegrss. It ompletely inhiited estlishment of ryegrss t the rte of 6.25 g/dm 2 nd ws the most effetive of the mterils tested. Germinted seeds in the Petri dish study were defined s those with rdile longer thn 0.5 m extruding from the seed ot. The gluten hydrolyste smple (smple 10) ws the most tive mong ll six test smples in this study (Tle 5). It inhiited germintion t ll test rtes nd stopped germintion t rtes of 0.056 g/dm 2 nd higher. The totl ion-exhnge gluten hydrolyste smple (smple Tle 5. Effet of mterils derived from orn gluten mel on germintion of perennil ryegrss iossyed in the growth hmer. % Germintion 0.028 0.056 0.139 0.222 0.278 Smple g/dm 2 g/dm 2 g/dm 2 g/dm 2 g/dm 2 8 81 100 38 0 0 9 100 100 94 50 4 10 12 0 0 0 0 11 62 12 0 0 0 12 75 88 100 94 88 13 100 94 12 0 0 MSD C 6 6 4 2 1 A germinted seed ws defined s seed with rdile longer thn 0.5 m. Numer is the men of three replites. The ontrol germin tion of perennil ryegrss ws 94 ± 5% (n = 6). MSD is the men of stndrd devitions of vlues in eh ol umn. The LSD (0.05) ws 12%. 11) hd slightly less tivity thn the gluten hydrolyste (smple 10). Two gluten mel smples (8 nd 9) were less tive thn gluten hydrolyste. The solule orn steep liquor smple (smple 12) hd no inhiitory effet t ll test rtes, wheres the insolule orn steep liquor (smple 13) inhiited germintion t rtes higher thn 0.139 g/dm 2. Study 3 In study 3, smples derived from orn gluten, whet gluten, nd soyen were iossyed y using perennil ryegrss, whih hd more rpid germintion nd served s more onservtive speies thn the
Nturl Heriide from Corn Gluten Mel 225 Tle 6. Effet of grin-derived mterils on fresh lipping weight of perennil ryegrss tested on soil in the greenhouse. Fresh lipping weight (g) 0.78 1.56 3.13 4.69 6.25 Smple g/dm 2 g/dm 2 g/dm 2 g/dm 2 g/dm 2 14 2.20 1.99 2.10 2.08 1.87 15 2.39 1.99 2.25 2.01 1.81 16 3.16 2.45 1.04 0 0 17 3.87 2.80 1.76 0 0 MSD 0.45 0.34 0.25 0.13 0.10 Tle 7. Effet of grin-derived mterils on germintion of perennil ryegrss iossyed in the growth hmer. % Germintion 0.028 0.056 0.139 0.222 0.278 Smple g/dm 2 g/dm 2 g/dm 2 g/dm 2 g/dm 2 14 100 100 100 88 44 15 94 100 62 19 6 16 94 12 0 0 0 17 69 62 19 0 0 MSDC 7 6 4 2 1 Numer is the men of three replites for eh level of smples. The ontrol hd 1.78 ± 0.44 g (n = 4). A germinted seed ws defined s seed with rdile longer thn 0.5 m. Numer is the men of three replites for eh level of smples. The ontrol germintion of perennil ryegrss ws 96 +± 6% (n = 4) MSD is the men of stndrd devitions of vlues in eh olumn. The LSD (0.05) ws 12%. MSD is the men of stndrd devitions of vlues in eh olumn. The LSD (0.05) ws 0.86 g. other two test grss speies. All test smples hd greter fresh lipping weight thn the ontrol t 28 DAS t the rtes of 0.78 nd 1.56 g/dm 2 (Tle 6). Corn gluten hydrolyste (smple 16) nd soyen mel (smple 17) strted to show inhiition of plnt growth t the rte of 3.13 g/dm 2 nd ompletely stopped plnt estlishment t higher levels of tretment. Whet gluten (smple 14) nd its hydrolyste (smple 15) showed no inhiition. In the Petri dish study, the soyen mel nd orn gluten hydrolyste showed inhiition t ll test rtes, nd orn gluten hydrolyste ompletely inhiited rdile emergene t rtes of 0.139 g/dm 2 nd higher (Tle 7). However, the whet gluten showed no inhiition t the rte of 0.139 g/dm 2, nd its hydrolyste hd less thn 40% redution in germintion of perennil ryegrss. The whet gluten hydrolyste hd more inhiitory tivity thn whet gluten itself. meets with rtes ove 4.69 g/dm 2 of the gluten hydrolystes resulted in omplete inhiition of plnt estlishment. However, rtes of 6.88 g/dm 2 were required for the orn gluten mel to hieve omplete inhiition. In the Petri dish iossy, the orn gluten hydrolyste prevented germintion of reeping entgrss t pplition levels ove 0.118 g/dm 2 nd prevented germintion of rgrss t pplition levels ove 0.236 g/dm 2, s seen in study 1 (smple 3). The sme mteril ws tested o n perennil ryegrss in study 2 (smple 10), nd it ompletely inhiited root emergene of perennil ryegrss seed t 0.056 g/dm 2. The Petri dish iossy required less of the smple thn the soil iossy to inhiit germintion. The Petri dish iossy is fster nd more eonomil method to monitor inhiitory tivity thn soil iossy. Root-length redution of seedlings ws the est mesurement of heriidl tivity. The teril proteinse gluten hydrolyste ws the most highly iotive wter-solule mteril. This mde it suitle for the isoltion nd identifition of the iotive ompounds. This mteril lso hs the potentil of eing used s nturl heriide, nd further work to evlute its effetiveness in vrious ropping systems is plnned. Conlusion It ws demonstrted in ll three studies tht the orn gluten hydrolyste ws more effetive inhiitor thn orn gluten mel. The orn gluten hydrolyste is proteinse-hydrolyzed produt of orn gluten mel. It is formed y proess of treting the orn gluten slurry with mylses nd then treting it with one or more proteinses, followed y the removl of the wter (Christins et l. 1993). The proteinse used to hydrolyze orn gluten mel did not ontriute to the inhiition of the germintion of perennil ryegrss (dt not shown). On the sis of the results from four studies, oth in the greenhouse nd growth hmer, the orn gluten hydrolyste prepred with teril proteinse ontined the most potent inhiitory tivity for three test grss speies. In soil iossys, tret Referenes Blogh JC, Anderson JL (1992) Environmentl impts of turfgrss pestiides. In: Blogh JC, Wlker WJ (eds) Golf ourse mngement nd onstrution Environmentl issues. Lewis Pulishers, Chelse, pp 221-354 Christins NE (1991) Preemergene weed ontrol using orn gluten mel. US Ptent No. 5,030,268
226 D. L.-Y. Liu et l. Christins, NE (1993) The use of orn gluten mel s nturl preemergene weed ontrol in turf. In: Crrow RN, Christins NE, Shermn RC (eds) Interntionl Turfgrss Soiety Reserh Journl. Interte Pulishing Corp., Overlnd Prk, pp 284-290 Christins NE, Grutt JT, Liu DL-Y (1993) Preemergene weed ontrol using plnt protein hydrolyste. US Ptent No. 5,290,749 Hupptz JL (1990) Essentil mino id iosynthesis provides multiple trgets for seletive heriides. In Csid JE (ed) Pestiides nd lterntives: Innovtive hemil nd io- logil pprohes to pest ontrol. Elsevier, New York pp 563-572 Lovett JV (1991) Chnging pereptions of llelopthy nd iologil ontrol. Biol Agri Hort 8:89-100 Lydon J, Duke SO (1987) Progress towrd nturl heriides from plnts. Hers Spies Med Plnts Dig 5:1-4 Pimentel D (1986) Sttus of integrted pest mngement. In Pimentel D (ed) Some spets of integrted pest mngement. Cornell University Press, Ith, pp 29-68 Snedeor GW, Cohrn WG (1989) Sttistil methods, 8th ed. Iow Stte University Press, Ames