JASP 1 -- 1983.pdf - International Herbage Seed Group

Journal of AppliedSeed ProductionVolume I (1) 1983Table of ContentsReduction of Hardseededness in Crownvetch (Coronilla Varia L.) Seed Lots.A. E. Garay and D. F. Grabe ... ... . ..... . ...... ... ........... .. ... .......... .... . .... .A Proposed Severity Index for Seed Handling.N. Robert Brandenburg . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 8The Effect of Cutting in Spring and Application of Alar on Red Clover(Trifolium pratense L.) Seed Production.K. P. Puri and A. S. Laidlaw ........ . ... .... . . . . ... ................ . . .... . . .. ........ 12Effects of Sunlight Reduction and Post-Harvest Residue Accumulationson Seed Yields of Kentucky BluegrassR. D . Ensign, V. G. Hickey and M . D. Bernardo .... . ............. ..... ..... . . ..... . . .... 19The Effects of the Environment at Anthesis on the Seed Yield and Yield Componentsof Perennial Ryegrass (Lolium perenne L.) cv. S. 24.J. G. Hampton and P. D. Hebblethwaite . . ... ...... . ................... ............. ..... 21Yield Components of the Perennial Ryegrass (Lolium perenne L.) Seed Crop.J. G. Hampton and P. D. Hebblethwaite ..... ......... .. ................ .. . . ....... . .. ... 23Preliminary Investigations of Adhesive Sprays to Improve Seed Retention in Tropical GrassesD. S. Loch and G. L. Harvey . ..... ...... . ........... .. . . ......... ... ..... . ........... 26The Effects of Method and Time of Sowing on Seed Production in Perennial RyegrassP. D. Hebblethwaite and S. D. Peirson . ... ....................... ... ..... . . . ............ 30Growth, Floral Induction and Reproductive Development in Selected Perennial Ryegrass(Lolium perenne L.) Cultivars.A. S. Gangi, D. 0. Chilcote, and R. V. Frakes ............. . . ........ ................... 34Chemical Dwarfing Effects on Seed Yield of Tall Fescue (Festuca arundinacea) cv. Fawn, FineFescue (Festuca rubra) cv. Cascade, and Kentucky Bluegrass (Poa pratensis) cv. Newport.D. W. Albeke, D. 0. Chilcote, and H. W. Youngberg .............. . ...... .......... ... . .. 39Reproductive Growth and Development in Selected Kentucky BluegrassCultivars Under Different Environmental Conditions.J. K. Turner, D. 0. Chilcote and R. V. Frakes .... . ....... .... ......... .. ... . . . ... . ..... .43Effects of Chemical Dwarfing Application Under Different Nitrogen Levels on Seed Yieldof Fine Fescue (Festuca rubra) cv. Cascade.D. W . Albeke, D. 0. Chilcote and H. W. Youngberg ..... ..... . .... .. . . .... ... . . . ... .... .47Page

JOURNAL OF APPLIED SEED PRODUCTIONEDITORIAL BOARDHarold W. Youngberg, Editor,Oregon State University, U.S.A.David 0. Chilcote,Oregon State University, U.S.A.Donald F. Grabe, Oregon State University, U.S.A.Paul D. Hebblethwaite,University of Nottingham, U.K.Murray J. Hill, Massey University, N.Z.Vern L. Marble, University of California, U.S.A.Editorial and subscription correspondenceHarold Youngberg, EditorJournal of Applied Seed ProductionCrop Science DepartmentOregon State UniversityCorvallis, Oregon 97331SubscriptionsThe Journal of Applied Seed Production ispublished annually. Subscription rate: US $10.00postpaid surface mail, US $15.00 airmail toforeign address. Subscription requests should beaddressed to the Editor at the above address.From The Editor's CornerThe need for a publication to communicateresults of seed production research amongscientists, specialists and producers on an internationalbasis has been discussed for some time.Much of the currently published work is availableon a regional or national basis but the informationmay also have a much broader interest andapplication.The International Herbage Seed ProductionResearch Group (IHSPRG), organized during theEaster School Program at the University ofNottingham in 1979 brought together scientistsfrom around the world with scientific interest inseed production. Under the leadership of its firstchairman, Dr. Paul D. Hebblethwaite of theUniversity of Nottingham, U.K., IHSPRG startedan exchange of information through its newsletter.Dr. David 0. Chilcote, the present IHSPRG chairmanhas continued this important form of communication.The newsletter can and will continueto serve a valuable role, but some felt there wasstill the need for a publication with a morescientific format to supplement the newsletter.In 1981, An Oregon seed growers organization,the Oregon Seed Growers League,expressed a willingness to support a publicationof research information pertaining to seed production.They have underwritten a grant insupport of the publication of a Journal of AppliedSeed Production. The objective is to broaden theexchange of ideas among scientists, interestedspecialists, and seed producers. For their support,the Editor wishes to thank the following Directorsof the Oregon Seed Growers League for theirencouragement and financial support:Mark Hagman Ron von der HellenDon Wirth William PanikeDon Fisher George Royes Jr.Don Bowers John RossnerFred Kaser Toby IrwinOther Financial support came from IHSPRGwhich in turn is financed by its subscriptionmembers. They are listed at the end in thisissue. Our appreciation is extended to them fortheir support of IHSPRG and its activities.An Editorial Board was selected from severalmajor seed producing areas to guide the policiesof the Journal and to encourage support of thepublication. It was their recommendation thatinitially the publication be issued on an annualbasis.Subject matter considered suitable for thepublication includes topics related to seedproduction technology in the broad sense. Seedquality control and research methodology are alsosuitable subjects. Seed laboratory methods, seedtesting and other topics already adequatelycovered by existing Journals will receive lessattention. Papers submitted will be subject to peerreview.A preliminary survey of interest indicatedconsiderable support for the concept (see lettersto the editor). An invitation for submission ofpapers through the IHSPRG Newsletter brought anumber of inquiries and papers. I trust this firstissue will stimulate more interest on the part ofother scientists to support the Journal with papers.A section for brief papers in a more narrativestyle and review articles is being considered forfuture issues. Letters to the editor are invitedcommenting on subject matter, format, or content.The Editor

Reduction of Hardseededness in Crownvetch (Coronilla Varia L.)Seed Lots 1A. E. Garay and D. F. Grabe 2ABSTRACTThe use of crownvetch (Coronilla varia L.) as a forage cropmight be increased considerably if seed with low hard seedcontent could be made available at reasonable cost. Sincemechanical scarification damages a large percentage of crownvetchseeds, this study was initiated to find other ways ofreducing hardseededness. Two approaches to the problem werestudied: (a) separation of hard and soft seeds on the basis ofdifferences in physical properties, and (b) boiling water scarificationfollowed by separation of soft seeds after soaking inwater.Attempts to separate hard and soft seeds on the basis oflength and width were only partially successful. Long-narrowseeds contained up to 33% less hard seeds than short-wide seedsin some lots, but this relationship did not exist in other lots.Immersion for 5 seconds in boiling water reduced hardseedednesswithout lowering the viability of soft seeds. Thepercentage of hard seeds remaining after a 5-second treatmentwas further lowered by a second treatment, again withoutreducing viability. It appears that the abrupt change intemperature, rather than long exposure to high temperature, isimportant in breaking hardseededness. The viability of hardseeds averaged 24% higher than that of naturally soft seeds.Soft seeds were successfully separated from hard seeds, aftersoaking in tap water, by removing the swollen soft seeds withround-hole screens. After separation, swollen soft seeds can beair-dried to their initial moisture content without apparentdamage. Soaked and dried seeds showed reduced dormancy,good storability, and good emergence from soil in greenhousestudies.Techniques for boiling water scarification and separation ofhard and soft seeds are suggested.Additional index words: scarification, seed size, seed conditioning,dormancy.----~--------~-----------INTRODUCTIONCrownvetch (Coronilla varia L.) is a perennial leguminousplant used for erosion control and beautification, butwhich has a limited use as a forage crop. High percentagesof hardseededness necessitate high seeding rates which leadto excessive costs of stand establishment. Use as a forageI. Contribution of the Oregon Agric. Exp. Stn., Corvallis.Technical paper no. 6690. Received 19 January 1983.2· Former graduate research assistant and professor of agronomy,Crop Science Dept., Oregon State Univ., Corvallis, OR 97331.crop might be increased considerably if seed lots with lowhard seed content could be made available at reasonablecost. Most crownvetch seed lots are mechanically scarifiedto lower the hard seed content, but this practice damages alarge percentage of the seed because of their narrow oblongshape. (Peiffer, McKee and Ditmer, 1972)Hard seed content of crown vetch seed lots could conceivablybe reduced by development of a technique for removalof hard seeds with conventional conditioning equipment.This would require a physical difference between hard andsoft seeds that could be sensed by the equipment. Suchdifferences exist to some extent in several small-seededlegumes. Vaughan (1962) demonstrated that hardseedednesswas associated with small seed size in white clover(Trifolium repens), red clover (T. pratense) and crimsonclover (T. incarnatum), while Middleton (1933) found thesame to be true for Korean lespedeza (Lespedeza stipulata).Another approach is to develop an improved scarificationtechnique that will not lower seed viability. Brant (1968)was successful in lowering hard seed content in crownvetchby soaking seeds in concentrated sulfuric acid for 15minutes, immersing in boiling water for 15 seconds, ordipping in liquid nitrogen for 1 minute. Casey (1939)increased germination of bur clover (Medicago arabica andM. maculata) by immersing the seed in boiling water for 1minute.The primary objectives of this study were: (a) to determinethe degree of association between hardseededness ofcrownvetch seeds and certain physical characteristics(length and width) that can be used as the basis forupgrading the germinability of seed lots and (b) to determinethe feasibility of mechanically separating hard andnon-hard seeds by soaking them to artificially create sizedifferences that would allow them to be separated by seedsizing equipment.MATERIALS AND METHODSSeed LotsTwo lots each of three commonly grown cultivars ofcrownvetch (Coronilla varia L.) were obtained from commercialseed sources. The characteristics of these lots wereas follows: 'Chemung' Lot A- Idaho origin, commercialclass; Chemung Lot B -Idaho origin, commerical class;'Emerald' Lot A- Iowa origin, Certified class; EmeraldLot B- Iowa origin, Certified class; 'Penngift' Lot A­Pennsylvania origin, Foundation class; Penngift Lot B -Pennsylvania origin, Foundation class.

3gallon cans filled with unfertilized river alluvial soil.Temperature in the greenhouse ranged from 22 to 25 C. Theexperiment was set up in a completely randomized designwith each treatment repeated four times. Seedlings werecounted daily for 15 days. Since a number of the seedlingsdied during the period, the figures represent only thenumber of seedlings that survived. Data were analyzed as afactorial arrangement of treatments.RESULTSRelationship of SeedLength and Width to HardseedednessThe relationship of seed length and width to hardseedednessis shown in Table 1. When the hard seed percentagesTable 1. Relation of length and width to percent hard seed incrownvetch. Average of two lots each of three cultivars.WidthLength Wide Intermediate Narrow Average-------------------%-----------------Long 47.33 46.50 43.70Medium 51.91 51.37 46.00Short 56.70 51.83 51.91Average 51.98 49.90 47.20LSD 01for all 9 length-width combinations: 3.58LSD 05for all 9 length-width combinations: 2.7245.8549.7653.48were averaged over the three widths, short seeds containedmore hard seeds than the medium and long seeds. Sinceshort seeds of crownvetch are the smaller seeds, the resultsare in agreement with data presented by Middleton (1933)and Vaughan (1962) which showed that smaller seeds ofKorean lespedeza and clovers were also higher in hard seedcontent. When hard seed percentages were averaged overlengths, wide seeds (which could be considered largerseeds) contained more hard seeds than intermediate ornarrow seeds. This suggests that in this species, hardseedednessis associated with shape of the seed rather thansize itself, with long-narrow seeds containing the lowesthard seed content and the short-wide seeds the highestamount.Variation in hard seed content for the nine length-widthgroups within each seed lot and cultivar is shown in Table2. In Chemung Lot A and B, Emerald Lot A and PenngiftLot B, the wide seed group contained more hard seeds thanthe narrow groups; but in Emerald Lot B and Penngift LotA, this tendency was not present. In every cultivar and lotthere was a definite association between length and hardseededness,short seeds being more hard-seeded than longseeds in each width group.Scarification of Hard Seeds by Boiling WaterEffect of a Single Boiling Water TreatmentThe effects of a single boiling water treatment onTable 2. Relation of length and width to percent hard seed insix lots of crownvetch.WidthCultivar Lot Length Wide Intermediate Narrow------------%----------Long 48.25 44.25 36.00Chemung A Medium 57.50 53.25 38.50Short 68.50 54.00 50.76Long 87.75 82.75 80.75Chemung B Medium 88.00 83.75 85.00Short 91.25 89.75 84.50Long 25.75 22.75 20.00Emerald A Medium 33.25 28.50 22.25Short 38.00 32.25 30.50Long 46.50 55.00 55.75Emerald B Medium 53.75 62.00 55.75Short 54.50 62.00 68.50Long 35.75 36.00 39.00Penn gift A Medium 36.00 40.50 34.75Short 38.00 37.50 40.50Long 41.00 38.25 30.75Penngift B Medium 45.00 40.25 39.75Short 50.00 41.50 42.75hardseededness are shown in Figure 1. After 5 seconds oftreatment, hard seed content decreased about 25% inChemung Lot B and 40% in Emerald Lot B. The decline inhardseededness with longer periods of soaking was negligible.Soaking longer than 10 seconds increased the numberof dead seeds in both seed lots and also decreased the speedof germination (as indicated by the first count germination)in Chemung Lot B. These results indicate that 5 secondswould be the optimum period of treatment to render themost hard seeds permeable without killing any soft seeds.Effect of a Second Boiling Water Treatmenton Remaining Hard SeedsThe hard seeds remaining after a 5-second treatment inboiling water were subjected to a second boiling watertreatment of 10 to 180 seconds (Figure 2).The hard seed content of these seeds decreased withtreatments up to 30 seconds after which no further reductionof hardseededness occurred. Dead seed content was notincreased until the 2-minute period in this second cycle, butthe speed of germination was decreased with periods longerthan 30 seconds. Thus, periods of soaking longer than 30seconds should be avoided in the second cycle of boilingwater scarification.The cumulative changes in amount of hard seeds andgermination after each of the two optimum treatmentperiods are shown in Figure 3. The initial 82% hard seedcontent of Chemung was reduced to 55% with a 5-secondtreatment in boiling water. Since the remaining hard seedswere lowered to 52% with 10 seconds in the second cycle,only 28% (52% x 55%) hard seed would be left in this lot.

4Chemung Lot BEmerald Lot B501:: 40CDu'-CD 30a..0 10 20 30 40 50 600 10 20 30 40 50 60Seconds of boi I i n·g water treatmentsFig. 1. Effects of a single boiling water treatment on crownvetch seeds.8Chemung Lot 8+­c:CDu'­CDa..30200 10 30 60 120 180 010 30 60Seconds of boi I ing water treatmentFig. 2. Effect of a second cycle of boiling water treatments on crownvetch seeds.120 180

570600 Hard seeds~ Germinationseeds. Separation can then be made with the same screensoriginally used to divide the seed lot into width groups.Using this approach, all hard seeds passed through thescreen and the soft seeds remained on the screen, when thescreens used were as follows (Table 3): 1.49 diameterUn- One Two Un· One TwoTreated Cycle Cycles Treated Cycle CyclesChemung Lot BEmerald Lot BFig. 3. Effects of one and two cycles of boiling water treatmenton hard seed content and germination of Chemung Lot B andEmerald Lot B.In Emerald Lot B, the initial 52% was lowered to 12% withthe first cycle and to 3% (12% x 25%) after the secondcycle. Germination increased in proportion to the reductionin hardseededness.The sudden reduction in hardseededness following thetwo short (10 and 30 seconds) soaking periods compared toa single 40 second treatment indicates that it is the abruptchange in temperature rather than long exposure to hightemperature that is important in breaking hardseededness.This further emphasizes that rather than trying to break mosthardseededness with a single deleteriously long treatment inboiling water, more beneficial effects could be achievedwith two or more short cycles. If this approach was to betried on a practical basis, the seed lot should first besubjected to a 5-second treatment in boiling water. If thistreatment does not reduce hardseededness to an acceptablelevel, the seeds that become soft by this treatment should beseparated. Only the remaining hard seeds should be subjectedto a second treatment of 10 to 30 seconds in boilingwater.Mechanical Separation of Hard and Soft-SeedsSeveral techniques or methods were tried with the aim ofseparating hard seeds from swollen soft seeds. The mostsuccessful approach was to use round-hole screens whichseparate on the basis of seed width and do not cause damageto the seed.In a mixture of swollen seeds and hard seeds there isconsiderable overlapping in width between widehard seedsand narrow soft seeds after they have swollen.This overlapping can be avoided by dividing the seed lotinto several width groups before scarification treatment.Soft seeds will increase in width while hard seeds will not,creating a definite width difference between soft and hardTable 3. Percent separation of soft seeds obtained with roundholescreens from each initial width class in crownvetch lots.Seed width Diameter of screen in mmCultivar Lot class 1.81 1.49 1.41. 1.34 1.27 1.21(mm) --------------%-------------Chemung B Over 1.49 100Chemung B 1.49-1.41 92 100Chemung B 1.41-1.34 64 100 100Chemung B 1.34-1.27 100 100 100Chemung B 1.27-1.21 90 100 98 100Chemung B Thru 1.21 60 78 85 100Emerald A Over 1.49 90Emerald A 1.49-1.41 72 100Emerald A 1.41-1.34 57 98 100Emerald A 1.34-1.27 97 100 100Emerald A 1.27-1.21 82 99 100 100Emerald A Thru 1.21 72 80 86 100Penn gift B Over 1.49 90Penngift B 1.49-1.41 70 100Penn gift B 1.41-1.34 58 88 100Penngift B 1.34-1.27 46 100 100Penn gift B 1.27-1.21 82 93 100Penngift B Thru 1.21 56 72 96screen for seeds that were 1.49-1.41 width before soaking,1. 41 screen for seeds that were 1.41-1. 34 width beforesoaking, 1.34 screen for seeds that were 1.34-1.27 widthbefore soaking, and 1.27 screen for seeds that were 1.27-1. 21 width before soaking.A good but not perfect separation was obtained from thewidest groups of Emerald Lot A and Penngift Lot B. Thesewide groups included seeds ranging from 1.81 to 1.49 mm(1/14 to 1/17 inch) in diameter. These extreme widthclasses represented a very small proportion of the seed lots.Viability of Naturally Soft and Hard SeedsThe germination percentage of the "Naturally Soft Seeds"(seeds that po not require scarification to become permeableto water) and scarified hard seeds of six seed lots ispresented in Table 4. Hard seeds averaged 92% germinationafter boiling water scarification with a range of 85% to100%. Naturally soft seeds averaged 68% germination witha range of 46% to 88%.Relative Storability of Soft SeedsAlthough hardseededness can be broken effectively andsoft seeds can be mechanically separated from the hardseeds, the practicability of the technique is also dependenton the storability of the treated seeds. When stored in sealed

6Table 4. Percent germination of "naturally soft seeds" and hardseeds in crownvetch cultivars and lots.Naturally soft HardCultivarLot seedsseeds------------------ % ------------------ChemungChemungEmeraldEmeraldPenngiftPenn giftABABAB6578785388461009293859788Average68 92glass containers, both "Naturally Soft Seeds" and "BoilingWater Soft Seeds" (hard seeds made permeable by boilingwater), showed a slight decline in germination after 10months of storage at 30 C (Table 5). Neither seed categoryTable 5. Percent germination of air-dried soft seeds of crownvetchlots stored at 30 and 10 C. Average of Chemung Lot B,Emerald Lot A, and Penngift Lot A.Storage Months of storageSeed catego!}' temJ.>erature 0 1 4 10oc -------------- % --------------Untreated 30 32 32 32 38Naturally Soft 30 92 89 87 85Untreated 10 32 30 32 43Naturally Soft 10 84 85 82 86Boiling Water Soft 10 92 90 91 91showed a decline in germination after 10 months at 10 C.The increased germination of the untreated seeds after 10months is attributed to some hard seeds becoming permeableduring this period.Effect of Selected Seed Treatmentson Emergence From SoilThe average daily emergence-survival of selected seedclasses from three seed lots when planted in the greenhouseis presented in Figure 4.Emergence-survival of every seed category was lowerthan germination in the laboratory, but in general, seedcategories that showed higher germination in the laboratorystudies also showed higher emergence-survival in the greenhouse.The "Boiling Water Soft Seeds and "Naturally SoftSeeds'' showed faster emergence and about 40 and 30%higher final emergence-survival than the untreated seeds.The lower emergence of "Naturally Soft Seeds" can beattributed to the higher dead seed content of this seedcategory.On the average, the long-narrow seeds (which showedhighest germination in the length-width studies) emerged aswell as the seeds scarified in boiling water for 5 seconds andbetter than the short-wide seeds (which showed the lowestgermination in the length-width studies). Untreated seeds,0>60> 50 '-:::»II)I40CIJuc::CIJ300'-CIJE 20CIJ-r::::CIJ10u'-Q)a.. 06 7 8 9 10 IIDays after plantingTreatments- Boiling water soft seeds2 - Naturally soft seeds3- Scarified 5 seconds4- Long-narrow seeds5 - Untreated6- Short- wide seeds7- Scarified with sand paperFig. 4. Average emergence-survival of selected seed classes ofcrownvetch in the greenhouse. Average of Chemung LotB, Emerald Lot A, and Penngift Lot B.seeds hand scarified between sand paper and short-wideseeds showed the lowest emergence-survival.DISCUSSIONAttempts to separate hard and soft seeds on the basis oflength and width conditioning procedures were only partiallysuccessful. In some lots, hardseededness tended to beconcentrated in certain sizes or shapes, but in other lots, theproportion of hardseededness did not vary much betweensizes or shapes. In those lots where hardseededness isassociated with shape, a sublot with reduced hardseedednesscould be created by making the proper length andwidth separation. The remainder of the seed lot, of course,would have an increased percentage of hardseededness. Thesublot with the lowest hard seed content would be the mostsuitable for planting, while the more hardseeded sublotcould remain in storage or be scarified to reduce hardseededness.In commercial seed conditioning, artificial size differencesare sometimes created when they do not exist naturally.An example of this is the addition of sawdust to moistenedbuckhorn plantain (Plantago lanceolata L.) seeds to makethem larger so they can be separated from alfalfa (Medicago12

7sativa) seeds. Similarly, soft seeds of crownvetch may bemade wider than hard seeds by soaking them in water. Withthis width difference, hard and soft seeds may be easilyseparated with round-hole screens.It is possible to scarify the hard seeds of a crownvetchseed lot without killing the soft seeds, by combining boilingwater scarification and separation of soft seeds. A suggestedprocedure for doing this is as follows:1. Before scarification, subdivide the seed lot intoseveral width groups (sublots) with a series ofround-hole screens.2. Submerge each sublot in boiling water for 5 seconds.3. After removal from boiling water, soak the seeds for16 hours in water at room temperature to allow thesoft seeds to swell. Remove from the water andsurface-dry the seeds.4. Pass each sublot over the same screen that was usedto make the original width groups. For example, thesublot that passed through a 1.49 mm screen but washeld by a 1.41 mm screen, should be passed over a1.49 mm screen. Hard seeds will fall through as theydid before soaking, but the swollen soft seeds will beheld.5. The hard seeds remaining may be put through asecond cycle of boiling water scarification, soakingat room temperature and screening.6. A third cycle may be necessary for lots with anextreme degree of hardseededness. One cycle maybe sufficient for other lots.7. Swollen seeds should be dried to their originalmoisture content for safe storage and to permit theuse of normal seeding equipment.As a variation of this procedure, sublots with no hardseeds could also be created without boiling water scarificationmerely by soaking in water and screening off theswollen seeds. The procedure could also be adapted toeliminate fast swelling seeds which are lowest in viability.(Peiffer, McKee and Ditmer, 1972)Since hard seeds are higher in viability than soft seeds,and constitute a high proportion of the seedlot, there isconsiderable value in scarifying crownvetch seed beforeplanting. The practicability of the procedure described willbe dependent on the size of the seed lot and its intended use,equipment available, need, economics and the extent ofhardseededness present. In seed testing, for example,boiling water scarification provides an efficient way ofovercoming hardseededness so that viability of hard seedsmay be determined in a germination test.In research, boiling water scarification could be verypractical for small lots of seed for experimental plantings.The procedure could likewise be practical for small- tomedium-sized lots used for erosion control and ornamentalplantings. The feasibility of using these techniques for verylarge seed lots may be questionable, however.Additional research may show that the several-cycleapproach to boiling water scarification could be applied toother species of legumes as well. Previous workers havefrequently concluded that boiling water scarification isimpractical because the long treatments required to overcomehardseededness frequently reduced the viability of thesoft seeds in the lot. This disadvantage is overcome with therepeated-cycle approach.REFERENCES1. Association of Official Seed Analysts. 1978. Rules fortesting seeds. J. Seed Techno!. 3:1-126.2. Brant, R.E. 1968. The effect of chemical and physicaltreatments on impermeable seed of Coronilla varia L.Master's Thesis. The Pennsylvania State University. 85numb. leaves.3. Casey, J.E. 1939. Boiling water treatment an aid in germinatingbur clover (Medicago arabica and M. maculata)seed. Proc. Assoc. Off. Seed Anal. 31:114-115.4. Gray, S.G. 1962. Hot water seed treatment for Leucaenaglauca (L.) Beenth. Austral. J. Exp. Agr. Animal Husbandry2:178-180.5. Middleton, O.K. 1933. Size of Korean lespedeza seed inrelation to germination and hard seed. J. Amer. Soc. Agron.25:173-177 06. Peiffer, R.A., G.W. McKee, and W.P. Ditmer. 1972.Detection of processing injury and viability in crownvetchseed by staining with fast green. Proc. Assoc. Off. SeedAnal. 62:109-115.7. ---. 1972. The quick swell test as an indication ofpotential germination in crownvetch seed. Proc. Assoc. Off.Seed Anal. 62: 101-108.8. Vaughan, C.E. 1962. Physical and physiological propertiesof seeds associated with viability in small seeded legumes.Master's Thesis. The Mississippi State College. 67 numb.leaves.9. Wilson, J .K. 1937. Scarification and germination of blacklocust seed. J. Forestry 35:241-246.10 .---. 1944. Immersing seeds of species of Robinia inboiling water hastens germination. J. Forestry 42:453-454.

A Proposed Severity Index for Seed Handling 1N. Robert Brandenburg 2ABSTRACTSeeds are exposed to various mechanical actions or forces asthey pass through handling operations like threshing, separating,conveying, and scarifying. Some actions are relatively severeand cause seed damage. Other actions are relatively gentle andmay improve germination of seed species having hard seeds. Asa means of minimizing seed damage and improving qualitycontrol in seed production, a technique was developed toevaluate severity of handling operations. A formula was devisedto relate the severity of mechanical actions to the germinationresponse of seeds handled and permit the calculation of severityindex numbers. A comparison of these numbers for a seed lotbefore and after exposure to handling operations can providethe following information: (1) The severity of a given system;(2) Seed damage, if any, that has occurred; (3) The potential forseed damage that might occur under less favorable conditions;(4) The identity of specific machines or processes that areactual or potential contributors to seed damage. This informationis useful in reducing severity of those harvesting orconditioning operations that damage or may damage seed.Additional index words: seed harvesting, conditioning, injury,mechanical damage, germination, quality control, hardseed.INTRODUCTIONSeeds are subjected to various mechanical actions orforces like impact, abrasion, shear, and compression as theyare harvested and conditioned (processed). Such actionsmay influence the germination capability of seeds and, insome cases, severely injure them. This is particularly truefor certain small legume seeds like alfalfa (Medicago sativaL.) and clover (Trifolium spp.), for vegetable seeds like beans(Phaseolus vulgaris L.) and peas (Pisum sativum L.), andfor oil seeds like soybeans (Glycine max L. Merrill) and cotton(Gossypirm spp.).Seed injury causes a substantial loss to the U.S. seedindustry. If present damage percentages based on total seedproduction were reduced by one percentage point, thefollowing amounts of additional viable seed would beobtained annually: 453,000 kg of alfalfa; 384,000 kg ofgarden beans; 635,000 kg of peas; and 19,195,000 kg of1Contribution of Agricultural Research Service, USDA, incooperation with the Oregon Agricultural Experiment Station.Technical Paper No. 6662 of the latter. Received 24 January 1983.2Agricultural Engineer, Agricultural Research Service, USDA,Department of Agricultural Engineering, Oregon State University,Corvallis, 97331.soybeans. (Production figures from U.S. Department ofAgriculture, 1980)Seed damage resulting from mechanical injury can takedifferent forms. Damaged seeds may show external cracksand breaks, or they may appear sound but have internalbreakage or other damage. They may germinate and produceabnormal seedlings, or they may be dead. Standardgermination tests determine which seeds in a sample germinatenormally, abnormally, or not at all. Some seeds failto germinate because they are dead, while others arerestricted from germinating within a prescribed period bynatural physiological characteristics of the seeds. A commongermination-restricting mechanism for some seed types is aseed coat that is impermeable to water. Seeds with thisproperty are called hard seeds and are generally assumed tobe alive and capable of germinating after their seed coatsbecome permeable. Hard seeds and normally germinatingseeds are traditionally reported in official germination tests.Upon request, testing laboratories will also report abnormallygerminating seeds and dead seeds.Many investigators have studied seed response to themechanical actions involved in such handling operations asthreshing, separating, conveying, and scarifying. In threshingof bean seeds, there was general agreement that mechanicalinjury resulting in abnormal seedlings was influenced criticallyby cylinder speed and seed moisture content (Bainerand Borthwick, 1934; Asgrow, 1949; Toole and Toole,1960; Dorrell and Adams, 1969). Similar relationships ofcylinder speed and seed injury were found for threshingalfalfa and red clover (Bunnelle, Jones and Goss, 1954), forthreshing soybeans (Park and Webb, 1958; Moore, 1972),and for threshing crimson clover, (Klein and Harmond,1966). Seed damage with resultant drop in normal germinationwas also found related to feed rates for threshing redclover (Bunnelle et al. , 1954) and for beans (Asgrow,1949). In threshing of crimson clover, Park and Webb(1958) noted that hard seed percentages were decreased byaggressive threshing. Franca and Potts (1978) also reporteda lowering of hard seed content in threshing of soybeans.Seed response to mechanical actions associated with theseparating or cleaning process has been studied by simulatingthese actions through various drop or impact techniques.Asgrow (1949) and Atkin (1958) found that thenormal germination percentages of bean seeds were reducedsharply by dropping the seeds repeatedly onto a hard surfacefrom heights of 0.61 m and 0.71 m, respectively. Burris( 1980) showed that the amount of seed damage caused bydropping soybeans was related to seed moisture content.Similar moisture-damage relationships were found for beanseed by Asgrow (1949) and Perry and Hall (1960).Silbernagel (1973) observed that dropping of bean seeds8

9stimulated seedling growth slightly.The influence of conveying operations on seed germinationalso has been investigated. Pneumatic conveyors werefound to produce seed damage that increased with airvelocities when handling alfalfa (Goss and Jones, 1956) andwhen handling beans (Brandenburg and Harmond, 1964).Asgrow (1949) determined that seed damage in beans wascumulative in repeated passes through an airlift. Andrews(1965) reported similar cumulative damage in repeatedpasses of seeds through elevators.The germination response of seeds to mechanical scarificationhas been the subject of many studies. Scarification iscommonly performed on seed of alfalfa, clover, lespedeza(Lespedeza spp.), crownvetch (Coronilla varia L.) andother hard-seeded legumes to improve germination. Scarifyingis accomplished mechanically by abrading or crackingthe seed coat to increase permeability. It has been welldocumented that mechanical scarification can effectivelyinduce some hard seeds to germinate (Battle, 1948; Parkand Webb, 1958; McKee, Peiffer, and Mohsenin, 1977). Itis also well documented that scarification can be tooaggressive and cause seed injury resulting in excessivelevels of abnormal germination and/or dead seed (Arnold,1945; Jones, 1971; Moore, 1972).In many cases, seed injury from scarification does notappear immediately but shows up later after storage of theseed. Graber (1922), Bret (1952), and others have foundthat mechanically scarified seeds show greater decreases innormal germination during storage than unscarified seeds.Consequently, this type of scarification is best performedshortly before the seeds are planted.A different method of reducing hard seed content usingradiofrequency electric energy was investigated extensivelyby Nelson (1976). At optimum exposures, this techniquesharply increased normal germination percentages in alfalfawhile producing essentially no seed damage. In addition,seed lots treated in this manner showed no deterioration inseed- quality after storage periods of 4 years.Other processes that subject seeds to mechanical actionsthat may influence germination include drying, debearding,hulling, polishing, treating, blending, and storing.Mechanical injury to seeds cannot be avoided totally inharvesting and conditioning. However, it is possible toreduce injury by minimizing the severity of operations thatexpose seeds to potentially damaging forces. The objectiveof this research study was to develop a technique forevaluating severity and potential for damage in the handlingof legume and other seed species having hard seeds.Knowledge of the potential for damage can be used toreduce the severity of a process or specific machine thatotherwise may injure seeds and reduce seed value.DEVELOPMENT AND DISCUSSIONOF SEVERITY INDEXThe effect of a mechanical action on seed species havinghard seeds can be damaging, beneficial, or neither. Arelatively severe action may cause damage, either by killingseeds or causing them to germinate abnormally. A relativelymild action may have no effect or may be beneficial bycausing hard seeds to germinate normally. Since the severityof a mechanical action is important, the followingformula was devised to rate relative severities of mechanicalactions based upon germination response of seeds:Severity Index No. = ----'l'------'--+.-'("-'A'-"'G';-;+~D::..::S:L)_ ------- (1)1 + HSWhere: AG = abnormally germinating seeds, percentageby numberDS = dead seeds, percentage by numberHS = hard seeds, percentage by number[As determined in standard tests for seed germination(International Seed Testing Association,1976)]A term for normal germination of seeds is not included inthe formula because any change in this measure caused by amechanical action will be reflected in revised values for theabnormal, dead, or hard seed terms. The arbitrary figure of1 appears in both numerator and denominator to providefinite numbers and a mathematically rational conditionwhen the hard seed content or abnormal germination plusdead seed percentage is zero.When applied to seed lots before and after handling, theformula can sense and measure actions that damage seeds(change the numerator) or cause hard seeds to germinate(change the denominator). The initial index number for aseed lot will change when the action produces any of thefollowing events: ( 1) Initially hard seeds germinate normally,germinate abnormally, or are killed; (2) Initially normallygerminating seeds germinate abnormally or are killed.These events are shown schematically in Figure 1 alongwith another event where initially abnormally germinatingseeds are killed (no change in index number). Some seedsof each germination category shown in Figure 1 may bepresent originally in a given lot of seed. Exposure to amechanical action can shift seeds from one category toanother in the directions indicated.Considering all of the described possible germinationresponses, the numerator of the formula either increases orstays the same, and the denominator either decreases orstays the same. As a result, the calculated index numbermust either increase or stay the same for seeds subjected toany mechanical action.As severities of actions increase, the first change (increase)in index number usually results from a decrease inhard seed content. This seed response is the basis for thepractice of scarification. Hard seeds are encouraged togerminate because the mechanical action tends to increasepermeability of the seed coat. Greater levels of severity tendto eliminate more hard seeds but also may cause seeddamage by increasing abnormal germination and/or deadseed percentages. These germination responses, which maybe seen in the flow diagram of Figure 1, will result in higherindex numbers. For purposes of this discussion, damagemay be considered an extreme form of severity.Some severity is desirable in harvesting and conditioningseed lots that contain hard seeds. Ideally, the mechanicalactions involved should be severe enough to reduce hardseeds, but not severe enough to damage seeds. This level ofseverity will cause an increase in normal germination. As

10HARDGERMINATE ..,....NORMALLY""- GERMINATE-ABNORMALLY...~-...DEAD-..,Fig. 1. Possible changes in germination categories of seedssubjected to a mechanical action.reflected in the formula, there will be a decrease indenominator but no increase in numerator, resulting in anincreased severity index number but no damage.A different type of seed index that relates damagevulnerability to seed characteristics was reported by Potts(1972). In this case, a number denoting seed structure wasmultiplied by a number denoting chemical composition toobtain an index value called the damage susceptibility index(DSI). Considering seed properties only, DSI values canindicate the need for greater care in handling certaindamage-sensitive seed types. In contrast to the DSI, theseverity index developed here is concerned with a givenseed lot and its germination response to the mechanicalactions of a given seed handling operation.An equivalent form of formula (1) can be stated to permitcomputation of severity index numbers using the termsnormal germination and hard seed that are traditionallyreported by seed testing laboratories. Since the four categoriesof germination as indicated in Figure 1 must total 100percent in standard tests,100 = NG + HS + AG + DS(AG + DS) = 100 - (NG + HS)- - - - - - - - - -(2)Substituting for (AG + DS) in (1):Severity Index No.1 + 100 - (NG + HS)1 + HS101 - (NG + HS) - - - - - -(3)1 + HSWhere: NG = normally germinating seeds, percentageby numberHS = hard seeds, percentage by numberAG = abnormally germinating seeds, percentageby numberDS = dead seeds, percentage by number(As determined in standard tests for seed germination)The severity index number is useful in determining thedegree of severity in a harvesting or conditioning operation.A relatively large increase in index number due to ahandling operation indicates a substantial degree of severityin the operation. And, the higher the level of severity, thegreater the chance for seed damage. In a practical application,the severity index formula was used to rate severityand potential for damage in a pneumatic conveying systemtransporting crimson clover. The system was operated atdifferent levels of aggressiveness produced by varying airvelocities, conveying rates, and number of passes. Resultsare presented in Figure 2. The severity index curve clearly1001- 60zw(.)500::wa. 403020NORMAL GERMINATION(%)HARD SEEDS (%)ABNORMALGERMINATION PLUS10 "_j_ l DEAD SEEDS (%)QL_ __·-----==--·~·L_ l_ l_ ~-~~-~NONE LOW MED.- MED. MED.-(CONTROLS) LOW HIGHAGGRESSIVENESS OF CONVEYINGFig. 2. Severity index numbers and seed germinations inpneumatic conveying of crimson clover (T. incarnatum.)Plotted points are averages of two or more determinations.shows that the potential for damage increased as aggressivenessincreased, even though no change was apparent inactual damage (abnormal germination plus dead seeds) asmeasured by standard germination tests.0::w00::::!!:JzXwc~>­!:::::0::w>w(/)

11Studies referred to earlier showed that damage varieswith seed moisture, rate of handling, number of passesthrough a machine, and other factors. Since the point atwhich damage first occurs can change with these variables,it is a sensible precaution to closely monitor and perhapsreduce severity of a handling operation that causes a largeincrease in index number.Handling severities can be reduced either by changingoperating techniques in an existing system or by changingthe system. The potential for seed damage can be lessenedby conditioning seeds at their optimum moisture content forminimum damage, lowering the flow rate, and reducingnumber of passes. System changes that tend to decreasedamage potential include eliminating some conditioningmachines, reducing height or impact of seed drops, decreasingdischarge velocities of bucket elevators, and decreasingair velocities in pneumatic conveyors.The severity index number also can be used to compareseed handling operations according to degree of severity.On this basis, relative severity ratings for handling operationscan be determined either for a total system or forcomponents of a system. For example, consider two alternativeconditioning systems that show no seed damage forthe same seed lot as determined by usual measures. Althoughboth systems appear satisfactory, severity indexratings may reveal that one system is less severe than theother. In this case, the system showing lower severitywould be preferable since it is less likely to produce seeddamage under more adverse operating conditions.Similarly, components of a particular conditioning systemcan be numerically rated and compared individually toidentify specific machines or processes that are potentialcontributors to seed damage. Once identified, the machineor process may then be modified to reduce its severity.REFERENCES1. Andrews, C. H. 1965. Mechanical injury in seeds. ShortCourse for Seedsmen Proceedings. Mississippi State Univ.,State College, Mississippi, May 3-6. p. 125-130.2. Arnold, H. A. 1945. Seed scarifiers. University of Tennessee,Agricultural Experiment Station Bulletin No. 194, Knoxville,Tennessee.3. Asgrow. 1949. A study of mechanical injury to seed beans.Asgrow Monograph No. 1. Associated Seed Growers, Incorporated,New Haven, Connecticut.4. Atkin, J. D. 1958. Relative susceptibility of snap beanvarieties to mechanical injury of seed. Amer. Soc. forHorticultural Sci. Proceedings 72:370-373.5. Bainer, R., and H. A. Borthwick. 1934. Thresher and othermechanical injury to seed beans of the Lima type. CaliforniaBul. No. 580.6. Battle. W. R. 1948. Effect of scarification on longevity ofalfalfa seed. J. Amer. Soc. Agron. 40:758-759.7. Brandenburg, N. R., and H. E. Harmond. 1964. Fluidizedconveying of seed. U.S. Dept. of Agric. Tech. Bul. No.1315.8. Brett, C. C. 1952. Factors affecting the viability of grass andlegume seed in storage and during shipment. InternationalGrasslands Conference 6:878-884.9. Bunnelle, P. R., L. G. Jones, and J. R. Goss. 1954.Combine harvesting of small-seed legumes. AgriculturalEngineering 35:554-558.10. Burris, J. S. 1980. Free fall and conditioning environment­Is seed quality affected? Seedsmens Digest 31:44-46.11. Dorrell, D. G. , and M. W. Adams. 1969. Effect of someseed characteristics on mechanically induced seed coat damagein navy beans. J. Amer. Soc. Agron. 61:672-673.12. Franca, J. B., and H. C. Potts. 1978. Response of hardseeded soybeans to combine harvest and artificial drying.Agronomy Abstr., p. 109.13. Goss, J. R., and L. G. Jones. 1956. Problems in pneumaticlifts. Seed Processing Conference Proceedings, University ofCalifornia, Davis, February 9-10. P. 14-22.14. Graber, L. F. 1922. Scarification as it affects longevity ofalfalfa seed. J. Amer. Soc. Agron. 14:298-302.15. International Seed Testing Association. 1976. Internationalrules for seed testing. Seed. Sci. and Tech. 4:3-49.16. Jones, M. E. 1971. Seed scarification. Rhodesia AgriculturalJournal 68:25 and 31.17. Klein, L. M., and J. E. Harmond. 1966. Effect of varyingcylinder speed and clearance on threshing cylinders incombining crimson clover. Trans. American Soc. of AgriculturalEngineers 9:499-500 and 506.18. McKee, G. W., R. A. Peiffer, and N. N. Mohsenin. 1977.Seed coat structure in Coronilla varia L. and its relation tohard seed. J. Amer. Soc. Agron. 69:53-58.19. Moore, R. P. 1972. Effects of mechanical injuries onviability, P. 94-113. In: E. H. Roberts (Ed.). Viability ofSeeds, Syracuse University Press, New York.20. Nelson, S. 0. 1976. Use of microwave and lower frequencyrf energy for improving alfalfa seed germination. J. ofMicrowave Power 11:271-277.21. Park, J. K., and B. K. Webb. 1958. Seed harvesting in theSoutheast. South Carolina Agric. Expt. Sta. Bul. 461,Clemson, South Carolina.22. Perry, J. S., and C. W. Hall. 1960. Germination of peabeans as affected by moisture and temperature at impactloading. Michigan Agric. Expt. Sta. Quarterly Bul. 43:33-39.23. Potts, H. C. 1972. A closer look at seeds. Short Course forSeedsmen Proceedings, Mississippi State University, StateCollege, May 1-3, 1972. P. 113-123.24. Silbernagel, J. J. 1973. Harvesting high-quality bean seedwith a rubber-belt thresher. Washington Agric. Expt. Sta.Bul. No. 777.25. Toole, E. H., and V. K. Toole. 1960. Viability of snap beanseed as affected by threshing and processing injury. U. S.Department of Agriculture Tech. Bul. 1213.26. U. S. Department of Agriculture. 1980. Agricultural Statistics,U. S. Government Printing Office, Washington, D. C.

The Effect of Cutting in Spring and Application of Alaron Red Clover (Trifolium pratense L.) Seed Production 1K. P. Pori and A. S. Laidlaw 2ABSTRACTThe combined effect of cutting in spring and level and timingof Alar application on plant morphology, bee activity, andcomponents of seed yield in the first harvest year (1981) wasinvestigated in a red clover field trial. Small plots of cv.Sabtoron were left uncut in spring (C 0 ), cut on 15(C 15 ) or 30 (C 30 ) May and Alar at 0 (L 0 ), 4 (L 4 ) or 8 (L 8 ) kga.i. ha ·1 was applied on 10 (T 10 ) or 25 (T 25 ) June. Inflorescenceappearance rate and components of yield were assessedon plants within a fixed quadrat of 0.25 m 2 and plots wereharvested on 15 September.Alar reduced stem length and corolla length but had abeneficial effect on seed setting only when applied early in theflowering season.Cutting in spring resulted in a delay in inflorescence appearanceproducing a distinct period of rapid appearance. Alarincreased appearance rate especially within the period of rapidappearance. Bee activity was closely related to inflorescenceappearance rate. C 15 produced heavier seeds than C 0 resultingin a significant increase of 11% in seed production (597compared to 538 kg ha ·1 ). C 30 produced 12% less seed thanC 15 due to a significant reduction in inflorescence number (868compared to 975 m ·2 ).L 4 produced 582 kg ha ·1 of seed compared to 474 kg ha ·1by L 0 , the increase being due to a significant increase ininflorescence number (29%) and a slight increase in number ofseeds per inflorescence. Overall, yields from L 8 (604 kg ha ·1 )were not significantly different to those from L 4 • Early applicationof Alar was more effective than late application in C 0and less effective in C 30 •Mean potential seed yield calculated from components ofyield was 2. 7 times that of actual seed yield, Alar having aneffect on potential yield similar to that on actual yield, althoughL 8 produced higher potential yields than L 4 at the earlyapplication and T 25 produced a lower potential yield thanT 10 •It is concluded that cutting has a small effect on seed yieldcompared to the effect of Alar and that Alar is less effectivewhen applied to 10-day regrowth than when applied to 25- or40-day regrowth.Additional index words: Red clover, seed production, cutting(time of), Alar level/timing, seed yield components.LA contribution of the Queen's University of Belfast, NorthernIreland. Received 14 March 1983.2· Present address: Department of Animal Science, PunjabAgricultural University, Ludhiana, India and Department of AgriculturalBotany, Queen's University of Belfast, Newforge Lane,Belfast BT9 5PX, Northern Ireland, respectively.INTRODUCTIONCutting primary growth of red clover grown for seedproduction is the recommended practice to delay inflorescenceappearance so that flowering coincides with the timewhen pollinators are most likely to be prevalent (Williams,1925). However, the treatment may not necessarily result inincreased yields of seed. Briggs (1955), Hawkins (1958),Stoddart (1961) and Rincker et al. (1977) found that cuttingin spring adversely affected seed production. The effect ofspring cutting could be made more predictable if theinfluence of spring cutting on the development of thecomponents, which contribute to seed yield in red cloverfrom cutting to harvest, were known.The application of the growth retardant Alar has beenshown to increase seed yield in red clover (Picard andSigwalt, 1967; Hulewicz and Hortynski, 1970; Fojtik et al.,1976 and Wioncek et al., 1977). Mela (1969) in Finlandachieved increases of up to 64% by application of 10 kgha ·1due to higher inflorescence number, delay in inflorescenceappearance, increase in floret number per infloresenceand reduced lodging. Holm (1972) and Wawryn(1978) have shown corolla tube length to be shortened andbee activity to be increased by Alar. However, a systematicapproach to determine the optimum level of Alar applicationon red clover has not been attempted.Cutting time in spring and level and timing of growthretardant application are management factors which can bevaried. This study was carried out to investigate thecombined effect of these factors and to elucidate thedevelopmental changes which take place between impositionof treatments and harvesting of seed, paying particularattention to flowering.MATERIALS AND METHODSCulturalThe trial was sown on 29 July 1980 by broadcasting at 15kg ha ·1 seed of red clover cv. Sabtoron previously inoculatedwith Rhizobium trifolii Rothamsted strain 5. A basaldressing of25 kg N, 50 kg P 2 0 5 and 75 kg K 2 0 ha·1 wasapplied to the seed bed and a further 20 kg P 2 0 5 and 40 kgK 2 0 ha·1 were broadcast on 20 April 1981. Plots weresprayed with Asulox (asulam) at 1.12 I a.i. ha -I on 17September 1980 to control Rumex obtusifolius, cut on 17October 1980 and sprayed with Clout (alloxydim-sodium)at 1. 88 kg ha -I to control invading grasses on 12 March1981.TreatmentsPlots were subjected to one of three cutting treatments in12

13spring viz. not cut (C 0 ), cut on 15 May (C 15 ) or cut on 30May (C 30 ). Alar was applied at 3 rates viz. 0,(L 0 ), 4 (L 4 ) or 8 (L 8 ) kg a.i. ha - 1 in the equivalent of1000 1 ha - 1 water, with 100 ml Agral/1000 I as a wettingagent, on 10 June (T 10 ) or 25 June (T 25 ) using a pressurizedknapsack sprayer through a fan type nozzle held25-30 em above the herbage. The stages of development ofthe plants at spraying (mean of 10 stems per plot) arepresented in Table 1. The dry and wind-free weatherTable 1. Stage of development of Red Clover plants at time ofAlar application.Time of application10 June (T 10 ) 25 June (T 25 )Cutting Height of Number of Height of Number ofdate plant (em) extended plant (em) extendedinternodes/internodes/stemstemNot cut (C 0 ) 51.5 5.5 58.9 5.915 May (C 15 ) 20.2 1.5 36.2 4.430 May (C 30 ) 11.0 0.0 23.5 2.3conditions subsequent to spraying were suitable for uptakeof Alar. The treatments were in 4 blocks (72 plots) withcutting and time of application as main plots and level ofapplication randomized within these main plots as subplots.Morphological CharacteristicsBetween 22 and 29 July the length of corolla tube, heightof nectar level, diameter of corolla tube and depths ofhiding nectar (length of corolla tube - height of nectarlevel) were measured under a stereo-microscope on fifteenflorets at the distal part of each of 4 inflorescences per plot,the calyx having been carefully removed with forceps. Stemlength, number of nodes and inflorescences per stem wereassessed on 15 stems within a fixed quadrat at 0.5 x 0.5m/plot at harvest time. The number of stems per unit areawas assessed within quadrats. The diameter of internodeswas measured with a micrometer screw guage, to thenearest 0.5 mm., only in stems from plots cut in spring.Seed Yield Components and Potential Seed YieldDuring the initial stages of growth a randomly placedquadrat 0.5 x 0.5m was fixed in each plot. Canes wereinserted at the comer of each quadrat, and the inflorescencesseparated from the adjoining area by cord fixedaround the canes. Within each quadrat inflorescence appearancerate was recorded and all newly opened inflorescences(75 to 100% florets opened) were marked weeklyfrom 5 July to 30 August, being tagged with PVC coatedwire differing in color for each week. The quadrat area washarvested on 5 September, 10 days before the remainder ofthe plot.Seed yield components- number of florets, seed settingpercentage and thousand-seed-weight (M-weight) - wereassessed on 20 inflorescences (or less if 20 were notavailable) which appeared during weeks 3 (13-19 July), 5(27 July - 2 August) and 7 (10-16 August), these observationsbeing analyzed as sub-sub treatments. The productof these components was taken as a measure of seedproduction potential of inflorescences which appearedduring weeks 1-3, 4-6, and 7-9 respectively. Potential seedyield was calculated by integrating the weekly potentialseed production data over the period of assessment ofinflorescences.Bee ActivityBee (Bombus and Apis spp.) activity was recorded in aspecific one m 2 area in each plot from 10 July to 3September by counting bees twice daily at 10.00 and 14.00hours and taking the daily average.LodgingA crop lodging index was calculated from data recordedon 20 August from whole plots using the method describedby Caldicott and Nuttal (1979).Actual Seed YieldHerbage on a net plot area of 1.5 x 1.5 m was harvestedon 15 September with hand clippers, tied in small bundlesand stacked to dry. The harvested material was threshedwith an Alvan Blanch Midget II thresher and the strawdiscarded. The remainder was rubbed between the roughfaces of two rubber mats, passed through a series of sievesand finally cleaned by blowing light debris from the seedsample. The seed was weighed and its moisture contentdetermined.RESULTSStem and Floret CharacteristicsThe main effect of cutting in spring and of applying Alarwas a reduction in stem length and lodging, Alar treatedplants having fewer extended internodes than untreatedplants (Table 2a and 2b). The pattern of response to Alardepended upon the cutting treatment. Alar applied at L 8 ateither date on early cut plots and at L 0 and L 4 on the latecut plots at the late application date gave rise to the shorteststems.Cutting early or not at all resulted in a higher response ofinflorescences per stem to Alar than cutting late (Table 2c).Stems per unit area (mean 160m- 2 ) were not significantlyaffected by any of the treatments.Other than late cutting reducing corolla tube diameter(1.47 mm compared to 1.58 mm (S.E. mean = 0.011) forthe uncut plots), only Alar level affected floral partssignificantly (Table 3). All were reduced by Alar application,the highest level of Alar reducing corolla tube lengthand diameter by 18 and 6% respectively, height of nectarlevel by 7% and depth of hiding nectar by 18%.Components of Seed Yielda) Inflorescence number. Delaying the time of cutting inspring resulted in a reduction in total inflorescence numberin contrast to increasing level of Alar which increased thenumber of inflorescences although the difference between

14Table 2. The effect of cutting, Alar level and time of Alar application on stem morphology of Red Clover.a) Number of extended internodes and lodging indexDate of Nodes/stem Lodging index Alar Nodes/stem Lodging indexcutting(kg ha -1)Not cut (C 0 )7.0600 7.0 6615 May (C 15 ) 6.9544 6.8 4830 May (C 30 ) 6.9498 6.7 49s.e. mean 0.09 NS 2.1 * s.e. mean 0.05* 1.4*b) Length of stem (em)Date of cuttingTiming ofAlar level Not cutapplication(kg ha -1) (Co)10 June (T 10 ) 0 (L 0) 75.34 (L4) 66.58 (L 8 ) 67.2c) Number of inflorescences per stem25 June (T 25 )Not CutAlar level (kg ha - 1 )(Co)0 5.3 (L o)4 6.5 (L 4 )8 7.2(L 8 )Timing of application10 June25 June6.4 (T ) 106.3 (T 25 )0 (L 0)4 (L 4 )8 (L 8)74.871.771.615 May 30 May s.e. mean(C 15) (C 3o)74.9 74.266.8 68.462.5 67.71.24**76.1 75.264.7 62.561.7 59.3Date of cutting15 May 30May s.e. mean(C 15) (C 3o)5.2 5.07.5 6.2 0.28*7.0 5.86.4 5.0 0.24*6.8 6.4Table 3. The effect of rate of Alar application on floretmorphology of Red Clover (mean of cutting and time ofapplication treatments)Alar Corolla Height of Depth of Diameter oflevel tube nectar hiding corolla(kg a.i. ha - 1 ) length level nectar tube- - - - - - - - - - - - - - - - - (mm) - - - - - - - - - - - - - - - - -9.2 1.4 7.8 1.68.0 1.3 6.6 1.57.7 1.3 6.4 1.5s.e. mean 0.07*** 0.01*** 0.06*** 0.01***L 4 and L 8 was not significant (Table 4). The interactionbetween time of application of Alar and cutting in springwas significant due to the uncut treatment producing moreand the late cut treatment producing less inflorescenceswhen Alar was applied on 10 June than on 25 June (Table5).Rate of inflorescence appearance in the three cuttingtreatments was highest during the first three weeks of Julyand lowest during early August in the C 0 plots, the earlyapplication of Alar producing a significantly faster rate ofinflorescence production in C 0 than when applied on 25June (Fig 1). The late cut treatment (C 30 ) had the lowestrate of production during the first three weeks of July butthe highest during early August, the late application of Alaraccentuating this high rate.At both application times and at either rate Alar increasedinflorescence production in C 0 in the third and fourthweeks of July, from late July to mid August in C 15 and forone week in mid August in C 30 .b) Floret and seed number per inflorescence. Significantlymore florets per inflorescence occurred on C 15 andC 30 than on C 0 treatments, cutting having no significanteffect on the number of seeds per inflorescence (Table 4).Consequently, seed setting percentage was significantlyhigher in the uncut than cut treatments. In C 15 treatmentsinflorescences produced early in the season had the highestnumber of florets and seeds whereas, due to the delay incutting, the inflorescences produced in mid season had most(Table 6).Alar inceased the mean number of florets and seeds perinflorescence (Table 6) and increased seed setting early inthe season relative to untreated plots although the time ofapplication had no effect.c) Seed weight. Determined on the basis of markedinflorescence throughout the season cutting resulted inheavier seeds, other treatments having no significant effect(Table 4). So compensation between seed yield componentshad taken place within cutting treatments viz. between seedweight and inflorescence number. Further, the mean M­weight of seed harvested and weighed after cleaning wassignificantly higher in cut than uncut plots and higher in

15Table 4. The effect of cutting, Alar level, time of Alar application and time of inflorescence appearance on seed yield components ofRed Clover. (Mean of weeks 3, 5 and 7 except for inflorescence number}M-weightTotal Number of: Seed Seeds from HarvestedTreatments inflorescence florets/Qroduction inflorescenceDate of Cutting (m -2)Not cut (C ) 01019 89.315 May (C ) 15975 97.630 May (C 30) 868 98.9s.e. mean 31.6* 1.81 **The relative total number of bees on the plots for theseason recorded was 34 and 40% higher for L 4 and L 8 thanfor L 0 and 22 and 16% higher for C 15 and C 30 than C 0 .Bee number response to Alar was higher at C 15 and C 30than C 0 , L 4 having no effect at C 0 • Cutting in the absenceof Alar had little effect (Table 7). The pattern of bee activityover the flowering period followed closely that for infloresseeds/Setting individual seedinflorescenceinflorescences(%) (g) (g)91.3 102.5 1.54 1.3695.7 98.2 1.70 1.4696.9 97.8 1.79 1.441.57 NS 0.71*** 0.034*** 0.017**Alar level (kg ha -I )0 (L ) 0803 90.24 (L ) 41037 97.28 (L ) 81021 98.5s.e. mean 21.2*** 1.05**Time of Alar application10 June (T ) 10955 97.025 June (T 25) 953 93.6s.e. mean 26.0 NS 1.48 NSTime of inflorescence appearance(13/7 to 19/7) 92 99.7(27/7 to 2/8) 204 97.7(10/8 to 16/8) 148 88.5s.e. mean 2.5*** 1.09***88.4 98.2 1.72 1.4796.6 99.6 1.67 1.4098.9 100.6 1.65 1.391.22** 0.99 NS 0.272 NS 0.020*96.0 99.0 1.70 1.4493.2 100.0 1.65 1.391.28 NS 0.58 NS O.Q28 NS 0.018 NS95.2 95.3 1.9499.9 102.2 1.7288.8 100.9 1.381.51 *** 1.06*** 0.022***Table 5. Interaction effect on Red Clover in inflorescencenumber (m ·2 ) for time of application of Alar and a) cuttingand b) level of Alar applicationa)Time ofapplication10 June (T 10)25 June (T 25 )b)Time ofapplication10 June (T 10)25 June (T 25 )Not cut(Co)10929460(L o)772834Date of cutting15 May 30 May(C t5) (C 3o)983 791966 946s.e. mean44.8*Level of Alar (kg a.i. ha-t)4 8 s.e. mean(L4) (L 8 )10151058107696430.0*plots not treated with Alar than in those to which Alar wasapplied. In this instance compensation was between seedweight and inflorescence and floret number.Bee ActivityTable 6. Significant interactions between time of inflorescenceappearance and cutting or Alar level for selected componentsof seed yield in Red Clover.Time of inflorescence aQQearance13-19 27 July- 10-16 s.e.Jul:t 2 August August meanFlorets/inflorescenceNot cut (C 0 ) 92.6 88.9 86.4Date of 15 May (C 15 ) 108.6 99.0 85.2 1.89***cutting: 30 May (C 30 ) 97.8 105.1 93.8Seeds/inflorescenceNot cut (C 0 ) 90.3 93.9 89.6Date of 15 May (C 15 ) 103.9 99.5 83.8 2.91***cutting: 30 May (C 30 ) 91.4 106.1 93.1Seed setting (%)0 (L 0 ) 90.7 102.3 101.7Level of Alar 4 (L 4 ) 96.5 99.6 102.3 1.85*(kg a.i. ha-t ) 8 (L 8 ) 98.6 104.5 98.9cence appearance and a correlation analysis showed beeactivity to be positively correlated with inflorescencenumber and negatively with corolla tube length and depth ofhiding nectar.Seed YieldCutting in mid May (C 15 ), meaned over all treatments,produced 10.9% more seed yield than the uncut treatment(597 kg ha·1 compared to 538 kg ha- 1 ) and 13.9% more

16300a. No! cui Alar application0 10 Juneto. 25 June200 0 none applied(control)100Q)'""' 0Q)"'3:'E(f)w 200(.)zw(.)(f)w0:::0_JLL.~300 b. Cui 15 May100300 c. Cui 30 May2001000~==~~~--~L_ __ L_ __ _L __ _L __ ~ __ __.J

17Table 8. Actual and potential Red Clover seed yields in response to cutting and Alar application levels and dates.a) Actual seed yields (kg ha - 1 )Date ofapplication10 June (T 10)25 June (T 25)MeanMeanb) Calculated seed yields (kg ha - 1 )Date ofapplicationAlarlevel(kg ha - 1 )0 (L 0)4 (L 4)8 (L 8)0 (L 0)4 (L 4)8 (L 8)Not cut(Co)480587644570459488570506Cutting date15 May 30 May(C 15) (C 3o)493 436658 524631 532594497507675615599466555630550s.e. mean and significanceCutting (C) 11. 9*** XL 22.1 NSAlar level (L) 12.7*** c X T 16.8*Time of applicationLX T 9.7 NS(T) 9.7 NSc XL X T 31.2 NSAlarlevel(kg ha - 1 )0 (L ) 010 June (T ) 104 (L ) 48 (L ) 8Mean0 (L ) 025 June (T ) 254 (L ) 48 (L ) 8MeanNot cut(Co)1233153420881618114716131287Mean470590602554477573605552Cutting date15 May 30 May Mean(C 15) (C 3o)1253 1100 11951877 1482 16311941 1706 19121690 1429 15791275185514001259164715431227170514091349 1510 1483 1447s.e. mean and significanceCutting (C) 49.0 NS CxL 92.7NSAlar level (L) 53.5*** CxT 69.3NSTime of applicationLX T 75.7 ***(T) 40.0*c XL X T 131.1 NSstimulation of inflorescence production late in the season byspring cutting may be related to environmental conditions,Roberts and Lewis (1979) having shown a promotive effecton seed yield by cutting in late May.Although inflorescence production was a major determinantof seed yield, other components contributed to orcountered its effect. For example, seed yield produced bythe late cut treatment in plots to which Alar was appliedearly or not al all, was not so adversely affected asinflorescence production, due to an increase in seeds perhead and seed weight relative to the uncut plots. Thepositive effect of Alar on seed production was a consequenceof high inflorescence production and, to a lesserextent, more seeds per head relative to the untreated plots.The mean M-weight of the seed harvested at the normaltime was 0.25 glower than the mean from the sample areasover the three inflorescence appearance times presumablydue to the methodology employed in the trial, the formerbeing harvested 10 days later and so having drier seeds.Cutting in spring resulted in a contraction of the period ofdistribution of inflorescence appearance. This permitted thetiming of harvest for optimum seed yield to be predicted(Puri, 1982) so that the loss of early produced seed byshedding or sprouting and the inclusion of late producedseed due to immaturity is minimized. However, it couldalso increase the risk of very low yields being producedbecause of low pollination if, for any reason, the period ofmaximum rate of inflorescence appearance was coincidentwith low bee activity.In the year of this trial (1981) rainfall and hours ofsunshine were below average and temperatures wereaverage except for the last two weeks in August when thetemperatures were slightly above normal. If the weather hadbeen stormy the benefits of Alar and cutting would likelyhave been accentuated. The data on stem size presentedhere (stem length and thickness) could make a usefulcontribution to the prediction of the effect of heavy rain onthe crop when suitable models have been developed.In the uncut treatment the higher response to earlyrelative to late Alar application was presumably due to the

18stand being too advanced for the late application to beeffective, there being a high proportion of the leaf surfacecomprising old leaves and effective uptake of some othergrowth regulators having been shown to be reduced by oldleaves e.g. succinic acid (Schonherr and Bukorac, 1978).The lack of cover in the late cut treatment at the earlyapplicaton date presumably reduced the interception of Alarand might even have damaged the young regrowth(Wioncek, et al. 1977).There was also evidence that the late application of Alarat the high rate had an inhibitory effect on inflorescencedevelopment which in tum reduced potential seed yield.However, it did not affect actual yield harvested.In conclusion Alar would seem to have a greater effect onseed yield than cutting in spring, the effect of Alar beinggreater when applied to regrowth of 25 to 40 days thanwhen applied 10 days after cutting. So the biologicalresponse of red clover seed yield to Alar would seem to beindependent of cutting in spring when Alar is applied at astage when regrowth is adequate to intercept it.ACKNOWLEDGEMENTSThe authors are indebted to the Association of CommonwealthUniversities for financial support to K. P. Purl during the periodwhen this work was carried out and to Professor C. E. Wright forconstructive criticism.REFERENCES1. Briggs, G. W. G. 1955. Production of high quality seeds offield crops in relation to climatic and other conditions. InHigh quality seed, its production, control and distribution.Report Workshop Wageningen OEEC/EPA. 1954.2. Caldicott, J. J. B. and A.M. Nuttall. 1979. A method for theassessment of lodging in cereal crops. Jour. of the NationalInst. of Agric. Bot. 15:88-91.3. Fojtik, A., V. Svetlik, and H. Haslbachova. 1976. [Effect ofAlar and B9 on the growth characteristics and seed yieldcapacity of tetraploid red clover (Trifolium pratense L.)].Rostlinna vyroba, 22:701-708.4. Hawkins,R. P. 1958. A survey of late-flowering and singlecut red clover seed crops. Jour. of the Nat. Inst. of Agric.Bot. 8:450-461.5. Hawkins, R. P. 1971. Selection for height of nectar in thecorolla tube of English single-cut red clover. Jour. of Agric.Sci. Cambridge. 77:347-350.6. Holm, S. N. 1972. [Seed yields in red clover in relation tothe number of pollinating bees as influenced by a growthregulator]. Arsskrift Kongelige Veterinaer- og Landboh¢jskole.127-141.7. Hulewicz, T. and J. Hortynski. 1970. [The effect of B995(succinic acid 2, 2-dimethyl hydrazide) and DMSO (dimethylsulphoxide)on the seed yield of polyploid red clover].Zeitschrift fiir Acker-und Pflanzenbau. 132:2-15.8. Mela, T. 1969. The effects of N. dimethylamino-succinarnicacid (B-995) on the seed cultivation characteristic of lateflowering red clover. Acta agralia fennica. 115:1-1149. Picard, J. and C. S. Sigwalt. 1967. [Effect of growthretardants on seed yields of red clover. Preliminary results].A cad d 'Agriculture de France. Ex trait du proces verbal de laSeance due Janvier. 141-148.10. Puri, K. P. 1982. Agronomic studies in seed production ofred clover. Ph.D. thesis. The Queen's University of Belfast.11. Rinckler, C. M., J. G. Dean, C. S. Garrison, and R. G.May. 1977. Influence of environment and clipping on theseed-yield potential of three red clover cultivars. Crop Sci.17:58-60.12. Roberts, H. M. and D. A. Lewis. 1979. Seed productionstudies. Red clover. Report Welsh Plant Breeding Station.103-105.13. Schonherr. J. S. and M. J. Bukovac. 1978. Foliar penetrationof succinic acid- 2,2-dirnethylhydrazide: mechanismand rate limiting step. Physiologia Plantarum. 42:243-251.14. Skirde, W. 1964. [Reactions of grass species and varietiesand of clover to growth retarding substances]. Zeitschrift fUrAcker-und Pflanzenbau. 119:263-282.15. Stoddart, J. L. 1961. Management factors influencing flowerand seed production in red clover. Ph.D. thesis, UniversityCollege of Wales, Aberystwyth.16. Wawryn, T. 1978. [The effect of Alar-85 on flower structureand seed yield of tetraploid red clover]. Biuletyn InstytutuHodowli I Aklimatyzacji Roslin. 134:107-113.17. Williams, R. D. 1925. Studies concerning the pollination,fertilisation and breeding of red clover. Bulletin of the WelshPlant Breeding Station. H2, No 4.18. Wioncek, J., B. Kacperek, M. Krzaczek, J. Hortynski, B.Dys, and T. Hulewicz. 1977. The effect of Alar-85 on theseed yield and yield components in Di- and Tetraploid redclover (Trifolium pratense L.). Zeitschrift fUr Acter-undPflanzenbau. 144:113-129.

Effects of Sunlight Reduction and Post-Harvest ResidueAccumulations on Seed Yields of Kentucky Bluegrass 1R. D. Ensign, V. G. Hickey, and M.D. Bernardo 2ABSTRACTKentucky bluegrass seed growers have recognized the necessityof complete removal of post-harvest residue to maintain seedproductivity of this species from year to year. The residuesrestrict subsequent tiller growth and thereby seed yields aresignificantly reduced. Residue remaining on fields shadesplants and thereby may restrict tiller growth and subsequentseed yield.To explore this theory, polyethylene shade screens whichexcluded 30 and 67% of sunlight were placed over August fieldburned Kentucky bluegrass cv. Baron plants for 75 and 130days. Other treatments included mechanical vacuum clippingat 2.5, 7.6 and 15.2 em levels, field burning of residue and noresidue removal.Tiller numbers of artificial shaded plants and tillers from noresidue removed plots were significantly less than from openfieldburned plants. There were more tillers from plants of close(2.5 and 7.6 em) residue removal than no residue removal butthey were not significantly different from open-field burnplants.The sheath lengths of plants from artificial shaded and noresidue removed plots were greater than open-field burn plants.Panicle numbers of plants from the 67% shaded plants werecomparable to the plants of no residue removed plots. Paniclenumbers from these artificial shaded plants were 41% less thanopen-field burned plants.Seed yields of 67% shaded plants were 24% less than plantssubjected to open-field burning.It was concluded that light reduction due either to shadingfrom post-harvest residue accumulations or by artificial meansaffected Kentucky bluegrass growth and development.Additional index words: Artificial shading, tillering, panicles,field burning, Poa Pratensis, air quality.INTRODUCTIONIdaho, Oregon and Washington produce approximately95% of the Kentucky bluegrass (Poa pratensis L.) seed inthe United States (USDA, 1979). Seed production of thisspecies is enhanced by complete removal of post-harvestresidue. This is best accomplished by open-field, early fallburning.Field burning is the most rapid and economical means toremove post-harvest residueResearch in Idaho (Ensign, et al., 1978) has shown thatmechanical vacuum clipping may maintain subsequent yearseed yields, during the first or second seed year, with some!.Approved for publication by the director of the Idaho Agr.Exp. Stn. as Exp. Stn. J. Article No. 83719. Received 11November 1982.2-Agronomist-Professor, former scientific aide and former graduateassistant, respectively, Dept. of Plant, Soil and EntomologicalSciences, Univ. of Idaho, Moscow, ID 83843.cultivars. Where mechanical removal was practiced forseveral consecutive years, seed yields were reduced (Canodeand Law, 1977).Tiller elongation and growth has been shown to beaffected by residue accumulation. Tillers under shade areelongated, poorly developed, erect, have narrow leaves,and lacking dark green color when grown in shade or heavyresidue accumulations (Cordukes and Fisher, 1974), and(Canode and Law, 1979). Ensign, et al., (1974) alsoreported that early complete residue removal by open-fieldburning will produce larger healthy tillers than whereresidue is not removed. It has been assumed that burningwill remove harvest residue from around the crown of theplants, allowing light penetration into the canopy andthereby promotes vigorous tiller growth and improved seedyields.Idaho research was designed to compare the effects ofexcluding various amounts of sunlight and comparing thesetreatments to residue accumulations as they may alter plantgrowth and seed productivity. The effects of tiller number,sheath length, panicle number, and seed yield of 'Baron'Kentucky bluegrass were studied.MATERIALS AND METHODSOn a fifth year of seed crop residue, shading and residueremoval treatments were established in a field of cultivarBaron Kentucky bluegrass within 3 x 6 m plots using a 7 x 9randomized complete block design. The grass had beenplanted in 1974 in 30 em rows on a Thatuna silt loam,fine-silty, mixed, mesic Ultic Argixerols near Tensed,Idaho.Post-harvest residue was mechanically removed at heightsof 2.5, 7.5, and 15.2 em on 2 August 1979 with a rotarypower mower at a vacuum attachment. A small propaneburner was used to ignite plots which were burned on 14August. One set of plots had no residue removal. All plotswere fertilized with 168 kg/ha of N in the ammonium nitrateform on 29 August and immediately sprinkler irrigated with10.6 em of water.Black polyethylene shade screens were superimposedover burned plots on 17 September 1979. Light intensitieswere reduced by 30% for 75 days, 30% for 130 days, 67%for 75 days, and 67% for 130 days. The shade screens weresuspended 0.5 m above the ground on 2.4 x 3 m woodenframes. During October an additional10.6 cm/ha of irrigationwater was applied.Soil moisture readings were obtained by removing soilcores 2.5 em in diameter and 15 em in depth. The sampleswere weighed, air dried, and reweighed to determine thepercentage of soil moisture. On 2 March 1980, five grasscore samples, 9.5 em in diameter and approximately 15 emin depth, were removed from each plot. They were dissectedand the number of large tillers per core counted. Sheath19

20Table 1. Effect of residue removal and shading on soil moisture and Kentucky bluegrass yield components. 1Treatment Soil Moisture Tillers Sheath Length Panicles Seed Yeilds- - - % - - - - - no. core- 1 - - - - - - em - - - - - - no. m- 2 - - - - kg ha- 1 - -Open-field burned 12.6 c 20.9 a 2.7 d 5308 a 1266 abc30% shade-75 days 14.2 be 12.4 bcde 3.2 cd 4683 ab 1399 ab30% shade-130 days 14.2 be 14.8 bed 3.3 cd 3898 be 1194 abc67% shade-75 days 18.3 a 9.8 de 3.6 c 3349 de 1339 abc67% shade-130 days 18.3 a 7.8 e 3.5 c 3145 de 966 dClip to 2.5 em 13.0 be 17.5 ab 2.7 d 5241 a 1542 aClip to 7.6 em 12.9 be 17.0 abc 3.0 de 4762 ab 1304 abcClip to 15.2 em 15.4 b 11.3 cde 5.6 b 3246 de 983 dNo residue removal 15.4 b 10.5 de 6.3 a 2447 d 1079 d1 Means within each column followed by the same leter are not significantly different at the 5% level of probability based on Duncan'sMultiple Range Test.length, measured from the crown to collar of first leaf wasrecorded in early April 1979. Panicle counts were recordedin the field on 19 June 1980. Seed yields were obtained byremoving a 0.6 x 6.0 m swath of panicles from each plot on10 July. These samples were then thrashed, cleaned, andweighed.Data were analyzed as a randomized complete blockdesign. Duncan's multiple range test was performed toseparate treatment means.RESULTS AND DISCUSSIONSoil MoistureSince moisture was applied from August - October, thesoil moisture level was adequate for excellent autumn tillergrowth. The soil moisture was also significantly increasedby 67% shading in the fall. The effect of artificial shading atthe 30% level on soil moisture was comparable to the effectof clipping or not removing the residue (Table 1).TillersTillering was suppressed most by 67% light exclusion.Shading at the 30% level restricted tiller development equalto clipping to 15.2 em and the no residue removal treatment.Open-field burned plants gave the largest number of tillersper 9. 5 em diameter core (Ensign et al. , 197 4). Suppressionof tillering as the result of reduced light agrees with Ryle(1961) and Sprietz and Ellen (1972).Sheath LengthSheath lengths were longest where no residue was removed(Table 1). The open-field burned treatments and theshort clipping to 2.5 em produced short sheaths which wereless than shading at the 67% light exclusion. Thus shadingeither with artificial screens or by excessive residue accumulationincreased sheaf length. It is assumed thatshading increased leaf and sheath elongation due to lightexclusion as reported by Wilkinson, et al., (1975). He alsoreported a reduction in net photosynthesis by shading whichwould be expected to affect potential seed production.PaniclesOpen-field burning and clipping to 2.5 em produced thelargest numbers of panicles. No-residue removal and 75%exclusion of the sunlight produced relatively less panicles.These results correspond with Ryle (1961) who reportedthat panicle numbers were reduced 15% by shading. Thesedata indicate that residue accumulation may reduce lightpenetration into the canopy and cause a reduction in paniclenumbers.Seed YieldsSeed yields of the 67% shaded, the clipped to 15.2 emlevel, and the no residue removal plants were significantlyless than open field burn plants. Seed yields of 67% shadedplants were less than 24% of the open-field burned plants.Seed yields of the low residue clip (2.5 em) were notdifferent than the open-field burned plants. Thus, completemechanical residue removal offers a possible alternative tofield burning although such mechanical removal may not beeconomically feasible.Seed yields seem to be reflected in the increase numbersof panicles as were observed in open-field burning and closemechanical clipping of residue.REFERENCES1. Canode, C. L. and A. G. Law. 1977. Post-harvest residuemanagement in Kentucky bluegrass seed production.Washington State Univ. Coil. Agric. Res. Ctr. Bull 850.2. , and . 1979. Thatch and tiller size asinfluenced by residue management in Kentucky bluegrassseed production. Agron. J. 71:289-291.3. Cordukes, W. E. and J. E. Fisher. 1974. Effects of shadingof the leaf sheath on the growth and development of the tillerstems of Kentucky bluegrass. Can. J. Plant Sci. 54:47-53.4. , B. Augustin, M. Buettner, P. Gray, R. Hall, andR. Nelson. 1974. Burning and alternative treatments forKentucky bluegrass seed production. Idaho Agr\c. Res.Prog. Rpt. No. 188.5. Ensign, R. D., V. G. Hickey, and M. D. Bernardo. 1978.Seed yields of Kentucky bluegrass as influenced by methodsof post-harvest residue removal. Idaho Agric. Res. Prog.Rpt. 204, p. 10-14.6. Ryle, G. J. A. 1961. Effects of light intensity on reproductionin S. 48 timothy (Phleum pratense L.). Nature 191:1%-197.7. Spiertz, J. H. J. and J. Elen. 1972. The effect of lightintensity on some morphological and physiological aspects ofthe crop perennial rye grass (Lolium perenne L. var.'Cropper') and its effect on seed production. Neth. J. Agric.Sci. 20:232-246.8. Wilkinson, J. F., J. B. Beard, and J. V. Krans. 1975.Photosynthetic-respiratory responses of 'Merion' Kentuckybluegrass and 'Pennlawn' red fescue at reduced light intensities.. Crop Sci. 15:165-168.9. USDA. 1979. Statistical reporting service. Boise, ID.

The Effects of the Environment at Anthesis on the Seed Yield and Yield Componentsof Perennial Ryegrass (Lolium perenne L.) cv. S.24. 1J. G. Hampton and P. D. Hebblethwaite 2ABSTRACTAn examination of macro-environmental factors duringan thesis of perennial ryegrass (Lolium perenne L.) cv. S.24 seedcrops at Sutton Bonington, UK over the period 1971-80 showedthat minimum screen temperature (C) accounted for 70% of thevariance in seed numbers recorded. Wind velocity in the weekfollowing peak anthesis and rainfall during the week of peakanthesis were significantly related to seed number and yieldrespectively. Further work should determine the effects of themicro-environment at anthesis on seed number and yield.Additional index words: Temperature, wind, velocity, rainfall.INTRODUCTIONHampton and Hebblethwaite (1983) have shown that inperennial ryegrass seed crops which lodge either before orduring anthesis, seed number, and hence seed yield isdependent largely on the number of seeds per spikelet andnot the number of fertile tillers. Final seed yield is often lessthan one-tenth of the theoretical potential seed yield(Hebblethwaite, Wright and Noble, 1980). The maximumnumber of seeds per spikelet is determined shortly beforeear emergence, by which time the number of florets hasbeen fixed. The subsequent events of anthesis, pollinationand fertilization follow to determine the seed-set componentof the total yield. Hill (1980) recently reviewed the effectsof environmental conditions over this period on seed yieldin perennial ryegrasses. In this paper, we have examined theeffects of climate at anthesis on seeds per spikelet, seednumber and seed yield of perennial ryegrass cv. S.24 atSutton Bonington over a 10 year period.MATERIALS AND METHODSSeed yield and component data were drawn from thesame series of experiments described by Hampton andHebblethwaite (1983), carried out at the University ofNottingham, School of Agriculture, Sutton Bonington,Loughborough, Leics., UK between 1971-80. Meterologi-I. Contribution from Department of Agriculture and Horticulture,School of Agriculture, University of Nottingham, SuttonBonington, Loughborough, Leics., U.K. Received 31 August1982.2· Graduate research fellow and senior Lecturer in Agronomyrespectively, University of Nottingham, School of Agriculture,Sutton Bonington, Loughborough, Leics., U.K.cal data were obtained from records for the years 1971-80(Reports, School of Agriculture, University of Nottingham1970/71-1976/79; unpub. data). Data for minimum, maximumand mean screen temperature (C), daily wind run(km h- 1 ), rainfall (mm), sunshine (h), mean daily radiation(MJ m- 2 ) and relative humidity, for the month of June(during which anthesis occurs in perennial ryegrass cv.S.24), the week of peak anthesis (usually the third week ofJune - peak anthesis was recorded from 11-23 June overthe 10 years), and the week following peak anthesis wasused. Meteorological data were recorded from a site whichwas always within 1 km of experimental plots, and therange of data recorded is presented in Table 1. Yield datawere taken from control (i.e. lodged) plots over the 10years.RESULTSMinimum screen temperature was significantly and positivelycorrelated with seed number and yield at all threetime periods (Table 2) and with seeds per spikelet at andafter anthesis. This one factor accounted for over 70% ofthe variance in seed numbers recorded. Some of thevariations in the data recorded are presented in Table 3. Theonly other factor significantly related to seed number waswind velocity in the week following anthesis (Table 2)where seed numbers were reduced by increasing windvelocity.Rainfall during the week of peak anthesis significantlyaffected seed yield (Table 2), but not seed numbers or seedsper spikelet. Mean daily radiation, sunshine, and relativehumidity were not significantly related to either yield or itscomponents.DISCUSSIONThis examination of the effects of macro-environmentalchanges on seeds per spikelet, seed number and yield,although accurate, was crude in relation to the microenvironmentwithin the crop. However, it has been adequateto highlight the importance of low temperature indetermining seed numbers at this location. Low temperaturesare known to slow down or even inhibit anthesis inperennial ryegrass (Jones and Brown, 1951; Hill, 1980),and Hill (1971) found that low night temperatures decreasedflowering intensity and reduced the number of florets openat peak anthesis. Emecz (1961) found that a threshold levelof 14 C was required before anthesis would begin inperennial ryegrass cv. S.24. Wheat plants are known to bemost susceptible to damage by low temperature when theyare at the point of anthesis (Olugbemi, 1968), because at21

22Table 1. Range of meteorological data associated with seed number, yield and seeds per spikelet in perennial ryegrass cv. S.24, SuttonBonington, 1971-80.Month of JuneWeek of Qeak anthesisWeek after peak anthesismin.max.min. max. min. max.Min. screen temp. (C) 7.7 9.97.8 10.0 8.0 10.2Max. screen temp. (C) 15.7 23.215.2 22.8 15.5 24.1Mean screen temp. (C) 11.9 17.011.1 15.7 ll.8 16.1Windrun (km h- 1 ) 7.2 9.96.5 10.9 5.1 10.1Sunshine (h) 122 22924.5 52.9 15.9 52.8Rainfall {mm} 10 1090.1 44.7 1.7 23.8Table 2. Relationship between environmental factors duringJune; week of peak anthesis; week after peak anthesis, and a)seeds m·2,b) seed yield, c) seeds per spikelet.Month Week of Week afterof peak peakJune anthesis an thesisa) Seed numberMin. screen temp. (C) 0.871** 1 0.872**Windrun (km h- 1 ) -0.467 0.109Sunshine (h) 0.361 0.566Rainfall (mm) 0.356 0.287Relative humidity -0.453 0.327Mean daily radiation(MJ m·2 ) -0.450 -0.195b) Seed yieldMin. screen temp. (C) 0.737* 0.627*Windrun (km h- 1 ) -0.322 0.251Sunshine (h) 0.176 0.144Rainfall (mm) 0.603 0.666*Relative humidity -0.423 -0.121Mean daily radiation(MJ m·2 ) -0.416 -0.324c) Seeds per SpikeletMin. screen temp. (C) 0.456 0.749*Windrun (km h- 1 ) 0.469 0.567Sunshine (h) 0.443 0.516Rainfall (mm) 0.017 O.ll7Relative humidity -0.317 0.295Mean daily radiation{MJ m·2 } -0.404 -0.5061correlation between data sets d.f. = 80.842**-0.644*0.4990.076-0.030-0.2070.629*-0.3870.3420.118-0.320-0.2540.806**-0.5010.596-0.2590.381-0.457this stage the reproductive organs are fully developed andtherefore most susceptible to temperature stress (Hill, 1980).Temperature may also affect processes after anthesis. Hill(1971) found that a 2 C frost during the first few days ofseed development reduced seed set in perennial ryegrass.Wind velocity was the only other factor found to significantlyaffect seed number. Wind velocity is important inhindering or even inhibiting anthesis in a range of grasses(Hill, 1980). A velocity of 28 km h·' inhibits anthesis inperennial ryegrass cv. S.24 (Emecz, 1961) although velocitiesrecorded in this study were only between 5-10 km h- 1 •Hill (1980) has reviewed the effects of rainfall, lightintensity and relative humidity on ear development ingrasses, and all have been shown to influence anthesis(Smith, 1944; Emecz, 1961; Hill, 1971; Vough, 1975). Inthis study, none of these factors were significantly related toseed numbers or seeds per spikelet, although rainfall duringTable 3. Variations in temperature at anthesis and seed numberand f!eld, perennial !}'e~ass cv. S.24, Sutton Boning!on.Seed number Seed yield Mean screen temperature, YearC June{m·2 x 10 4 } {t ha· 1 } (Min.} {Max.} (Mean}3.48 0.69 7.8 16.7 12.3 19774.16 0.98 7.7 16.7 12.2 19726.37 1.24 9.6 18.1 13.8 19808.66 1.33 9.9 19.6 14.8 1973the week of peak anthesis significantly increased seed yield.However, this was probably linked to events after anthesis,through a reduction in soil moisture deficits. It is evidentthat further work is required to determine the relativeimportance of these climatic factors in the micro-environmentof the crop, and their interactive relationship with yield.ACKNOWLEDGEMENTSWe wish to thank all staff, in particular Mrs. S. Manison, andstudents associated with herbage seed production experiments atSutton Bonington from 1971-80. The financial assistance of theBritish Seeds Council over the 10 years is acknowledged. J. G. H.gratefully acknowledges the support of the New Zealand NationalResearch Advisory Council.REFERENCES1. Emecz, T. J. 1961. Meteorological factors and anthesis ofgrasses. Report, Welsh Plant Breeding Station for 1960. pp.125-126.2. Hampton, J. G., and P. D. Hebblethwaite. 1983. Yieldcomponents of the perennial rye grass (Lolium perenne L.)seed crop. J. of Appl. Seed Production. 1:23-25.3. Hebblethwaite, P. D., D. Wright, and A. Noble. 1980.Some physiological aspects of seed yield in Lolium perenneL.ln Seed Production pp. 71-90. P. D. Hebblethwaite (ed.),Butterworths, London.4. Hill, M. J. 1971. Ph.D. Thesis, Massey University, NewZealand.5. Hill, M. J. 1980. Temperate pasture grass-seed crops:formative factors. In Seed Production pp. 137-149. P. D.Hebblethwaite (ed.), Butterworths, London.6. Jones, M. D. and J. G. Brown. 1951. Pollination cycles ofsome grasses in Oklahoma. Agron. J. 43:218-222.7. Olugbemi, L. B. 1968. M.Agri.Sc. Thesis, University ofCanterbury, New Zealand.8. Smith, D. C. 1944. Pollination and seed formation ingrasses. J. Agric. Res. 68:79-95.9. Vough, L. R. 1975. Herbage Abstracts. 45:3383.

Yield Components of the Perennial Ryegrass (Lolium perenne L.) Seed Crop 1J. G. Hampton and P. D. Hebblethwaite 2ABSTRACTVariance in seed yield of perennial ryegrass (Lolium perenneL.) cv. S.24 at Sutton Hollington, Leics., U.K. between 1971-1980 was best explained by the number of seeds per unit area.In lodged crops, seed numbers were dependent on the numberof seeds per spikelet, and were only poorly related to thenumber of fertile tillers present at final harvest. In non-lodgedcrops, however, both the number of seeds per spikelet and thenumber of fertile tillers per unit area were positively andsignificantly related to seed numbers. Neither spikelets pertiller or thousand seed weight was significantly correlated withseed number or yield.Most perennial ryegrass seeds crops in U.K. are lodged at orbefore anthesis. Providing fertile tiller numbers are not limiting(i.e. are between 2000-4000 m·2 ), then seed numbers andhence yield depend primarily on conditions governing thenumber of seeds set per spikelet.Additional index words: lodging, seed numbers, spikelets,fertile tillers.INTRODUCTIONThe seed yield of Lolium perenne L. , like that of wheatand other cereals, (Gallagher, Biscoe and Scott, 1975)depends primarily on the number of seeds produced per unitarea (Hebblethwaite, Hampton and McLaren, 1982). Recentdata for perennial ryegrass cv. S.23 and S.24 hasshown that seed number per unit area accounted for 60-98%of yield variance (Hebblethwaite and Hampton, 1982).The yield components which constitute seed number arefertile tillers per unit area, spikelets per fertile tiller, andseeds per spikelet. Langer (1980), suggested that seednumbers, and hence seed yield depended primarily on thenumber of fertile tillers per unit area, as both Field-Dodgson(1971) and Spiertz and Ellen (1972) found highly significantcorrelations between fertile tiller population and seedyield. However, Hebblethwaite and Hampton (1982) sug-I. Contribution from Department of Agriculture and Horticulture,School of Agriculture, University of Nottingham, SuttonHanington, Loughborough, Leics., U.K. Received 31 August1982.2· Graduate research fellow and senior lecturer in Agronomyrespectively, University of Nottingham, School of Agriculture,Sutton Hanington, Loughborough, Leics., U.K.gested that once fertile tiller numbers had reached around2000 m·2 , increasing tiller numbers had little effect on eitherseed numbers or yield up to around 4000 fertile tillers m·2 •This paper discusses some of the factors affecting seednumber and hence seed yield in perennial ryegrass cv. S.24.The data are drawn from a series of experiments carried outat the University of Nottingham School of Agriculture,Sutton Bonington, Loughborough, Leicestershire, U.K.,between 1971-1980 (Hebblethwaite, Wright and Noble,1980; Hebblethwaite and Hampton, 1982).MATERIALS AND METHODSThe relationship between seed numbers and yield componentswas examined using regression analysis on datafrom 20 experiments where the crop had lodged at or beforeanthesis, and 10 experiments where, either by mechanical(Hebblethwaite, Burbidge and Wright, 1978) or chemicalmethods (Hebblethwaite, Hampton and McLaren, 1982),lodging was prevented.RESULTSIn lodged crops, fertile tiller numbers were not significantlyrelated to seed numbers (Table 1), but in non-lodgedcrops, fertile tiller numbers accounted for 75% of thevariance in seed number m·2 • In both lodged and nonlodgedcrops, seeds per spikelet were significantly correlatedwith seed number (Table 2), whereas the relationshipbetween spikelets per tiller and seed number wasnegative and not significant (r = -0.264 and -0.607 forlodged and non-lodged crops respectively). At SuttonBonington, thousand seed weight (TSW) of perennial ryegrasscv. S. 24 varied from 1. 56 - 2.18 g over the 10 years,but was not significantly related to seed number or yield.DISCUSSIONIn lodged perennial ryegrass seed crops between 1971-1981, fertile tiller numbers were not significantly related toseed numbers or to seed yield (Hebblethwaite and Hampton,1982). This conflicts with the results of Field-Dodgson(1971) and Spiertz and Ellen (1972). At Sutton Bonington,management practices stayed relatively constant over the 10years in terms of sowing time, rate, amount of nitrogen,phosphate and potassium applied (Hebblethwaite, Burbidgeand Wright, 1978; Hebblethwaite and Ivins, 1978; Wrightand Hebblethwaite, 1979; Hebblethwaite, Hampton and23

24Table 1. Relationship between seed number m·2 (Y) and fertile tiller number m·2 (X) in lodged and non-lodged perennial ryegrass(cv. S.24) crops, 1971-80.Relationship showing best fit Significance %variance data rangeaccounted for:lodged (d.f = 18) Y = 5.47( ± 2.02)+0.0003( ± 0.0007)X NS -5.2 seeds m·2 X 10 4 : 3.5 - 10.5fertile tillers m·2 : 1338 - 3865non-lodged (d.f = 8) Y = 1.01( ± 0.52)+0.002( ± O.OOl)X ** 73.7 seeds m·2 X 10 4 : 5.5 - 16.3fertile tillers m·2 : 1540- 4049Table 2. Relationship between seed number m·2 (Y) and seeds per spikelet (X) in lodged and non-lodged perennial ryegrass (cv.S.24) crops, 1971-80.Relationship showing best fit Significance %varianceaccounted for:data rangelodged (d.f = 18) Y = 2.46( ± 0.79)+2.69( ± 0.52)non-lodged (d.f = 8) Y = 2.78( ±0.24)+3.82( ± 1.10)****59.855.1seeds m·2 x 10 4 :seeds per spikelet:seeds m·2 x 10 4 :seeds per spikelet:3.5 - 10.50.76- 2.505.5 - 16.30.99- 3.39McLaren, 1982), so that the environment provided thebiggest effects on seed yield and its components. Incontrast, Fie1d-Dodgson (1971) obtained greater seed yieldsthrough the production of more fertile tillers in response tonitrogen application, while Spiertz and Ellen (1972)achieved the same result through manipulation of lightintensities in the spring. In both cases, overcoming deficienciesin the crop through the application of nitrogen andmore light respectively provided the linear relationshipbetween fertile tiller numbers and yield. Spiertz and Ellen(1972) also found that when light intensities were changedin autumn, the relationship between fertile tiller numbersand seed yield was negative and non-significant, becauserestricted autumn tiller development produced a flush oflow yielding spring-formed fertile tillers.Fertile tiller numbers may be decreased in lodged cropsthrough rotting of the base of the stem in a wet year(Burbidge, 1977), but in 6 of the 10 experiments, fertiletiller numbers were not significantly different at finalharvest in lodged and non-lodged crops.In non-lodged crops, fertile tiller numbers were significantlyrelated to seed number, but not to seed yield(Hebblethwaite and Hampton, 1982), because althoughseeds per spikelet did not change significantly, spikelets pertiller decreased significantly with increasing fertile tillernumbers (Hampton, unpub. data).In both lodged and non-lodged crops, spikelets per tillerwere not significantly related to seed number. The numberof spikelets per tiller is dependent on the number ofreproductive primordia accumulated at the shoot apex at theonset of reproductive development in the spring, and on thenumber of additional primordia delimited during apicaldevelopment (Ryle, 1964). Factors affecting spikelet numberper tiller have been recently discussed (Hebblethwaite,Wright and Noble, 1980).Langer and Lambert (1963) noted that in grass, TSWappeared remarkably uniform. Although affected by the ageof tillers (Anslow, 1964), timing of nitrogen (Hebblethwaiteand Ivins, 1978) and irrigation (Wright, 1978), variation inTSW from season to season, which was not related to seednumber or yield, is possibly best explained through theflexibility of the rye grass plant in its adjustment of the yieldcomponents in response to various environmental factors.Thus in some seasons, seed number explained over 95% ofthe seed yield variation (Hebblethwaite and Hampton,1982) while in others, TSW also had a limited influence.Lodging occurs in most perennial ryegrass seed cropsgrown in U.K. Providing fertile tiller numbers are not alimiting factor (because of poor establishment, moisturestress or a shortage of nitrogen), then it appears that seednumbers are determined primarily by conditions governingthe number of seeds per spikelet (Hill, 1980), and not byfertile tiller number.ACKNOWLEDGEMENTSWe wish to thank all staff, in particular Mrs. S. Manison, andstudents associated with herbage seed production experiments atSutton Bonington from 1971-1980. The financial assistance overthe ten years by the British Seeds Council is gratefully acknowledged.J.G.H. also gratefully acknowledges the support of theNew Zealand National Research Advisory Council.REFERENCESI. Anslow, R.C. 1964. Seed formation in perennial rye grass. II.Maturation of seed. J. of the British Grassland Soc. 19:349-357.2. Burbidge, A. 1977. Lodging and its control in Lolium perenne,grown for seed. Ph.D. Thesis, Univ. of Nottingham, U.K.3. Field-Dodgson, J.R.C. 1971. The effect of nitrogen onryegrass seed production. M. Agr. Sc. Thesis, LincolnCollege, New Zealand.4. Gallager, J. N., P. V. Biscoe, and R. K. Scott. 1975. Barleyand its environment. V. Stability of grain weight. J. Appl.

25Ecol. 12:319-336.5. Hebblethwaite, P. D., A. Burbidge, and D. Wright. 1978.Lodging studies in Lolium perenne grown for seed. 1. Seedyield and yield components. J. of Agric. Sci., Cambridge.90:261-267.6. Hebblethwaite, P. D., and J.D. Ivins. 1978. Nitrogen studiesin Lolium perenne grown for seed. II. Timing of nitrogenapplication. J. of the British Grassland Soc. 33:159-166.7. Hebblethwaite, P. D., D. Wright, and A. Noble. 1980. Somephysiological aspects of seed yield in Lolium perenne L. InSeed Production. pp. 71-90. P. D. Hebblethwaite (ed.).Butterworths, London.8. Hebblethwaite, P. D., andJ. G. Hampton. 1982. Physiologicalaspects of seed production in perennial ryegrasses. In BreedingHigh Yielding Forage Varieties combined with High SeedYield. pp. 17-32. Report Fodder Crops Section Meeting,Eucarpia, Gent, Belgium, 1981.9. Hebblethwaite, P. D., J. G. Hampton, and J. S. McLaren.1982. The chemical control of growth, development and yieldof Lolium perenne grown for seed. In Chemical Manipulationof Crop Growth and Development. pp. 505-523. J. S. McLaren(ed.). Butterworths, London.10. Hill, M. J. 1980. Temperate pasture grass-seed crops: forma-tive factors. In Seed Production pp. 137-149. P. D. Hebblethwaite(ed.). Butterworths, London.11. Langer, R. H. M., and D. A. Lambert. 1963. The physiologicalsignificance of population density in grass-seed production.J. of the British Grassland Soc. 18:177-180.12. Langer, R. H. M. 1980. Growth of the grass plant in relation toseed production. In Herbage Seed Production. pp. 6-11.Grassland Research and Practice Series No. 1. New ZealandGrassland Association.13. Ryle, G. J. A. 1964. The influence of date of origin of theshoot and level of nitrogen on ear size in three perennialgrasses. Annals of Appl. Bioi. 53:311-323.14. Spiertz, J. H. J., and J. Ellen. 1972. The effect of lightintensity on some morphological and physiological aspects ofthe crop perennial ryegrass (Lolium perenne L. var. 'Cropper')and its effects on seed production. Netherlands J. of Agric.Sci. 20:232-246.15. Wright, D. 1978. Control of growth, development and seedproduction in Lolium perenne. Ph.D. Thesis, University ofNottingham, U.K.16. Wright, D., and P. D. Hebblethwaite. 1979. Lodging studiesin Lolium perenne grown for seed. 3. Chemical control oflodging. J. of Agric. Sci., Cambridge. 93:669-679.

Preliminary Investigations of Adhesive Sprays To ImproveSeed Retention in Tropical Grasses 'D. S. Loch and G. L. Harvey 2ABSTRACTPreliminary experiments on Chloris gayana Kunth cv.Callide and Setaria sphacelata (Schumach.) Stapf & Hubbardex M. B. Moss var. sericea (Stapf) W. D. Clayton cv. Narok insouthern Queensland, Australia assessed the effects of glue andpaint-derived sprays on seed retention. InS. sphacelata var.sericea, adhesive sprays formed droplets on inflorescence hairsand did not improve seed retention. In contrast, spray coverageand penetration of C. gayana inflorescences were more effectiveand results were encouraging, with reduced shattering and anextension of the potential harvest period by about two weeks. Inany future work, high priority should be given to the possibleuse of additives (e.g. wetting agents) to achieve better coverageand penetration of inflorescences by adhesive mixtures.Additional index words: shattering, seed production.INTRODUCTIONShattering, or the loss of ripe seed from inflorescences,generally allows little margin for error in the timing ofharvest for tropical grass seed crops. Together with theproblem of uneven crop ripening caused by the spread ofmaturity between and within inflorescences, shatteringcontributes to the relatively low seed yields from tropicalgrasses and the high cost of seed of many cultivars.McWilliam and Schroeder (1974) sprayed inflorescencesof Phalaris aquatica L. (formerly Phalaris tuberosa L.)with a quick-setting plastic lacquer (PTSB or Phalaristuberosa seed coat builder, a paint-derived product- H. E.Schroeder, personal communication) just prior to seedmaturity. This treatment reduced shattering and is thereforea promising technique to improve both the yield and qualityof harvested seed, although more developmental work isnecessary before it can be applied successfully on a commercialscale.Essentially the same approach was used by Williams(1976, 1978, 1979) in South Australia on vegetable crops.In his work, a polyvinyl acetate (PV A) glue was applied andl. Contribution from Agriculture Branch, Queensland Departmentof Primary Industries, Brisbane, Queensland, Australia.Received 12 November 1982.2· Senior Agrostologist and Experimentalist respectively, QueenslandDepartment of Primary Industries. Mailing Address: Departmentof Primary Industries, P.O. Box 395, Gyrnpie, Queensland,Australia, 4570.showed promise even on a commercial scale.The high value of tropical grass seeds makes this techniquepotentially attractive for such crops. Unlike temperategrasses, however, the seed of tropical grasses abscissesabove the glumes (Hacker and Jones, 1971; Strickland,1971; Humphreys, 1979), thereby reducing the possibilityof improving seed retention through the use of adhesivesprays, especially in very open panicles (e.g. Panicumspp.). Grasses with light, chaffy "seeds" (e.g. Chlorisgayana Kunth) or with tight heads incorporating inflorescencehairs [e.g. Setaria sphacelata (Schumach.) Stapf &Hubbard ex M. B. Moss var. sericea (Stapf) W. D.Clayton] (formerly Setaria anceps Stapf) therefore offer thebest chances of success. This paper covers the results ofpreliminary experiments evaluating the use of glue andpaint-derived sprays on such species.BACKGROUNDChloris gayana and Setaria sphacelata var. sericea aremajor components of sown pastures in tropical regions andthe particular cultivars used have been released as superiorpasture types (Barnard, 1972; Bogdan, 1977). Both grassesstand about 1.8-2.0 m high in full head, butS. sphacelatavar. sericea forms distinct tussocks and produces tightspike-like panicles with dense, free-flowing seeds, in contrastto C. gayana which spreads by stolons and producesdigitate or sub-digitate panicles comprising one-sidedsessile spikes with light, chaffy seeds. Because of theperennial nature of individual plants, successive croppingcycles- conventionally designated as "crops" - can beimposed for a number of years on the same seed area asoutlined by Loch (1980, 1982).MATERIALS AND METHODSExperiment 1An established 39 x 16m area of C. gayana cv. Callide(sown November 1971) was mown to 15 em on 16 February1976 and fertilized with 150 kg N (as ammonium nitrate),150 kg superphosphate, and 125 kg muriate of potash perha. This was situated on a Udic Argiustoll (Soil SurveyStaff, 1975; B. Powell, personal communication) at UpperWonga (26°10'S, 152°25'E, with 1000 mm average annualrainfall) in southern Queensland.Walkways (1 m wide) were mown out on 6 April 1976leaving 12 plots, each of 8 x 1.5 m, which were allocatedrandomly within pairs to either an unsprayed control (TreatmentA) or a sprayed Treatment B. Spraying (with 1 part26

27PTSB in 7 parts of water) was carried out at a rate of·approximately 1400 1/ha of mixture on 18 May 1976, justprior to crop maturity.Quadrat samples (0.5 m 2 ) were cut at 14 day intervalsfrom randomly allocated sub-plots (1.6 x 1.5 m) and thepercentage of empty inflorescences (i.e. with seed completelyshed) was calculated as follows: -No. empty inflorescences x 100Total no. inflorescencesTo facilitate hand-stripping of seed, freshly-cut inflorescencesamples were "sweated" for three days by storingthem between sheets of damp newspaper to maintain a highmoisture content. Samples were then dried at 40 C and theseed stripped and weighed. The proportion of pure seed wasdetermined in two alternative ways. Method 1 (the Irishmethod) is used for testing chaffy grass seeds for sale inAustralia and classifies as pure seed all intact spikeletsminus the unequal basal glumes, irrespective of whether acaryopsis is present or absent. Method 2 (the internationalmethod) imposes the additional refinement that pure seedmust contain a caryopsis, and is in accord with InternationalSeed Testing Association (ISTA) rules (Anon., 1976).Experiment 2A 24 x 24m area of S. sphacelata var. sericea cv. Narokwas sown in 50 em rows on 14 November 1975. This wassituated on a Paleustalf (Soil Survey Staff, 1975; R. C.McDonald, personal communication) at "Bungawatta"(25°57'S, 152°42'E, with 1250 mm average annual rainfall)in southern Queensland.Two cropping cycles were studied in this experiment.The stand was initially mown to 15 em on 18 August 1976(crop 1) and 10 January 1978 (crop 2) and fertilized with150 kg N/ha (as ammonium nitrate). In addition, crop 1received 250 kg superphosphate and 50 kg muriate ofpotash/ha, while crop 2 received 375 kg superphosphateand 125 kg muriate of potash/ha. Mown walkways (1 mwide) were maintained, leaving 27 plots each 6 x 1.5 m forthe following treatments arranged in nine randomizedblocks:-A - Unsprayed controlB - Sprayed once:Crop 1 - 30 December 1976Crop 2- 22 March 1978C - Sprayed twice:Crop 1 - 30 December 1976, 28 January 1977Crop 2- 22 March 1978, 19 April 1978Because PTSB was unavailable, a commercial watersolubleacrylic glaze (Duralex Aqua-Clear Flat) was used oncrop 1 at a dilution of 1 part in 7 parts of water and appliedat 2000 1/ha of mixture. For crop 2, this was changed to aPV A glue without toxic plasticisers (thalates) or acryliccompounds (Vinamul 9300) at 1 part in 10 parts of waterand the mixture applied at 1650 1/ha.As in experiment 1, quadrat samples (0.5 m 2 ) were cutat 14 day intervals and percentages of empty inflorescencescalculated. Seed was "sweated" and hand-stripped andsamples were lightly cleaned by conventional screening andwinnowing before recovering pure seed using a modifiedOttawa seed blower descibed by Hergert, Zillinsky andKemp (1966).RESULTSIn experiment 1, sprayed plots had significantly fewerempty inflorescences (Figure 1) and more seed per inflores-30en(.)c:Q)(.).....~ 200;;:::c:..::- 10a.Ew(f.0A = controlB = treated14 28 42Days Since SprayinglL.S.D.(P=0.05)Figure 1. Change in percent empty inflorescences of Chlorisgayana cv. Callide (experiment 1).cence (Figure 2) 28-56 days after spraying, indicating thatPTSB delayed seed shedding for around two weeks.Spraying also appeared to increase both the percentage andthe size of caryopses present in spikelets sampled duringthis period (data not presented). Our observations, however,suggested that spray penetration of inflorescences could beimproved, even though plots were sprayed from the sidesand (as appeared more effective) from above.In experiment 2, neither of the mixtures used waseffective in penetrating S. sphacelata var. sericea inflorescences.Although plots were thoroughly wet afterspraying, droplets merely formed on inflorescence hairsand, as indicated in Figure 3, spraying had no effect on seedretention. Spraying also appeared to increase both thepercentage and the size of caryopses present in spikeletssampled during this period (data not presented).DISCUSSIONEffective coverage and penetration of inflorescences areessential if adhesive sprays are to improve seed retention intropical grasses. In the preliminary experiments reportedhere, the importance of these aspects was clearly demonstrated:inS. sphacelata var. sericea where spray dropletsdid not appear to penetrate beyond the inflorescence hairs,adhesive sprays were ineffective; in contrast, spray coverageand penetration of C. gayana inflorescences were moreeffective and results were encouraging, with reduced shatteringand an extension of the potential harvest period byabout two weeks. In addition, a few S. sphacelata var.sericea inflorescences from crop 2 were dipped directly intothe adhesive spray mixture to ensure proper coverage andpenetration, and our subsequent observations of these indicatedthat seed retention was markedly improved. Similar56

28250 a) Purity Method Ia) Cropping Cycle I200Seed15010040(i)(,)c(I)(,)en(I)....500ij:: 0 14 28 42 56c'~Ec>0ww(/)w0::::lCl..Days Since Sprayingb) Purity Method 20 14 28 42 56Days Since SprayingFigure 2. Change in seed per inflorescence of Chloris gayana cv.Callide (experiment 1) determined by (a) purity method 1and (b) purity method 2.(i) 20(,)c(I)(,)en(I)0 0ij::c'c>-50w(/)w 600::::lCl..40........... :::::= .. ::::-;:::;:-- -·~--·' ,.,-

29to A. C. Hatrick Chemicals Pty. Ltd. for Vinamul 9300.REFERENCES1. Anon. 1976. International Rules for Seed Testing. Annexes1976. Seed Sci. and Tech. 4:51-177.2. Barnard, C. 1972. Register of Australian Herbage PlantCultivars. C.S.I.R.O. Division of Plant Industry, Canberra,Australia.3. Bogdan, A. V. 1977. Tropical Pasture and Fodder Plants.Tropical Agriculture Series, Longman, London and NewYork.4. Budd, E. G. 1981. Preliminary research note on sheddinglosses in grass seed crops. J. of the National Institute ofAgricultural Botany 15:552-555.5. Hacker, J. B., and R. J. Jones. 1971. The effect of nitrogenfertilizer and row spacing on seed production in Setariasphacelata. Tropical Grasslands 5:61-73.6. Hergert, G., F. J. Zillinsky, and J. K. Kemp. 1966. Anaspirator for cleaning small seed samples. Canadian J. of Pl.Sci. 46:570-572.7. Hopkinson, J. M., and B. H. English. 1982. Spikeletpopulation dynamics in seed crops of Panicum maximum'Gatton'. Seed Sci. and Tech. 10:379-403.8. Humphreys, L. R. 1979. Tropical Pasture Seed Production.2nd Ed. FAO Plant Production and Protection Paper 8.9. Loch, D. S. 1980. Selecton of environment and croppingsystem for tropical grass seed production. Tropical Grasslands14:159-168.10. Loch, D. S. 1982. Seed production research- agronomicaspects. Tropical Grasslands 16:88-90.11. McWilliam, J. R., and H. E. Schroeder. 1974. The yield andquality of phalaris seed harvested prior to maturity. Australian J.of Agric. Res. 25:259-264.12. Soil Survey Staff. 1975. Soil Taxonomy. United StatesDepartment of Agriculture. Agric. Handbook 436. U.S.Government Printing Office, Washington, D.C.13. Strickland, R. W. 1971. Seed production and testing problemsin tropical and sub-tropical pasture species. Proc. of theInt. Seed Testing Assoc. 36:189-199.14. Williams, C. 1976. Glue spraying increases vegetable seedyield. Fact Sheet. Department of Agriculture and Fisheries,South Australia.15. Williams, C. M. J. 1978. Seedhead glueing increases yieldand quality in vegetable seed crops. Abstracts, XX Int.Horticultural Cong. No. 1656.16. Williams, C. M. J. 1979. Glue spray applications to reducecarrot and onion seed shattering. Paper presented at 8thNational Onion and Carrot Seed Conference. Univ. ofCalifornia, Davis.

The Effects of Method and Time of Sowing on Seed Productionin Perennial Ryegrass 1P. D. Hebblethwaite and S. D. Peirson 2ABSTRACTThe effects of method and timing of sowing on seed yield ofperennial ryegrass CV. S.24 were investigated in a series offield experiments from 1978 to 1981. Undersowing in the springin barley resulted in similar yields to sowing direct except in1979 when yields were reduced. Yields from autumn directsowings when carried out before mid September were similar toundersowing and direct sowing in the spring. Sowing after midSeptember in most years decreased seed yield and these reductionswere due to decreases in seeds per unit area.Additional index words: undersowing, seed yield components,lodging, tillering.INTRODUCTIONThe most popular method of establishing ryegrass seedcrops in the United Kingdom is by undersowing in springbarley (Denchfield, 1972). However, undersowing canprovide a number of problems. Firstly, there is often intensecompetition for soil moisture, nutrients and light betweenthe cereal crop and establishing grass. This can either resultin excessive grass growth which decreases cereal yield andcauses problems at cereal harvest or more often the cerealcrop severely decreases the establishment and growth of theryegrass. To some extent the adverse effects of the nursecrop can be lessened by decreasing cereal seed rates,nitrogen levels, widening cereal row width or drilling afterthe grass crop has emerged. But any of these methods thatfavor the grass seed crop are likely to decrease cereal yields.In Holland and Denmark the majority of perennialryegrass seed crops are direct sown in the autumn after thecereal harvest (Bawcutt and Hebblethwaite, 1978;Hebblethwaite, 1978). In recent years there has been agrowing interest in autumn sowing in the United Kingdomparticularly with the large increase in area of winter barleywhich is harvested early. Little research has been carriedout in the United Kingdom comparing undersowing withautumn sowing and no work has been published defining1· Contribution from Department of Agriculture and Horticulture,School of Agriculture, University of Nottingham, Sutton Bonington,Loughborough, Leics., U.K. Received 9 June 1983.2· Senior lecturer in Agronomy and undergraduate student, respectively,University of Nottingham, School of Agriculture,Sutton Bonington, Loughborough, Leics., U.K.the optimum autumn sowing date.Trials were carried out from 1978 to 1981 to assess theeffects of sowing method and time on the growth, development,seed yield components and seed yields of perennialrye grass.MATERIALS AND METHODSThe experiments were carried out on the University ofNottingham Experimental Farm, Sutton Bonington,Loughborough, Leics. on soil of the Astley Hall Series. Allexperiments were sown with basic seed of cv. S.24 at 12 kgha- 1 on 15 centimeter rows in plots 1.5 x 12 m. Actualsowing date treatments were within a few days of each otherin each year and average time periods are given in Table 1.Table 1. The effect of method and date of sowing on the dryseed :y!eld t ha-l, 1979 to 1981Year1978 1979 1980 1981 Mean of1979 to 1981Undersownin spring barleyMarch/ April 1.9 1.3 1.4 1.5Drilledwith no cover cropMid March 1.2 2.4 1.4 1.4 1.7Mid Aug. 1.3 2.1 1.3 1.3 1.6Early Sept. 2.1 0.9 1.1 1.4Mid Sept. 1.0 1.2 1.1 1.1 1.1Late Sept./early Oct. 1.2 0.9 0.9 1.0Late Oct. Nov. 0.7 1.3 0.9 0.5 0.9LSD (5%) 0.28 0.34 0.22 0.34 0.21S.E.D. (21 D.F.)* 0.11 0.20 O.Q7 0.15 0.10c.v. {%2 10 18 8 16 35* 1978 (12 D.F.)-- no treatment applied.All experiments utilized a randomized block design replicatedfour times. All plots received 60 kg N ha- 1 , 40 kg Pz0 5ha- 1 and 40 kg KzO ha- 1 at establishment and 120 kg N ha- 1at spikelet initiation in the spring of the harvest year. In1979 and 1980 ethofumesate was applied at 10 1 ha- 1 duringthe autumn to control Poa annua and any possible Poatrivia/is. In 1981 a mixture of half rate (5 1 ha- 1 ) ethofumesateand T C A (1.1 kg ha- 1 ) were used in order tocontrol Poa trivia/us and high levels of volunteer winterbarley. A mixture of 2, 3, 6 TBA, dicamba, M CPA andmecoprop was applied in spring to control broadleaved30

31Table 2. The effect of method and date of sowing on the seed components (mean of 1979 to 1981}.Fertile Spikelets/ Seeds/ Seeds/ Seeds 1000 seed Dry seedtillers fertile spikelet fertile m2 wt. (gm) yieldm2 tiller tiller x 10-2 (t ha- 1 )Undersown in spring barleyMarch/ April 2569 19.6Drilled with no cover cropMid March 2471 19.5Mid August 2735 20.5Early Sept. 2437 17.8Mid Sept. 2975 16.2Late Sept./early Oct. 2518 14.6Late Oct./Nov. 2538 15.2L.S.D. (5%) 981 2.42S.E.D. (21 D.F.} 467 1.152.04 40 1023 2.10 1.52.30 44 1102 2.08 1.71.72 36 977 2.00 1.61.90 34 829 2.04 1.41.58 27 761 2.07 l.l1.78 26 622 2.11 1.01.56 24 619 2.12 0.90.698 25.3 309.2 0.227 0.210.337 12.2 149.4 0.109 0.10weeds, the major problem in autumn sowings being Stellariamedia.In 1979 and 1980 each plot was prepared by plowing inthe autumn followed by cultivating with spring tynes,harrowing and then rolling before sowing. Consequently,autumn sowings followed a fallow in these years. In 1981autumn sowings followed winter barley which was harvestedon 8 August. These plots were sprayed with paraquatand diquat to kill any regrowth and then immediately beforesowing were spike rotary cultivated twice and then Cambridgerolled. In all years plots were Cambridge rolled aftersowing. In all years spring sown treatments were defoliatedat the time of combining the undersown with barleytreatment.In 1979 plots were direct combined and in 1980 and 1981were Mayfield harvested and threshed. Details of harvesting,growth analysis, lodging scores and yield componentmeasurement techniques used at Sutton Boningtonhave been previously published (Hebblethwaite and Burbridge,1976; Hebblethwaite et al, 1978).RESULTSSeed Yield and Seed Yield ComponentsUndersowing in spring barley when compared to sowingwith no cover crop at the same date decreased seed yield in1979 but not in 1980 and 1981 and on average (Table 1).Sowing later than early September decreased yield in mostyears and on average (Table 1). Decreases in seed yieldwere due to reductions in seeds harvested per unit area(Table 2). In 1981 yield was not decreased until sowing wasdelayed to early October possibly because of the exceptionallywet and mild autumn and winter of 1980/81. Ineach year the decreased number of seeds per unit area wasassociated with a decrease in the number of spikelets pertiller, seeds per tiller and per spikelet and not with numberof fertile tillers and individual seed weight (Table 2). Asresponses were similar in all years only average data for1979 to 1981 are presented in Table 2.DevelopmentGrowth stage and time of harvest at 30 to 35% moisturecontent (m.c.) was similar in all treatments up to sowing inlate September/early October. Delaying sowing to lateOctober/November on average delayed the time at whichthe crop reached 70% ear emergence, first anthesis, peakanthesis and final harvest at 30 to 35% m.c. by 14, 13, 8and 7 days respectively.Culm length and lodgingNo differences in culm length (stem + ear) measured inthe field were found until sowing was delayed until lateSeptember. For example, on 23 May 1980 the averageheight for sowings prior to late September was 58 emcompared to 44 and 34 em in late September and earlyNovember sowings. By final harvest these differences wereno longer evident and all treatments had reached a height ofapproximately 73 em. The degree of lodging was closelyrelated to crop height; for example, on 29 May the percentageof crop that was lodged in sowings prior to lateSeptember was similar and averaged 58% compared to 25%and 0% in the later two sowings. By anthesis on 12 Junelodging percentages for these treatments were 75, 55 and35% respectively. However, by final harvest all treatmentswere completely lodged. Similar results were found in 1979and 1981 -data is therefore not presented.Tille ringTillering was recorded in 1979 and 1981, but as the resultswere similar in both years and for simplicity, data for aselective number of treatments only are presented in Figs. 1and 2. Total number of tillers during the late autumn andwinter period remained constant in all treatments butnumbers were considerably lower for sowings carried outfrom mid September onwards (Fig. 1). Spring growth wasassociated with a rapid increase in the number of tillers m- 2 ,the maximum number attained being greater in the earliersowings. Maximum number was obtained in late March inall treatments except the late October sowing where this wasdelayed by one month. Thereafter, when the first earsemerged, the number of tillers decreased markedly and at asimilar rate in all treatments except in the late Octobersowing where the decrease was small. In all treatments, andparticularly in the case of the mid September sowing, tillernumber increased towards final harvest due to secondary

323IIFig. 1. The effect of sowing method and date on the totalnumber of tillers. 0, March/April; .A., Mid-September;II , Late October/November; e, Undersown in springbarley in March/ April.vegetative tiller growth. By final harvest there was nodifference in tiller number between treatments.Delaying the time of sowing delayed ear emergence andthe time the crop reached maximum emergence (Fig. 2).March direct and undersown treatments reached maximumemergence in early June but emergence continued in midMarch and October sowings until final harvest (Fig. 2).Total dry matter accumulationIn all treatments no dry matter was accumulated duringthe winter period (Fig. 3). In late March rapid dry matteraccumulation took place in all treatments except in the lateOctober sowing where this increase was delayed until May(Fig. 3). Sowing in mid September and late Octoberdecreased crop dry weight at all sampling dates up to midMay. However, by early July there was no significantdifference between treatments in crop dry weight (Fig. 3).WeedsIn 1981 autumn sowing followed a winter barley crop(see Materials and Methods) and in spite of herbicideapplication (ethofumesate plus T C A), this resulted in highpopulations and consequently dry weights of winter barleyvolunteers and Poa trivialus particularly in the late sowntreatments (Fig. 3). The barley was almost completely eatenby birds before harvest particularly in the November sowntreatment but the Poa grew considerably just before harvestas the ryegrass senesced and this could have caused harvestingproblems. The effect of the volunteer barley on seedyield was assessed on the mid September sown treatment by(\J......sMI~(/) 2s:..aJrlrl.... ..,aJ.... >..,'00s:..0.aJs:.."-zMay June July AugFig. 2. The effect of sowing method and date on the number ofreproductive tillers. 0, March/ April; A., Mid-September;II , Late October/November; e, Undersown in springbarley in March/ April.1400120100N-;;~ 800~§~600li400200Nov Dec j Jan I FebFig. 3. The :ffect of sowing method and date on the total cropdry wetght. 0, March/April; A., Mid-September;II , Late October/November; e, Undersown in springbarley in March/ April.

33hand rogueing an area 2 m 2 on 21 April and harvestingseparately. Interestingly, there was no significant differencein seed yields between the two areas and this may have beenbecause rogueing was too late and possibly caused somecrop damage.In all years, populations of Stellaria media were higher inautumn sown plots but this was easily controlled byhormone herbicide (see Materials and Methods).DISCUSSIONThese results show that undersowing in spring barley is asatisfactory method of establishing ryegrass seed crops inmost years and similar results have been found by otherworkers (Roberts, 1958; Kellner et al, 1967; Nordestgaard,1981). However, this work shows that growers wishing tochange from spring to winter barley will not suffer a yieldreduction by direct sowing perennial ryegrass after thebarley unless they sow later than the first week ofSeptember. Work in Denmark on early and medium earlyperennial ryegrass has shown that any delay in sowing afterthe end of August will decrease yields (Nordestgaard,1981). Consequently growers sowing perennial ryegrassafter winter barley should sow as soon as possible after thebarley harvest.In years with mild winters reasonable seed yields can beobtained from so wings in late September and early October.Nevertheless, it is important to remember that the farm fieldsituation is not likely to be as precise as that found inexperimental plots. Similar results have been found byworkers from a series of trials at the National Institute ofAgricultural Botany, Cambridge (A.W. Evans, Personalcommunication).Decreases in yield for late sowings were due to decreasesin ear size and seed set and not fertile tiller numbers. Anumber of workers have shown that ear size (spikeletnumbers) decreases the later the date of origin of the shoot(Ryle, 1964; Hill and Watkin, 1975; Hebb1ethwaite andClemence, 1982). This decrease in spikelets was found tobe correlated with a decrease in the number of unexpandedleaf primordia accumulated at the shoot apex before induction(Ryle, 1966). Late sowings would therefore haveaccumulated a lower number of leaf primordia beforeinduction than early sowings.Autumn sowings, particularly when late, were shown tohave problems in relation to volunteer winter barley. Recentsurvey work on farms (N.I.A.B., unpublished) has shownthat spring sown grass seed crops have less cereal volunteerson average than autumn sowings. Cereal volunteerseven at fairly low populations have been found to decreaseseed yields (Wright and Hebblethwaite, 1983). Furtherwork is therefore needed on chemical control of cerealvolunteers in autumn sown ryegrass seed crops.REFERENCES1. Bawcutt, D. E. and P. D. Hebblethwaite. 1977. Report of avisit to Holland, ADAS. May 1977.2. Denchfield, H. 1972. Seed production as a method ofutilizing the grass break. B.Sc. Honours Dissertation, Universityof Nottingham.3. Hebblethwaite, P. D. 1978. Report on a visit to Denmark tostudy seed production research in grasses. Department ofAgriculture and Horticulture, University of Nottingham.4. Hebblethwaite, P. D. and A. Burbidge. 1976. The effect ofmaleic hydrazide and chlorocholine chloride on the growth,seed yield components and seed yield of S.23 ryegrass. J. ofAgric. Sci., Camb. 86: 343-353.5. Hebblethwaite, P. D., A. Burbidge and D. Wright. 1978.Lodging studies on Lolium perenne grown for seed. I. Seedyield and seed yield components. J. of Agric. Sci. , Camb.90: 261-267.6. Hebblethwaite, P. D. and T. G. A. Clemence. 1982. Theeffect of autumn and spring defoliation and defoliationmethod on seed yield of Lolium pererme. Proceedings of theInternational Grassland Congress, Lexington, Kentucky,USA, pp. 257-260.7. Hill, M. J. and B. R. Watkin. 1975. Seed production studieson perennial ryegrass, timothy and prairie grass. 1. Effect oftiller age on tiller survival, ear emergence and seed headcomponents. J. of the British Grassland Soc. 30: 63-71.8. Kellner, E., P. Varga, C. Galan, T. Popa, and M.Georgescu. 1967. Effect of fertilizers and nurse crops onseed yields of perennial ryegrass. Research Institute forcereals and industrial crops, Fundulea. Anal. Inst. Cer. Pl.Teh. 33: 69.9. Nordestgaard, A. 1981. Trials on sowing time in ryegrass(Lolium multijlorum) and perennial ryegrass (Lolium pererme)for seed production. Saertryk af Tidsskrift for Planteavl. 85:389-398.10. Roberts, H. M. 1958. The effect of defoliation on the seedproducing capacity of bred strains of grasses. 1. Timothy andperennial ryegrass. J. of the British Grassland Soc. 13:255-61.11. R yle, G. J. A. 1964. The influence of date of origin of theshot and level of nitrogen on ear size in three perennialgrasses. Annals of Applied Biology. 53: 311.12. Ryle, G. J. A. 1966. The physiological aspects of seed yieldin grasses. In: Growth of Cereals and Grasses. Proc. 12thEaster School in Agricultural Science, University ofNottingham. F. L. Milthorpe & J. D. Ivins (eds.) Butterworths,London. pp. 106-120.13. Wright, D. and P. D. Hebblethwaite. 1983. Volunteer winterwheat; its effects and control in ryegrass seed production.Weed Research (in press).

Growth, Floral Induction and Reproductive Development in Selected PerennialRyegrass Lolium perenne L. Cultivars 1A. S. Gangi, D. 0. Chilcote, and R. V. Frakes 2ABSTRACTThis research was designed to identify inductive requirementsand floral initiation characteristics of three perennialryegrass Lolium perenne cultivars. Information was also collectedon growth and development differences in response todifferent low temperature exposures (0, 1, 2, 4, 8 weeks) andphotoperiod treatments (12 and 20 hour) under growthchamber conditions. Response of imbibed seeds was comparedwith 1 and 4 weeks old seedlings exposed to low temperature todetermine the effect of plant growth stage on response to floralinductive conditions.Pennfine and Loretta cultivars were superior in tillering toLinn and Acclaim. Tiller number and dry matter productionper plant were higher in the 20 hour than the 12 hourphotoperiod.More fertile tillers were produced when the low temperatureexposure was given at the seedling stage rather than theimbibed seed stage. Inflorescence production increased as theexposure time increased.The failure of many tillers to produce spikes may have beendue to: 1. An obligate juvenile stage for these cultivars so theywere insensitive to low temperataure and/or short days at timeof exposure; 2. Insufficient duration of exposure. These cultivarsmay have requirements for low temperature exposurewhich exceeds the 8 weeks exposure treatment.Additional key words: floral initiation, photoperiod, temperature,spike production, growth chamber.INTRODUCTIONPerennial rye grass (Lolium perenne L.) is used mainly forpastures with increasing use for turf due to development ofacceptable turf types. The majority of the U.S. supply ofperennial ryegrass seed is produced in the WillametteValley of western Oregon of the U.S.A. Many perennialryegrass cultivars are entering the market, but very little isknown about environmental effects on growth or reproductivedevelopment. Knowledge of these requirementswould help improve the management practices, such asI. A contribution of the Crop Science Department, Oregon StateUniversity. Received for publication 23 May, 1983.2· Graduate student, Agriculture Experiment Station; Professor ofCrop Science, Department of Crop Science; and Associate Dean ofResearch, Oregon State University, Corvallis, Oregon, respectively.fertilizer timing and rate for improved seed production.Most investigators agree that there are three stages ofdevelopment in grass tillers: juvenile, floral induction, andfloral initiation stage, although the description of thesestages differs (Cooper, 1960; Evans, 1960; Calder, 1964;and Calder, 1966).The juvenile stage is the phase during which plants areinsensitive to environmental conditions that later promoteflowering, and its length depends on species (Calder, 1963;Calder, 1966). Induction is a physiological, chemical orhormonal change within the plant resulting from the fulfillmentof certain thermophotoperiodic requirements(Gardner and Loomis, 1953; Canode, Maun and Teare,1972; and Murray, Wilton and Powell, 1973). Conditionsgenerally required for floral induction in most cool-seasongrasses are a period of short photoperiod and cool temperatures.The requirement of perennial ryegrass for short dayand/or low temperature for flower induction has alreadybeen reported (Cooper, 1951; Cooper, 1957; Cooper 1960;and McCowan and Peterson, 1964). The response of plantsto inductive conditions differs with the species, cultivar,and tillers within the plant (McCown and Peterson, 1964;and Canode, Maun and Teare, 1972).Once the plant is fully induced, floral initiation can betriggered by the appropriate photoperiod and/or temperatureexposure (Cooper, 1951; Chilcote, 1966; Calder, 1964,1966; Canode, Maun and Teare, 1972; and Canode andPerkins, 1977). The elongation of the vegetative growingpoint is closely followed by clearly discernible morphologicalchanges which have been carefully documented forseveral grasses (Gardner and Loomis, 1953; Barnard, 1964;and Calder, 1966).Our research was designed to: 1) determine if juvenilityexists in certain perennial ryegrass cultivars; 2) identifyinductive requirements and floral initiation characteristicsof these cultivars; and 3) determine growth and developmentdifferences among cultivars in response to differentlow temperature and photoperiod treatments.MATERIALS AND METHODSFour forage and turf type cultivars of perennial ryegrasswere used, namely Linn, an early maturing cultivar, Pennfineand Acclaim, medium maturing cultivars, and Loretta,a late maturing cultivar. Two experiments were carried outin controlled environment chambers. The chambers wereequipped with automatic light and temperature controls.The photosynthetic photon flux density (PPFD) at the plantlevel was 350 uE sed- 1 m 2 •34

35Experiment OneSeeds of certified perennial ryegrass cultivars Linn,Pennfme, and Loretta were exposed to different coldtreatments in moist perlite in a dark cold room at 4-6 C,after free imbibition for 24 hours at 20 C (Salisbury, 1965).Once a week, air was circulated through the dishes toprevent accumulation of carbon dioxide with water added atthe same time if needed. Cold exposure treatments of 0, 1,2, 4, and 8 weeks were arranged so that all treatments werecompleted simultaneously and plants moved to a chamber at21 C constant temperature on the same date, 6 January,1978, for floral initiation treatments. Four seedlings weretransplanted into 10 x 10 em plastic pots filled withthoroughly mixed combination of sandy loan soil andgreenhouse peat. The pots were placed in the chamberaccording to a completely randomized design under 12- and20-hour photoperiods.Experiment twoAcclaim cultivar was included in this experiment insteadof the Linn cultivar because it is a better known turf typecultivar. Seeds were sown in 10 x 10 em plastic pots. Fourplants were grown in each pot in the greenhouse at aconstant 18 C and 14-hour photoperiod for different periodsof time, during the period from 19 April to 19 May, 1979.Plants were transferred to the coldroom at three differentstages (Table 1).RESULTS AND DISCUSSIONVegetative growthIn the first experiment, significant differences were foundamong cultivars and between photoperiods in number oftillers produced per plant (Figure 1). These differences are16-14-12-10--c::ca...... 8-Q)c.U'l...Q)6-~}Total Tillers(:}}}] Fertile TillersTable 1. Growth stage, age and description of plants at time ofexposure to inductive conditions.Age of plantsGrowth stage (weeks)I4nIII10DescriptionTwo-three leaf stage with2-3 tillers/plantSingle leaf stageImbibed seedsSowing dates were arranged so that all three stages werecompleted simultaneously and moved to the cold room at4-6 C and 8 hrs photoperiod at the same time. The plantswere exposed to cold treatments for durations of 0, 1, 2, 4,or 8 weeks, as in experiment one. After the end of the coldexposure, plants were preconditioned for a week in achamber at 15 C and 8-hour photoperiods and were thentransferred to the growth chamber. The chamber constanttemperature was 21 C and two photoperiods (12 and 20hours) were used.In both experiments, newly exerted heads were loggeddaily to record time to head exertion. When experimentswere discontinued, number of plants per pot, number oftillers per plant, total number of spikes appearing at thesame date, number of heads per plant, number of leaves atheading, and number of spikelets per spike were recorded.Dry weights were determined after the samples were ovendriedfor 48 hours at 60 C. The data from both experimentswere analyzed as a completely randomized design and themeans that were statistically significant were separated byDuncan's multiple range test.4-2-PennfineLinnLorettaFig. 1. Total and fertile tiller production for three ryegrasscultivars grown under long day conditions following exposureof the imbibed seed to low temperature ( 4-6 C).likely due to the genetic background of these cultivars.Loretta (late maturing) and Pennfine (medium maturing) areturf types, whereas, Linn is a forage type and earlymaturing. Higher tiller number was produced in the 20-hourphotoperiod in both experiments because of greater lightenergy in the 20-hour photoperiod which favors tillering. Inthe second experiment, significant differences were notedamong cultivars (Figure 2) and plant growth stages (Table2) in tiller number per plant. Pennfine again had the highesttiller number, followed by Acclaim and Loretta. Theobserved difference in tiller number per plant noted amonggrowth stages (Table 2) is understandable since the plantsexposed at growth stage I (4 weeks) which has the highesttiller number are four and three weeks older than growthstage Ill (0 week) and growth stage II (one week) plants,respectively. Higher numbers of tillers and larger size are

3614-12-L::;:;:;:::J} Total Tillersh::::::::::d Fertile Tillers2.0-D 20 Hour PhotoperiodIEJ12 Hour Photoperiod-c::00:::...Q)a.Ul...Q)i=10 -8~6t-41-21-0~-l::l:l:t:::::::Pennfine Acclaim Loretta-..c:01Q):s:>.,_ 1.0-0-c::00::PennfineLinnLorettaFig. 3. Effect of photoperiod and cultivar on dry matterproduction in perennial ryegrass cultivars exposed as imbibedseeds to low temperature conditions.Fig. 2. Total and fertile tiller production of three ryegrasscultivars exposed at the seedling stage to different lowtemperature 4-6 C durations and grown under differentphotoperiods.Table 2. Response of three turf type perennial ryegrass cultivarsexposed to low temperature at different ages and grownunder 12 and 20 hour photoperiods.Growth stageIITIllPlant age at timeof cold exposure(weeks)4I0Tillers*per plant11.4 a10.9 a7.6 b•Figures with different letters are significantly different at 5% level of probabilitythe prerequisites for good seed production.Dry Matter ProductionDays toHeading51.460.160.6Dry weight per plant is a function of number of tillers perplant and tiller weight which depends, in part, on number ofleaves per tiller, their expansion and duration of growth.The dry weight per plant was higher for the 20-hourphotoperiod than the 12-hour photoperiod following thesame trend of tiller number per plant. This was due to thehigher total energy in the 20-hour photoperiod.In experiment one, a variety x photoperiod interaction fordry weight per plant was significant due to variation amongcultivars ·within the 20-hour photoperiod (Figure 3). Thissmall but significant difference appears to be due to thereduced response of the Linn cultivar (forage type) to the20-hour photoperiod. The advanced plant growth stages(second experiment) resulted in significantly increased plantdry weight. Since tiller production for these stages isdifferent (Table 2), the higher tiller number for older plantsallowed greater leaf area, more light interception, thushigher dry matter production through the photosyntheticprocess. Cold treatments did not affect dry matter production,which is in contrast to findings by Salisbury (1965)who noted that low temperature exposure increased tilleringand dry weight per plant in Lolium. High dry matterproduction is also important in obtaining high seed yield if itis partitioned to economic yield (seed).Reproductive DevelopmentSeed production depends not only on growth and developmentof tillers (shoots), but these tillers must bereceptive and the environmental conditions necessary forinduction and initiation must be favorable for fertile tillerproduction. In this study, flowering response of plants,even when exposed to 8 weeks of low temperature treatment,was less than expected. The percentage of plantswhich succeeded in producing spikes under growth chamberconditions was quite low (1 0%) compared to that usuallyoccurring under field conditions. Since plants were startedfrom seeds and the species is cross pollinated, variabilitymay be the result of genetic variation within the plantpopulation. Fertile tiller production was poor in bothexperiments (Figures 1, 2). The percentage of fertile tillersproduced in experiment one was 2, 32 and 4 for Pennfine,Linn, and Loretta cultivars, respectively. Linn (the foragetype) was the only cultivar where spike production increasedas the duration of cold treatment increased (Table3). This is in agreement with Murray, et al., (1973), whofound that the number of panicles per plant increased anddays to heading decreased with increasing exposure to low

37Table 3. Response of Linn perennial ryegrass after differentcold duration exposures in the imbibed seed stage.Cold Total Fertile Fertile Leaves to Spikelet/Tillers Tillers Tiller (%) Heading Spike84210190162157200172826443261743402813102.652.202.603.773.6210.09.18.214.812.1temperature. Evans (1960), Canode, Maun and Teare (1972)and Canode and Perkins (1977) drew similar conclusions.However, in the other cultivars, e.g. Pennfine and Loretta(turf types), the "no cold treatment" was sometimes aseffective as the eight-week cold treatment. This may suggestinduction due to short days which can be independentof low temperature induction (Cooper, 1960). Cultivarsdiffered in spike production despite the low response tofloral induction conditions (Figure 4). Linn (forage type)-c::cc::....Q)c.rnQ).>

38imbibed seeds. This suggests that maintenance of adequatetiller number and size is a prerequisite for high seed yield.The failure of many tillers to produce spikes in thesestudies showed some of the problems of greenhouse research.Results may not relate to field conditions because ofthe artificial light and temperature regime present in thegrowth chamber. This failure to produce spikes may also bedue to:1. An obligate juvenile stage for these cultivars so theywere insensitive to inductive low temperatures and/orshort days at time of exposure.2. Insufficient duration of exposure for floral induction.These cultivars may have requirements for low temperatureexposure which exceeds the 8 weeks used.The results stress the fact that different cultivars ofperennial ryegrass have different requirements for vegetativeand reproductive growth and development and specificmanagement systems for cultivars needs to be better identified.REFERENCES1. Barnard, C. 1964. Form and structure. In Grasses andGrasslands. by C. Barnard. (ed.). p. 47-72. McMillan,London.2. Calder, D. M. 1963. Environmental control of flowering inDactylis glomerata L. Nature 197:882-883.3. Calder, D. M. 1964. Stage development and flowering inDactylis glomerata. Herbage Abstracts 32:1525.4. Calder, D. M. 1966. Inflorescence induction and initiation inthe grarninae. In The Growth of Cereals and Grasses, Proc.of the 12th Easter School of Agricultural Science, Universityof Nottingham. F. L. Milthorpe and J. D. Ivens. (eds.). p.59-73. Butterworths, London.5. Canode, C. L., M. A. Maun, and I. D. Teare. 1972.Initiation of inflorescences in cool season perennial grasses.Crop Sci. 12:19-22.6. Canode, C. L. and Merlin Perkins. 1977. Floral induction inKentucky bluegrass cultivars. Crop. Sci. 17:278-282.7. Chilcote, D. 0. 1966. Environmental control of flowering inselected genotypes of orchardgrass. (Dactylis glomerata).Ph.D. Thesis. Purdue University, Lafayette, Indiana.8. Cooper, J. P. 1951. Studies on growth and development inLoliwn. II. Pattern of bud development of the shoot apex andits ecological significance. J. of Ecol. 39:228-270.9. Cooper, J. P. 1957. Developmental analysis of populationsin the cereals and herbage grasses. II. Response to lowtemperature vernalization. J. of Agric. Sci. 49:361-383.10. Cooper, J.P. 1960. Short day and low temperature inductionin Lolium. Ann. of Bot. 24:232-236.11. Evans, L. T. 1960. The influence of temperature onflowering in species of Loliwn and Poa pratensis. J. ofAgric. Sci. 54:410-416.12. Gardner, F. P. and W. E. Loomis. 1953. Floral inductionand development in orchardgrass. Plant Phys. 28:201-217.13. Lindsey, K. E. and M. L. Peterson. 1962. High temperaturesuppression of flowering in Poa pratensis. Crop Sci. 2:71-74.14. McCown, R. L. and M. L. Peterson. 1964. Effects of lowtemperature and age of plant on flowering in Loliumperenne. Crop. Sci. 4:338-391.15. Murray, J. J., A. C. Wilton, and J. B. Powell. 1973. Floralinduction and development in Festuca rubra L. Differentialclonal response to environmental conditions. Crop Sci.13:645-648.16. Salisbury, J. H. 1965. Interaction in the growth and developmentof Loliwn. I. Some effects of vernalization on growthand development. Aust. J. of Agr. Res. 16:903-913.17. Templeton, W. C., G. 0. Mott, andR. J. Bula. 1961. Someeffects of temperature and light on growth and flowering oftall fescue Festuca arundinacea Schreb. II. Floral development.Crop Sci. 4:283-286.

Chemical Dwarfing Effects on Seed Yield of Tall Fescue (Festuca arundinacea)cv. Fawn, Fine Fescue (Festuca rubra) cv. Cascade,and Kentucky Bluegrass (Poa pratensis) cv. Newport 1D. W. Albeke, D. 0. CWlcote and H. W. Youngberg 2ABSTRACTNew cultivars are frequently not superior seed producers.Severe lodging resulting from nitrogen applied to maximizeseed production often reduces the seed yield potential.An experimental growth retardant, Imperial Chemical IndustriesPP 333, was applied to three species of cool seasongrasses grown for seed under western Oregon conditions toreduce lodging and enhance dry matter partitioning to the seed.Species studied were: tall fescue (Festuca arundinacea) cv.Fawn, fine fescue (Festuca rubra) cv. Cascade, and Kentuckybluegrass (Poa pratensis) cv. Newport.The chemical treatment reduced internode elongation inculms of species tested. Rates applied to bluegrass were excessivein regards to vegetative growth but did not change seedyield. Application to fine fescue reduced fertile tiller numberand final seed yield. Treatment caused an increase in tall fescueseed yield, an increase in potential yield (in some cases) andreduction of lodging.Additional key words: lodging, growth retardant, harvest index,seed yield components.INTRODUCTIONGrass breeders select primarily for characteristics whichcontribute to forage or turf yield and/or quality. Whendeveloping new cultivars, seed yield is often a secondaryconsideration. Thus, cultivars released may be below seedproduction potential for the species (Griffiths, et al, 1980).In addition, lodging often occurs when nitrogen is appliedto maximize seed yield. Lodging causes poor pollen dispersaland light interception in the crop canopy (Hebblethwaite,1977). The resulting closed canopy may also facilitatepathogen development resulting in high tiller mortality(Griffiths, 1967). Plant breeders working with cereal cropshave had a great deal of success in incorporating semi-dwarfgenes into cereal cultivars and thereby improving productionunder high nitrogen fertility by preventing lodging(Borlaug, 1968). Use of a growth retardant chemical mayoffer a means of reducing plant height and strengtheningI. A contribution of the Crop Science Department, Oregon StateUniversity. Received for publication 31 May, 1983.2· Formerly Graduate Assistant, Professor Crop Physiology, andProfessor of Crop Science, respectively, Department of CropScience, Oregon State University, Corvallis, Oregon 97331, USA.stems in grass crops grown for seed, thus providing amanagement tool for maximizing grass seed production. Anew plant growth retardant (2RS,3RS)-1-(4-chlorophenyl)-4,4-dimethy 1-2-1 ,2,3-triazol-1-yl) pentan-3-01, coded PP333 (trade name Parlay) is available from Imperial ChemicalIndustries which has shown promise for lodging controland yield enhancement in perennial ryegrass (Hebblethwaite,et al, 1981). This compound is soil active and inhibitsgibberellin biosynthesis in the plant.The objectives of this preliminary experiment were todetermine the effect of PP 333, when applied to three coolseason grass species, other than ryegrass, grown for seed inthe Willamette Valley of Oregon. The species selected forstudy were tall fescue (Festuca arundinacea) cv. Fawn, finefescue (Festuca rubra) cv. Cascade, and Kentucky bluegrass(Poa pratensis) cv. Newport. Effect of the dwarfingcompound on certain morphological characteristics, degreeof lodging as well as potential and actual seed yieldcharacteristics was measured.MATERIALS AND METHODSApplication of PP 333 (a soil active compound) wasmade in the spring of 1980 to the three species which wereproducing their sixth seed crop. Applications at rates of0. 75 and 1.5 kg ai ha- 1 were applied on 18 March and 7April at approximately spikelet initiation and floret initiationgrowth stage, respectively.The material was applied with a conventianal bicyclesprayer used for experimental herbicide applications. Linepressure was 2. 7 6 x 1 OS P A, which provided a spray volumeof 280 1 ha- 1 of solution. More than 15 em of rain fell within48 hours of application on both dates. The experiment wasdesigned as a completely randomized block with 3 replications.Plot size was 0.65 m 2 •Morphological characteristics were measured on 20 fullyemerged fertile tillers selected at random of tall fescue andfine fescue prior to anthesis. Measurements were taken ofpanicle length, culm length, distance between the bottomspikelet and first upper node, and the distance between thefirst and second nodes from the base of the panicle. Sincerates selected for bluegrass were excessive only the finalyield parameters were measured.After anthesis, four 9.29 cm 2 samples per plot were takento determine the potential yield parameters. The fertiletillers were counted and spikelets per panicle and florets perspikelet were determined from a 10 tiller subsample. Floretsper spikelet were determined by randomly selecting 1039

40Table 1. Morphological characteristics of Fawn tall fescue fertile tillers as affected by PP 333 treatments.*Length between Length betweenpp 333Total leaf Panicle Culm 1st and 2nd bottom spikeletRate Leaf number area length length nodes and 1st node(kg ha- 1 )(cm 2 ) - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - -(em)- - - - - - - - - - - - - - - - - - - - - - - - - - - - - -03.325.6 17.6 108.3 25.1 55.80.75 3.220.8 18.9 100.0 18.6 59.51.503.221.6 16.6 84.9 13.6 53.5L.S.D .. osNS3.7 1.4 7.3 2.8 4.6*Mean across two application dates x three replications.spikelets from the panicle and counting the number offlorets. Potential fertile florets per unit area was determinedby the function: Fertile tillers per meter square x spikeletsper panicle x florets per spikelet.Seed was cut from a .7 x .55 meter area at moisturecontents of 43% for tall fescue, 25% for fine fescue, and28% for Kentucky bluegrass and air dried for 15 days. Seedand straw were placed in heated forced air drier for 24 hoursat SOC before threshing. Weight of above-ground biomasswas determined. Seed was hand threshed, hand cleaned,weighed and mean seed weight determined by weighing 400seeds. Harvest Index was calculated by dividing clean seedweight by the total above-ground biomass. Lodging scoreswere obtained by visual inspection of the plots with a scoreof 1 indicating an upright canopy, and 10 a flat canopy.Intermediate scores were based on the degree of flatness ofthe canopy as well as the lodged area. Sixty kg ai ha- 1nitrogen in the form of ammonium sulfate was applied tothe plots on 27 March. On 18 April, an additional 60 kg aiha- 1 of N was applied as ammonium nitrate. This fertilityprogram is consistent with commercial management practicesfor seed crops in the Willamette Valley.An analysis of variance was performed on all parametersmeasured. A significant F value at p = .05 was used todetermine significant differences among treatments. Theleast significant difference (L.S.D.) at p = .05 was used tocompare means.RESULTS AND DISCUSSION1. Tall FescueThe application of PP 333 had no effect on the number ofleaves of fully emerged tall fescue fertile tillers (Table 1).However, total leaf blade area was reduced approximately20% by both rates of the chemical, thus reducing photosyntheticarea. Dates of application, however, did not varywith respect to leaf number or area.The culm length was reduced at both rates, but thepanicle length was not changed. Culm length was reducedby inhibition of internode elongation, primarily at the firstand second nodes (Table 1). The chemical apparentlyaffected intercalary meristem activity or stem cell elongationwithout an effect on differentiation of the inflorescence.Fertile tiller number per meter square at anthesis wasquite variable, consequently no statistically significant differencescould be attributed to application of PP 333.However, at the .75 kg ha- 1 rate, 19% more fertile tillerswere observed when compared to the high rate or thecontrol (Table 2).Table 2. Potential yield characteristics of Fawn tall fescuefertile tillers as affected by PP 333 treatments.*Potentialfertileflorets(10 3 x m- 2 }pp 333 FertileRate Tillers m- 2 Spikelets/ Florets/Qanicle SQikelet(kg ha- 1 )0 337 33.9 6.50.75 417 40.8 6.71.50 341 35.2 6.7L.S.D .. 05NS 5.8 NS*Mean across two application dates x three replications.74.1114.080.111.0The .75 kg ha- 1 rate also resulted in more spikelets perpanicle than the high rate or the control. However, dates ofapplication did not differ in spikelets per panicle. Thenumber of florets per spikelet was unaffected by PP 333treatment (data not shown). Potential yield (the number ofpotential fertile florets per unit area present after anthesis)was increased by 54% at the low rate of PP 333. Date ofapplication had no effect on potential fertile floret number.The .75 and 1.5 kg ha- 1 rate increased seed yieldsignificantly over the control in this experiment (Table 3).Table 3. Actual seed yield charcteristics of Fawn tall fescue asaffected by PP 333 treatments.*pp 333 Seed Seed Seeds/ Harvesttreatment yield weight SQikelet Index Germination(kg ha- 1 ) (gm-2) (mg) (%)0 78.9 3.16 2.18 0.079 910.75 (early) 119.8 3.15 2.18 0.124 920.75 (late) 129.8 3.12 2.51 0.136 921.50 (early) 137.6 3.05 3.60 0.149 841.50 (late) 101.0 3.02 2.93 0.125 87L.S.D .. 0511.3 NS 0.42 0.04 NS*Mean of three replicationsClean seed yield increased 58% and 44% at the low andhigh rate respectively. A rate by date interaction for seedyield showed that 1.5 kg ha- 1 applied in March is superior to

41an April application. At 0.75 kg ha- 1 , however, both Marchand April applications showed similar seed yields. Latespring applications (floret initiation stage) at high ratesreduce seed yield potential in tall fescue.The number of seeds per spikelet was increased significantlyby the 1.5 kg ha- 1 rate of PP 333 when compared toeither the low rate or control. Although the high rates of PP333 developed the same potential yield as the control, thelarger number of seeds per spikelet at harvest resulted froma greater percentage of florets actually developing seeds.The treatment at the low rate, had a greater potential yield (alarger number of potential fertile florets per unit area), butwas not able to realize its potential since seeds per spikeletwere not greater than the control.Mean seed weight was not affected by PP 333 treatmentseven though seed number increased. This is a positiveresponse since an increase in seed number per unit area isusually accompanied by a decrease in seed weight due toyield component compensation.The efficiency of the seed production process as measuredby Harvest Index (HI) is improved significantly by PP 333(Table 3). The low and high rate resulted in a 64 and 73%increase in HI, respectively.The germination percentage of tall fescue, (Table 3) seedwas not statistically different from the control.Under normal seed production conditions, tall fescuedoes not lodge as early or as severely as ryegrass or finefescue. Some lodging is common at high fertility ratesand/or during wet spring weather. It was noted in thisexperiment that the control plots began to lodge the secondweek of June (after anthesis), while treated plots did notlodge (Figure 1).Table 4. Morphological characteristics of Cascade fine fescuefertile tillers as affected by PP 333 treatments.* (1980)Length between Length betweenpp 333 Panicle Culm 1st and 2nd bottom spikeletRate length length nodes artd 1st node(kg ha- 1 ) -----------------(em)------------------011.2 74.7 9.7 48.50.75 10.5 65.1 8.0 45.91.50 10.0 51.8 6.8 36.9L.S.D·.os NS 7.6 2.3 7.3*Mean across two dates of application x three replications.and between nodes 1 and 2 was reduced with the growthretardant treatment at the high application rate. However, asin tall fescue, the panicle length was not affected by growthretardant treatment.The date of application altered potential yield characteristics.The late application reduced florets per metersquare as a result of a significant decrease in fertile tillernumber (Table 5).Table S. Potential yield parameters of Cascade fine fescue asaffected by PP 333 treatments* (1980)PP 333 Fertile Florets/ Potential fertileTreatments tillers m- 2 spikelet florets (1 ()3 x m- 2 )(kg ha- 1 )00.75 (early)0.75 (late)1.50 (early)1.50 (late)L.S.D·.os103310717481222844240*Mean of three replications.7.07.48.46.87.70.321522716329419824108Q)0 6()(f).!: "'01 ·!!!..., .,4

42Mean seed weight and germination percentage were notaffected by PP 333 treatments.PP 333 treatment delayed and, in some cases, preventedlodging (Figure 2). The lack of seed yield enhancement108~0 6(.)enOL-~---7~--~----~--~~--~~--~----~16 20 31 5 13 20 27 5May June JulyFig. 2. The effect of PP 333 on lodging in fine fescue (1980).with the growth retardant treatment in this experiment maybe related to reduced lodging severity as well as delay inlodging until after anthesis in the untreated plots whichwould reduce the effects of lodging on tiller mortality andseed filling.3. Kentucky BluegrassPP 333 treatment at 0.75 and 1.5 kg ha- 1 caused extremegrowth inhibition. Final seed yield data were collected(Table 7). Surprisingly, when final seed yield results wereTable 7. Yield and seed characteristics of Kentucky bluegrassas affected by PP 333 treatments* (1980)PP 333 Seed Seed HarvestRate Yield Weight Index Germination(kg ha- 1 ) (g m- 2 ) (mg)0 108.4 0.410.75 99.9 0.361.50 107.9 0.33L.S.D .. 05NS 0.050.1460.1930.2600.042*Mean across two application dates x three replications.(%)938992NScalculated, treated plots were not significantly differentfrom control plots, in spite of drastic growth reduction.Mean seed weight was lower as a result of PP 333treatment. This is evidence of excessive dose and reductionsin photosynthetic area affecting seed filling. Substantialreduction of above-ground biomass resulted in an increasein HI in the treated plots.Germination percentage was not affected by PP 333 atthese rates.Lower rates should be used in further tests of thischemical on Kentucky bluegrass.CONCLUSIONSThe growth retardant chemical had significantly biologicalactivity on all three species studied. A decrease inthe internode elongation resulting in shorter culms wasnoted in all three species.In tall fescue, . 75 kg ha- 1 of PP 333 applied in the springat either spikelet or floret initiation stage resulted in moreflorets being differentiated than in the control or the 1.5 kgha- 1 treatment. However, the 1.5 kg ha- 1 rate resulted in agreater percentage of florets actually producing seed thanthe lower rate or control. This difference in rate effect isdifficult to explain since lodging scores were very similar.Stem length differences were apparent and perhaps drymatter partitioning from stem to seed is a factor in thedifferences in seed site utilization. Prevention of lodging inPP 333 treated plots contributed to increased seed yield ofsubstantial magnitude even though lodging was not severein the control and occurred after anthesis. Greater seed yieldper unit area in tall fescue apparently resulted from moreefficient partitioning of dry matter to economic yield assupported by higher HI.Applications of PP 333 to fine fescue at these rates andgrowth stages reduced seed yield. This was the result of areduction in fertile tiller number and an increase in floretabortion. Lower rates and/or application at different growthstages should be examined.Bluegrass showed greater sensitivity to PP 333 treatmentand lower rates need to be studied. However, it is interestingto note that seed yield was not depressed below thecontrol at the rates tested and seed germination was also noteffected.ACKNOWLEDGMENTSWe would like to acknowledge the support of Imperial ChemicalIndustries in the conduct of this work. Thanks are also extended toC. C. Moon, W. Young, and D. Ehrensing for their assistance inthe collection and summarization of data.REFERENCESl. Borlaug, N. E. 1968. Wheat breeding and its impact on worldfood supply. p. 1-36. In: K. W. Finlay and K. W. Sheperd(eds.), 3rd Int. Wheat Genet. Symp. Butterworths, London.2. Griffiths, D. J. 1967. Review of section 3 - herbage plantbreeding and seed production. J. British Grassland Society22:17-21.3. Griffiths, D. J. , J. Lewis, and E. W. Bean. 1980. Problems ofbreeding for seed production in grasses. p. 37-50. In: SeedProduction. P. D. Hebblethwaite (ed.). Butterworths & Co.,London.4. Hebblethwaite, P. D. 1977. Irrigation and nitrogen studies inS .23 ryegrass grown for seed. 1. Growth, development, seedyield components, and seed yield. J. Agric. Sci., Camb.88:605-614.5. Hebblethwaite, P. D., J. G. Hampton, and J. S. McLaren.1981. The chemical control of growth, development and yieldof Lolium perenne grown for seed. Proceedings of 33rd EasterSchool in Agricultural Science. University of Nottingham. byJ. S. McLaren (ed.). Butterworths & Co., London.

Reproductive Growth and Development in Selected Kentucky BluegrassCultivars Under Different Environmental Conditions 1J. Kevin Turner, D. 0. Chilcote, and R. V. Frakes 2ABSTRACTKentucky bluegrass (Poa protensis L.) cultivars are known todiffer quantitatively in their photothermal inductive requirements.The extent to which these requirements are met influencesthe potential for subsequent panicle exertion. A betterunderstanding of the inductive requirements as they varyamong cultivars would aid in the selection and IDaii_agement ofcultivars for maximum seed yields. This study was designed toevaluate the response of three Kentucky bluegrass cultivars,Bristol, Victa, and Vantage, to differing photothermal inductivetreatments in a controlled environment, a mild, wet fieldenvironment at Gervais, Oregon, and a cold, dry environmentat Madras, Oregon.Sod plugs pulled from breeder's nurseries were used as thesource of plant material for this study. Plants of Bristol, Victa,and Vantage were exposed to ambient inductive conditions forperiods ranging from 23 to 209 days. Quantitative measurementson the number of panicles produced and the rate ofpanicle exertion were made after transfer from inductiveconditions to controlled environment chambers programmed topromote floral initiation.A significant interaction was observed between cultivars andlocations for the rate of panicle exertion. The environment atGervais was more favorable for Bristol, relative to Victa andVantage, than that of Madras in terms of the number ofpanicles exerted per plant and the rate of panicle exertion.Controlled environment studies showed that Victa washighly dependent on early-emerging tillers for panicle production.In contrast, panicle production in Bristol and Vantagewas equally distributed among tillers of all ages. The seed yieldsof cultivars such as Bristol and Vantage, which can exertpanicles from late formed tillers may benefit by placement intoproduction areas which promote fall and winter growth such asGervais.Additional key words: Kentucky bluegrass, Poa protensis L.,tiller, floral induction, initiation, juvenile stage.INTRODUCTIONThe seed yields of Kentucky bluegrass (Poa pratensis L.)cultivars vary among commercial production areas. Theseed yield of Kentucky bluegrass is greatly dependent onI. A contribution of the Crop Science Department, Oregon StateUniversity. Received for publication May 23, 1983.2· Research agronomist, the 0. M. Scott & Sons Co. Formergraduate student, Crop Science Dept., Professor of Crop Physiology,Crop Science Dept., and Associate Dean of Research,Oregon State University, Corvallis, Oregon, respectively.the number of tillers which are transformed from a vegetativestate to a potentially reproductive state (Langer andLambert, 1959). This transformation requires exposure toboth short days and cool temperatures (Peterson and Loomis,1949) and is termed floral induction.In many grass species, induction can occur only aftercompletion of a juvenile state (Calder, 1967) during whichthe plant is insensitive to photoperiods which would otherwisebe inductive. The juvenile stage varies greatly induration among species, but little is known concerning thefactors controlling its length (Bean, 1970; Gardner andLoomis, 1953; and, Calder, 1967). Working with tallfescue (F estuca arundinacea Schreb.) and meadow fescue(Festuca elatior L.), Bean (1970) observed that the youngerthe plant, the longer the inductive period required forsubsequent floral differentiation. Gardner and Loomis(1953) concluded that in orchardgrass, each tiller has itsown juvenile stage. Calder (1967), however, was not surewhether juvenility was a property of the individual tiller orthe plant as a whole.When tillers of meadow fescue, orchardgrass (Langerand Lambert, 1959) and Lolium perenne L. (Spiertz andEllen, 1977) were labelled at intervals over time, theearliest formed tillers contributed the largest proportion ofseed heads at harvest.Both the photothermal inductive requirement and thesubsequent response to lengthening photoperiods and warmtemperatures which result in floral initiation are polygenicallycontrolled in cool season grasses (Cooper, 1954).Lindsey and Peterson (1964) reported that in Kentuckybluegrass only early-emerging tillers are exposed to asufficiently long cold period to assure subsequent reproductivedevelopment. Canode, Maun, and Teare (1972)observed differences in the inductive exposure durationrequirements among four Kentucky bluegrass cultivars.Similar results have been reported in other grass species aswell (Bean, 1970; Chilcote, 1961; and, Murray, Wilton,and Powell, 1973).In this study, three Kentucky bluegrass cultivars wereexamined for differences in floral inductive requirements.Controlled environment chambers were used to determinethe relative contribution of early and late formed tillers topanicle exertion. The extent to which the winter fieldconditions in two contrasting field environments fulfilledthe induction requirements of the three cultivars wasexamined.MATERIALS AND METHODSOn 1 September 1977, 10 em diameter sod cores of'Bristol', 'Victa', and 'Vantage' Kentucky bluegrass were43

44removed from breeder's nurseries at Gervais, Oregon. 1 Thesod cores were placed in 15 em plastic pots filled with asandy loam soil mix. All pots were watered twice weeklywith tap water and le.ft in the field at Gervais, Oregon, tofacilitate vegetative growth. These sod cores were subsequentlyused for the field experiments at Gervais andMadras, Oregon, and the controlled environment experiment.Twelve of the cores were divided into single primarytillers on 25 September and placed in flats in a greenhouse.On 5 October, tillers were selected for uniformity andtransplanted into 10 em diameter plastic pots. Twenty-onepots of each cultivar were transferred on 19 October fromthe greenhouse to a controlled environment chamber used toprovide inductive conditions. The induction chamber wasprogrammed for a 16/6 C day/night temperature regime, an8-hour photoperiod, and a photosynthetic photon fluxdensity (PPFD) of 280 p, E m 2 s- 1 provided by a combinationof fluorescent tubes and 40-watt incandescent bulbs.The primary tillers and tillers which had emerged from thesoil between 25 September and 19 October were labelledwith colored loops of wire. Tillers which emerged after thisdate, during the inductive period, were labelled on 14November, 14 December and 14 January, thus designatingfour tiller age groups. A completely randomized designwith three replications per treatment was used and the datawere examined with analysis of variance techniques.On each of seven dates, three pots of each cultivar weretransferred from the inductive chamber to a controlledenvironment chamber programmed to stabilize the inducedstate (Murray et al., 1973). This chamber was set for aconstant 10 C and a photoperiod of 11 hours. After oneweek, the pots were placed in a third controlled environmentchamber programmed to promote floral initiation andwere subsequently observed for the number of paniclesexerted per tiller age group. Initiation conditions were18/13 C day/night temperature with a 16-hour photoperiod.The PPFD for the stabilizaton period and the initiationperiod was 300 p, E m 2 s- 1 provided by a combination offlorescent tubes and 40-watt incandescent bulbs. All chambersemployed sub-irrigation systems (Stanwood, Phillips,and Chilcote, 1974) which watered the plants once daily.Each pot was fertilized on 19 October with 30 kg ha- 1 eachof N, Pz0 5and ~0. Each pot was additionally fertilizedwith 25 kg ha- 1 N as it was transferred to the initiationcontrolled environment chamber.On 1 October, four sod cores of each cultivar were buriedto ground level in randomized complete block designs withfour replications for each of the ten induction exposureduration treatments at both field locations, Madras andGervais. Four pots of each cultivar were removed from thefield inductive environments at both locations on each of tendates and placed in the stabilization controlled environmentchamber (Lindsey and Peterson, 1964). After one week inthe stabilization chamber, the pots were transferred to thecontrolled environment chamber programmed to promote1Breeder's nurseries were located at the 0. M. Scott and SonsCompany Research Station at Gervais, Oregon.floral initiation and were observed for the total number ofpanicles exerted per pot and for the number of days requiredfor exertion of the first panicle of each plant. Exertion datewas recorded when a spikelet was visible beyond thesubtending leaf sheath.Plants exposed to the field inductive environments werefertilized with 40 kg ha- 1 each of N, P 20 5, and ~0 on 8October, 7 November, and 18 December. Each pot wasadditionally fertilized with 37 kg ha- 1 N when it was transferredjo the initiatipn chamber.RESULTS AND DISCUSSIONGervais has a Mediterranean climate dominated by themoderating influence of the Pacific Ocean and receives over100 em of annual precipitation with a winter distribution.One hundred and sixty kilometers east of Gervais is Madraswhich has a continental climate and receives about 25 emannual precipitation. Radiometry data obtained from theU.S. Weather Bureau at Portland, Oregon, and data obtainedfrom the USDA Experiment Station at Redmond,Oregon, were used as estimates of the daily light energyreceived at Gervais and Redmond, respectively.The light energy levels, mean daily minimum temperaturesand mean diurnal temperature ranges differed significantlybetween Gervais and Madras from 1 October 1977to 10 March 1978. The Madras location received a greateramount of sunlight over the course of the inductive seasonand also had lower minimum temperatures with a greatermean diurnal range than Gervais. The mean daily maximumtemperatures were not significantly different between thetwo locations. The mean daily maximum and minimumtemperaures recorded during the ten induction exposureduration treatments at Gervais and Madras is given in Table 1.Table 1. Mean maximum and minimum temperatures recordedduring induction treatments at Gervais and Madras.Induction treatment Mean maximum Mean minimumDuration temperature {q temperature {qBegin End Gervais Madras Gervais Madrasl Oct. 23 Oct. 16.7 17.8 5.0 1.124 Oct. 6 Nov. 11.1 12.2 4.4 0.67 Nov. 20 Nov. 7.8 10.0 1.1 -3.321 Nov. 4 Dec. 8.9 8.9 4.4 -0.65 Dec. 18 Dec. 8.9 8.3 3.9 -1.119 Dec. 31 Dec. 5.6 -0.6 1.1 -5.61 Jan. 14 Jan. 5.6 2.2 1.7 -3.915 Jan. 3 Feb. 6.7 6.1 2.2 -1.74 Feb. 24 Feb. 8.9 7.8 2.8 -1.125 Feb. 10 Mar. 10.0 7.8 2.8 -7.8At least 51 days of the fall-winter field environmentexposure were required before panicles were exerted by anycultivar in either area (Table 2). Once this minimum orthreshold level was achieved, Bristol, Victa and Vantage allproduced more panicles per pot as the field exposureduration was lengthened, indicating the quantitative natureof induction. This supports results found by other workers(Canode et al., 1972).

45Table 2. Mean number of panicles exerted per pot of Bristol,Victa, and Vantage after different field inductive exposuredurations at Madras and Gervais.InductiveExposure(Days)3MadrasBristolVi etaVantageGervaisBristolVi etaVantage23 37 51 65 79 92 106 126 147 1610 0 2.0 4.8 17.017.020.023.028.033.80 0 2.5 5.0 16.523.530.541.545.248.20 0 0 1.2 5.0 10.2 8.5 16.217.519.20 0 0 1.2 11.221.231.031.231.241.80 0 1.0 7.5 12.517.526.237.533.040.50 0 0 1.0 6.0 10.2 10.8 15.0 15.5 8.0LSD 05for differences among cultivars within columns is 5.76 forMadras and 6.23 for Gervais.5040'0tQ) 30>

4660c:: 500'tQ))(wQ) 40~c::&.,g 30II),.,00200 40Days of:.··· .......') . , .........80 120Inductive Exposure- MadrasBristol.V1cta..... --VantageFig. 4. Average number of days required for panicle exertionof Bristol, Victa, and Vantage Kentucky bluegrass aftertransfer to initiation controlled environment chamber fromambient inductive exposure treatments of varying durationsat Madras.60c:: 50:E~wQ)40~c::0 'Victa'·.0..',g'30'II),.,'c'0 ' ,_ --20Vantage1600 40 80 120 160Days of Inductive Exposure-GervaisFig. S. Average number of days required for panicle exertionof Bristol, Victa, and Vantage Kentucky bluegrass aftertransfer to initiation controlled environment chamber fromambient inductive exposure treatments of varying durationsat Gervais.placement into production areas which promote fall andwinter tillering.There presently are dozens of Kentucky bluegrass cultivarsbeing harvested for seed in the Pacific Northwest. Itseems unlikely that all will respond similarly to the differentclimatic areas where they may be produced. Cultivars areknown to differ in their inductive requirements. Victa, forinstance, appears to benefit from the cold minimum temperaturesand high light energy levels occurring in theMadras area. Bristol, on the other hand, seems well adaptedto the more moderate temperatures and lower light levelsrecorded at the Gervais site. The degree to which theTable 3. Mean number and percentage of tillers producing11anicles 11er cultivar by tillers of four labelled age grou11s.Tiller labelDate Victa % Bristol % Vantage %19 Oct. 3.33 77 1.50 34 0.33 3314 Nov. 0.67 15 1.17 26 0.08 814 Dec . 0.17 4 0.50 11 0.42 4214 Jan. 0.17 4 1.25 28 0.17 17LSD 050.93 0.93 0.93environment in a seed production area satisfies the inductiverequirements of a cultivar, in relation to other cultivars, canbe determined by examining factors such as panicle numberper unit area and panicle exertion rate. An understanding ofspecific cultivar induction requirements will facilitate theselective placement of cultivars in areas that are bestadapted for seed production. This will aid in optimizingyields.REFERENCES1. Bean, E. W. 1970. Short day and low temperature control offloral induction in Festuca. Annals of Botany 34:57-66.2. Calder, L. T. 1967. Inflorescence induction and initiation inthe Gramineae. In: The growth of cereals and grasses. F. L.Milthorpe and J. D. Ivins (ed). Proceedings of the 12thEaster School in Agricultural Science, Nottingham, Universityof Nottingham, 1965. p. 59-73. Butterworths, London.3. Canode, C. L., M. Anwar Maun and I. D. Teare. 1972.Initiation of inflorescences in cool-season perennial grasses.Crop Sci. 12:19-22.4. Chilcote, D. 0. 1961. Environmental control of flowering inselected genotypes of orchardgrass (Dactylis glomerata).Ph.D. Thesis. Purdue University. Lafayette, Ind.5. Cooper, J. P. 1954. Studies on growth and development inLolium. IV. Genetic control of heading responses in localpopulations. J. of Ecology 42:521-556.6. Gardner, F. P. and W. E. Loomis. 1953. Floral inductionand development in orchard grass. Plant Physiol. 28:201-217.7. Langer, R. H. M. and D. A. Lambert. 1959. Ear bearingcapacity of tillers arising at different times in herbage grassesgrown for seed. J. of Brit. Grassld. Soc. 14:137-140.8. Lindsey, K. and M. L. Peterson. 1964. Floral induction anddevelopment in Poa pratensis L. Crop Sci. 4:540-544.9. Murray, J. J., A. C. Wilton and J. B. Powell. 1973. Floralinduction and development in F estuca rubra L. Differentialclonal response to environmental conditions. Crop Sci.13:645-648.10. Peterson, M. L. and W. E. Loomis. 1949. Effects ofphotoperiod and temperature on growth and flowering ofKentucky bluegrass. Plant Phys. 24:31-43.11. Spiertz, J. H. J. and J. Ellen. 1977. The effect of lightintensity on some morphological and physiological aspects ofthe crop perennial ryegrass (Lolium perenne var. Cropper)and its effect on seed production. Neth. J. Agric. Sci.20:232-246.12. Stanwood, P. C., J. C. Phillips and D. 0. Chilcote. 1974.Fully automatic subirrigation system for glasshouse andgrowth chamber use. Crop. Sci. 14:773-774.

Effects of Chemical Dwarfing Application Under Different Nitrogen Levelson Seed Yield of Fine Fescue (Festuca rubra) cv. Cascade 1D. W. Albeke, D. 0. Chilcote and H. W. Youngberg 2ABSTRACTApplication of spring nitrogen often causes lodging in finefescue well before anthesis. Early lodging results in poor pollendispersal, poor light interception and facilitates pathogen development.An experimental growth retardant, Imperial Chemical IndustriesPP 333, was applied to Festuca rubra cv. Cascadefertilized in the spring with several nitrogen rates to determineeffect on lodging factors related to seed yield.PP 333 reduced lodging at all nitrogen (N) rates in thisexperiment. The high rate (0.8 kg ai ha' 1 ) chemical treatmentreduced the culm length and both rates reduced the distancebetween the top and second node irrespective of N rate. Morefertile tillers were present at anthesis in plots treated with 0.8kg ai ha·1 at all rates of N application. The growth retardantapplication increased potential yield at the two higher N ratesand increased harvested seed yield at all N rates.Additional key words: lodging, growth retardant, harvest index,seed yield components.INTRODUCTIONApplication of nitrogen fertilizer in the spring is acommon practice in attempting to maximize seed yield(Hebblethwaite, 1980). High fertile tiller mortality andrelatively low number of seeds per spikelet in ryegrass hasbeen attributed to the fact that crops receiving nitrogen inamounts that maximize seed yield have a dense closedcanopy which can become severely lodged (Hebblethwaite,1977). Higher rates of nitrogen fertilization accentuate thelodging problem (Hebblethwaite and Ivins, 1977).A new experimental growth retardant, ICI's (ImperialChemical Industries) PP 333 [(2RS,2RS)-1(4-chlorophenyl)-4,4-dimethyl-2-1 ,2,-4-triazol-1-yl)pentan-3-01] has shownpromise for lodging control in grosses. PP 333 is a soilactive chemical which must be taken up by the root systemto exhibit growth retardant properties (personal communication,ICI). PP 333 appears to be more reliable and predictablein its response, as long as the chemical is present in theroot zone (Hebblethwaite and Burbidge, 1976).I. A contribution of the Crop Science Department, Oregon StateUniversity. Received for publication 1 June, 1983.2· Formerly Graduate Assistant, Professor Crop Physiology, andProfessor of Crop Science, respectively, Department of CropScience, Oregon State University, Corvallis, Oregon 97331, USA.In experiments conducted in Great Britain, applicationsof PP 333 on perennial ryegrass delayed lodging, increasedseed yields up to 50%, and substantially increased HarvestIndex (Hebblethwaite et al., 1981).The objective of this experiment was to study theinteraction of growth retardant application with springnitrogen rate as it affects seed yield of fine fescue, Festucarubra. In this research, the effects of PP 333 on lodging,plant morphological characteristics, potential seed yield andactual seed yield were examined under several springnitrogen application rates.MATERIALS AND METHODSThe experiment on a nine-year-old stand of fine fescuecv. Cascade was a factorial arrangement of treatments witha split plot design. Nitrogen rates were treated as main plotsand growth retardant, PP 333, treatments as subplots. Theexperiment was replicated five times. Plot size was 3. 7 x4.5 m. No autumn nitrogen was applied. In February, 1980,30 kg ai ha·' of nitrogen as amonium sulfate was applieduniformly to all plots. On 3 April, the spring nitrogen andthe growth retardant PP 333 treatments were applied. A 2%granular formulation PP 333 was applied at 0.4 and 0.8 kgai ha·' using a lawn spreader. Supplemental nitrogen fertilizerwas applied at 60, 90, and 120 kg ai ha· 1 N in the formof ammonium sulfate. With the 30 kg applied earlier, thisgave total nitrogen treatments of 90, 120, and 150 kg ha·'and were designated as N 90, N 120and N 150treatments,respectively.Morphological characteristics were measured on 20 fullyemerged fertile tillers prior to anthesis in order to quantifytreatment effect on panicle length, culm length, distancebetween the bottom spikelet and the first node, and thelength between the first and second nodes at the top of theplant.At completion of anthesis, four 929 cm 2 samples weretaken from each plot to determine the potential yield. Thefertile tillers were counted. Ten tillers were selected atrandom from each sample and spikelets per panicle andflorets per spikelet, selected at random, on 10 spikeletswere counted. Results were combined giving one yieldcomponent value per sample. Fertile florets per unit areawere determined from the function: fertile tillers per metersquare x spikelets per panicle x florets per spikelet = floretsper unit area.Seed was cut at 25% moisture content and air dried for 15days. Seed and straw were placed in a heated forced airdrier for 24 hours at 50 C before threshing. Weight ofabove-ground biomass was determined. Seed was threshed,47

48cleaned, and weighed. Mean seed weight was determinedby weighing 400 seeds. Harvest Index was calculated bydividing clean seed weight by weight of total above-groundbiomass.Lodging scores were obtained by visual inspection of theplots with a score of 1 indicating an upright canopy and 10 atotally flat canopy. Intermediate scores were based on thedegree of flatness of the canopy as well as the lodged area.For conciseness, the sampling date values were combinedand a mean lodging score is presented.Analyses of variance were performed on all parametersmeasured. A significant F value at p X .05 was used todetermine significant differences among treatments. Theleast significant difference (L.S.D.) at p X .05 was used tocompare means.RESULTS AND DISCUSSIONApplication of PP 333 at the rate of0.8 kg ai ha- 1 reducedthe culm length of developing fertile tillers at all nitrogenapplication rates (Table 1). The 0.4 kg ai ha- 1 rate did notTable 1. Morphological characteristics of Cascade fine fescue asaffected by PP 333 and nitrogen treatments (1980).Length Distance between:Culm Panicle lowest spikelet 1st and 2ndTreatment and 1st node nodes(kg ha- 1 ) - - - - - - - - - - - - - - - - (em) - - - - - - - - - - - - - - - -Nitrogen 1Low 71.1 9.7 46.4 14.8Medium 70.1 9.5 44.4 15.2High 73.2 10.2 45.3 15.9L.S.D .. o5 NS NS NS NSpp 333 20 75.9 9.8 46.1 17.80.4 72.9 9.9 46.3 15.90.8 65.7 9.6 44.9 12.2L.S.D .. 053.7 NS NS 1.91Across three growth retardant treatments x five replications.2Across three nitrogen treatments x five replications.Table 2. Potential yield characteristics of Cascade fine fescueas affected by PP 333 and nitrogen treatments (1980}.Spikelets Florets Total potentialFertile per per fertileTreatment tillers m- 2 ~anicle s~ikelet florets (10 3 } m- 2(kg ha- 1 )Nitrogen 1Low 859 25.1 5.48 118.3Medium 963 23.9 5.68 128.1High 960 24.6 5.99 141.5L.S.D .. o5 NS NS 0.38 15.6pp 333 20 862 24.7 5.64 120.00.4 945 24.8 5.73 134.50.8 974 24,1 5.77 133.3L.S.D .. o5 101 NS NS 11.71Across three growth retardant treatments x five replications.2Across three nitrogen treatments x five replications.Table 3. Florets per spikelet of Cascade fine fescue as affectedby PP 333 and nitrogen treatments (1980}.PP 333Nitrogen levelLow Medium High5.75.45.4NS5.75.75.6NS5.66.16.30.4Table 4. Total florets (x 10 3 ) m- 2 of Cascade fine fescue asaffected by PP 333 and nitrogen treatments (1980}.PP 333Nitrogen levelLow Medium High116.8 122.1 121.2129.8 129.7 144.1108.3 132.6 159.1L.S.D .. o5 16.3 NS 16.3significantly reduce the total culm length, but did reduce thesecond internode length. The reduction of culm length isdue primarily to shortening the second internode in bothrates of PP 333. Nitrogen rate had no effect on culm orinternode length. Panicle length and the distance betweenthe lowest spikelet and the first node were not affected byeither nitrogen level or growth retardant application.Fertile tiller number m 2 after anthesis was increased withapplication of PP 333 at 0.8 kg ai ha- 1 (Table 2) over allnitrogen treatments. Spikelets per panicle were not affectedby either growth retardant or nitrogen treatments. However,the number of florets per spikelet was increased at N 150•There was significant nitrogen by growth regulator interactionfor this characteristic (Table 3). PP 333 applicationincreased the number of florets per spikelet only at N 150•Florets m 2 was affected both by growth retardant andnitrogen treatment (Table 4). Both rates of PP 333 increasedpotential yield significantly over the control (Table 2),while N 150increased potential yield significantly over N 90and N 120• There was significant growth retardant by nitrogeninteraction for potential yield. Application of 0.8 kgai ha- 1 PP 333 at N 90decreased potential yield whileapplications of the growth retardant at N 150increasedpotential yield significantly (Table 4).Clean seed yield was increased significantly by applicationof PP 333 only at 0.8 kg ai ha- 1 (Table 5). Optimumseed yield was obtruned with N 120; however, there was nonitrogen x PP 333 interaction for harvested seed yield.Mean seed weight was not affected by any treatment in thisexperiment.The efficiency of the seed production process as measuredby Harvest Index, (HI), was altered by both PP 333and nitrogen levels. Application of 0.4 kg ai ha- 1 and 0.8 kgai ha- 1 significantly increased the HI. N 150, however, lowered

49Table S. Actual yield components and harvest index of Cascadefine fescue as affected by PP 333 and nitrogen treatments(1980).Seed Seed Harvest Seeds/Treatment yield wt. index SQikelet(kg ha- 1 ) (g m-2) (mg)Nitrogen 1Low 72.7 1.24 0.150 2.34Medium 79.8 1.26 0.135 2.72High 74.4 1.27 0.126 2.48L.S.D .. o5 16.1 NS 0.021 0.23pp 333 20 63.2 1.27 0.118 2.330.4 71.1 1.27 0.142 2.500.8 78.9 1.24 0.151 2.72L.S.D .. o5 9.5 NS 0.013 0.161Across three growth retardant treatments x five replications.2 Across three nitrogen treatments x five replications.HI significantly when compared to N 90and N 120. There wasno interaction for HI and N level.Both growth retardant and nitrogen level inceased seedfilling. Applications of PP 333 at 0.4 and 0.8 kg ai ha- 1increased seeds per spikelet. N 120also increased seeds perspikelet. There was no nitrogen by PP 333 interaction forseeds per spikelet.Application of PP 333 reduced lodging significantly overthe control. Higher nitrogen levels caused more lodgingeven with retardant treatment (Table 6). However, therewas no significant PP 333 by nitrogen interaction.Table 6. Cascade fine fescue mean lodging scores 1 as affectedby PP 333 and nitrogen treatments (1980). 2Rate(kg ha- 1 ) LowN Medium N High N0 4.4 6.8 7.60.4 1.8 3.0 3.60.8 1.0 1.6 3.0Mean 2.4 3.8 4.7L.S.D .. 05nitrogen means -1.3L.S.D .. 05growth retardant means -1.111 = no lodging; 10 = completely lodged.2Across six observation dates.CONCLUSIONSMean6.32.81.9Nitrogen rates had no significant effect on plant mor-phology or tiller number. Florets per spikelet was the onlycharacteristic affected and by the high nitrogen rate only.However, only spring nitrogen rates were studied. Applicationof growth retardant at low nitrogen levels loweredpotential yield, while high nitrogen levels accompanied by agrowth retardant treatment showed an increased number offlorets per unit area. Fertile tiller number was increased atthe 0.8 kg ai ha- 1 rate of PP 333. Significant culm heightreduction was observed only at the high rate of PP 333application, although the distance between the first andsecond nodes was reduced by both rates. No effect on theuppermost internode was observed. Lodging scores suggestthat the stem is strengthened by the shortened lowerinternode,even though total culm height may not be reduced.Growth retardant application increased seed yield at allspring nitrogen levels in this experiment. The optimumnitrogen rate was 120 kg ai ha- 1 at all growth retardant rates.There was a greater number of seeds at the medium nitrogenrate when compared to the high or the low rates.Even though there was a growth retardant by nitrogeninteraction for potential seed yield, there was no interactionfor harvested seed yield. Higher rates of nitrogen andapplication at different plant growth stages in conjunctionwith growth retardant treatment need to be investigated.REFERENCES1. Hebblethwaite, P. D. 1977. Irrigation and nitrogen studies inS.23 ryegrass grown for seed. I. Growth, development, seedyield components and seed yield. J. Agric. Sci. Camb.88:605-614.2. Hebblethwaite, P. D., and J. D. Ivins. 1977. Nitrogenstudies in Lolium perenne grown for seed. I. Level ofapplication. J. British Grassland Society 32:195-204.3. Hebblethwaite, P. D. 1980. Some physiological aspects ofseed yield in Lolium perrene L. In: Seed Production, editedby P. D. Hebblethwaite. Butterworths. London.4. Hebblethwaite, P. D., J. G. Hampton, and J. S. McLaren.1981. The chemical control of growth, development andyield of Lolium perenne grown for seed. In: ChemicalManipulation of Crop Growth and Development. J. S.McLaren (ed.). Butterworths, London.ACKNOWLEDGEMENTSWe would like to acknowledge the support oflmperial ChemicalIndustries in the conduct of this work. Thanks are also exended toC. Moon, W. Young, and D. Ehrensing for their assistance in thecollection and summarization of data.

50IHSPRG SUBSCRIPTION MEMBERSAUSTRALIAThe Library BranchDepartment of Primary IndustriesBrisbane, QueenslandBRAZILCooperative Regional TriticolaCotrijuiIJUI- RSC.I.A.T. Tropical Pastures ProgrammePlanaltina, D.F.Agroceres S.A. Imp. Exp. Ind. & Com.Sao PauloEMBRAPA/SPSBBrasilia- D. F.DENMARKL. Daehnfeldt Ltd.OdenseA/S Dansk Fr¢handel Trifolium-SiloTaastrupDanish Plant Breeding Ltd.Store HeddingeDe samv. Danske Fr¢avlerforeningerRingstedD.L.F.RoskildeDorthealyst Fr¢aulM¢rk¢rHerf¢1ge Fr¢handel A/SHerf¢1geSN-FR0Nyk¢bing F.A/S Chr. Kehlets Fr¢forrentningBjertDanish Seeds CouncilCopenhagenFRANCElngenieur Regional FNAMSPont St. MarieFEDERAL REPUBLIC OF GERMANYDeutsche Saatveredelung Lippstadt­Bremen GmbH. zu LippstadtLippstadtDeutsche Saatveredelung Lippstadt­Bremen GmbH zu Lippstadt4 70 LippstadtITALYSisforaggeraBolognaNETHERLANDSZwaan & de Wiljes b.v.ScheemdaPlant Breeding StationZelder b.v.OttersumBarenbrug Holland b.v.OosterhootD. J. Vander Have b.v.Rilland-BathJ. Joordens' Zaadhandel b.v.KesselMommersteeg International b.v.VlijmenNEW ZEALANDWrightson NMA LimitedChristchurchNZ Grassland AssociationPalmerston NorthUNITED KINGDOMBayer U.K., Ltd.Agrochem. DivisionBury St. EdmundsSuffolkBritish Seed HousesBristolJohn Bryant (Romsey) Ltd.Romsey, HampshireCleanacres Ltd.Cheltenham, Glos.Dunns Seed & Grain Ltd.Corsham, Wilts.International Seed ProducersBury St. EdmundsSuffolkRegional Development and TechnicalServiceSandiacre, Nottingham

G. Gascoyne Seeds Ltd.WorcesterHoliwell Seed & Grain Co. Ltd.Ashford, KentHurst Crop Research & DevelopmentColchester, EssexI.C.I. Ltd.Bracknell, Berks.W.W. Johnson & Son Ltd.Boston, Lines.A. H. Marks & Co. Ltd.Bradford, West YorkshireMay & Baker Ltd.Ongar, EssexMommersteeg InternationalNr. WellingboroughNorth antsLincolnshire Seed Growers Assoc.Woodhall Spa.Lines.Nickersons Seed Specialists Ltd.Brims bySouth HumbersideJ. Picard & Co. Ltd.Uckfield, SussexRothwell Plant BreedersLincolnCharles Sharpe & Co. Ltd.Sleaford, Lines.Sinclair McGill (East) Ltd.Boston, Lines.Smith Brothers (Basingstoke) Ltd.BasingstokeHantsTwyford Seeds Ltd.BanburyOx on.W.H.D. Seed Growers Ltd.Staple GardenWinchesterWrightson NMA Ltd.LondonUNITED STATESNormark, Inc.Tangent, OregonNorth American Plant BreedersMission, KansasNorthrup King Co.Tangent, OregonNorthrup King Co.Othello, WashingtonNorthrup KingMinneapolis, MinnesotaO.M. Scott & Sons, Co.Gervais, OregonOregon Seed LeagueCentral Point, OregonPacific Seed Production Co.Albany, OregonPickseed West, Inc.Tangent, OregonRound Butte Seed GrowersCulver, OregonTurf-Seed, Inc.Hubbard, OregonWestern Seed Co.Albany, OregonWillamette Seed & Grain Co.Albany, OregonAmpac Seed Co.Tangent, OregonBlue Mountain Seeds, Inc.Imbler, OregonCenex Western CommoditiesSalem, OregonE. F. Burlingham & SonsForest Grove, OregonGreat Western Seed Co.Albany, OregonInternational Seeds, Inc.Halsey, OregonJenks-WhiteDiv. of Jacklin Seed Co.Tangent, OregonMeeker Seed & Grain Co.Amity, OregonMt. Emily Seeds, Inc.Imbler, OregonNORTHERN IRELANDGerminal Holdilngs Ltd.Banbridge, Co. DownField Botany Research Div.Dept. of Agriculturefor No. IrelandBelfast