and rotational grazing management

Differences among orchardgrass cultivars in response to hay and rotational grazing management y. A. Papadopoulosl, t.dJ'ffi_Bi,{#Jff'; h;;,$."ae', ...
Author: Angel Manning
2 downloads 0 Views 390KB Size
Differences among orchardgrass cultivars in response to hay and rotational grazing management y. A. Papadopoulosl,

t.dJ'ffi_Bi,{#Jff';

h;;,$."ae',

L. F. Larrammel'

lExperimental Farm, Agricutture and Agri-Food Canada, Nappan, Nova Scotia, Canadg BOL lCO;2Depa.rtment of sRese-arch Station, fEaz; and ptant science, Nova icoTii Agricunurit Coit"gi, iriio,' Nov'a S6otia, Canad-a Agriculture and Agri-Food CanZda, Kentville, Nioia Scotia, Canada B4N 1J5. Received 3 December 1993, accepted 19 December 1994.

Can. J. Plant Sci. Downloaded from www.nrcresearchpress.com by MICHIGAN STATE UNIV on 01/14/17 For personal use only.

a!

Y. A., Price, M. A., Hunter, G. M., McRae, K. B., Laflamme, L. F', Caldwell, c. D. and Fulton, N' R' 1995' Oiit"."oces among orchardgrass cultivars in response to hay and rotational grazing management. Can. J. Plant Sci' 75: 147-157. This study was conducted to compare orthardgrass (Dacrylis glomerata L.) cultivar performance under contrasting Papadopoulos,

orchardgrass managemenr sys@ms (hay vs. rotational grazing) during the first two production years. The trial consisted ofeight each replication perimeter of The replications. with three design, complete-block randomized cultivars broadcast seeded in a was fenced. and the area was divided into two sictions. The first section was managed as a simulated hay production system' (DM) yield was The second section was managed by simulating a rotational grazingsystem using crossbred ewes' Dry maBer four cultivars prior on (LAI) measured were (TD) area index leaf and grazing. Tillei density harvesting-or before determined hay management years under in both to harvesting and grazing. Eicettent ofralield and persistence were obtained from all cultivars was management yield under_hay DM to according ranking Cultivar varied under grazing.*ug"r"ni.

but

considerably {if{ere1t from ranking based on performancJunder pasture management. One group of four cultiva.. luq 3 decrease in yield of l9'9Vo in the second year of grLing, over the first year, but yield of the remaining group increased by .5..7.V" (SED = 9.83)' Production (CGR)' of hay was similar foi bottigroups. Early-maturing cultivars yielded 7% more DM and had higher crop growth rates LAI yield, CGR, DM by LAI and TD than late-maturiig cultivars. Growth characteristics of orchardgrass cultivars as measured system management hay under a varieties orchardgrass and TD were modified by management practice. As a result, evaluating was not effective in determining their performance under grazing'

Key words: Dactylis glomerata, cultivar evaluation, cultivar development, grazing, yield' yield

components

McRae, K. B., Laflamme, L. F., Caldwell, C. D' et Fulton, N' R' 1995' pour pelotonn6 I'aptitude I'exploitation en prairie de fauche et en paturage tourDiff6rences entre cultivars de dactyle nant. Can. J. plant Sci. 75: 14:,-157. i'objet de notre 6tude 6tait de comparer le comportement de divers cultivars de dactyle pelotonn6 (Dactylis glomeratal.) sous difffrents systbmes d'exploitation (foin ou pAturage tournant), durant les deux premidres i trois ann6es de productio-n. Huit cultivars de dactyle 6taient semds irla vol6e, selon un dispositif en bloc aldatoires complets prairie r6petitions. Chaque r6p6tition 6tait entour6e d'une clOture et I'emplacement divis6 en deux sections, I'une exploit6e en (m.s.) 6tait mesur6 de fauche (simulee), l';utre en paturage tournant utilisant des brebis crois6es. [r rendement en matibre seche quatre avant la fauche ou la mise en pAture.-La densit6 de tallage (DT) et I'indice de surface foliaire (ISF) dtaient mesur6s sur r6gime en long6vit6 de et rendement de cultivars avant Ia fauche et la p6ore. Tous les cultivars ont fourni des niveaux excellents classement des cultivars selon le rendement de fauche, mais leur comportement en r6gime de piture variait considdrablement. accusait de matidre sbche en regime Oe fauche dlff6rait de celui 6tabli en rdgime de paissance. Un groupe de quatre cultivars que rendement le tandis premidre ann6e, la par d p6turage rapport une chute de rendemeni d e 19,9% dans la deuxidme ann6e de groupes. des quatre autres vari6t6s augmentait de 5,7% (ETD = 9,83). La production de foin 6tait comparable chez les deux foliaire (TcR), indice un 6s cultivars precoces produisaient 7% phts de matidre sdche et manifestaient des taux de croissance mesur6s de croissance les caracGres modifiait d'exploitation r6gime tardifs. et une densit6 de tallage plus 6lev6s que les cultivars par le rendement de iraiidre sdche iar TCR, par ISF et par DT. Par cons6quent, l'6valuation des cultivars en rdgime de fauche ne convient pas pour 6tablir leur comportement en r6gime de pdturage. Papadopoulos,

Y. A.,

Price,

M. A., Hunter, G. M.,

i

k

k

Mots cl6s: Dactylis glomeratal., evaluation

des cultivars, mise au point de cultivars, pdturage, rendement, composantes du rendement

Orchardgrass (Dactylis glomerata L.) is an important forage species in eastern Canada. It is easy to establish, has excellent aftermath production even during the drier surnmer periods, and is well adapted to the region (Jung and Baket 1976; Suzuki 1989: Kunelius 1990). These attributes have led to its increased use in hay and pasture mixtures.

The composition and productivity of an orchardgrass sward plants

is a function of cultivar potential, the number of

Abbreviations: CGR, crop growth rate; DM, dry matter; LAI, leaf area index; TD, tiller density; TDM, total dry matter 147

148

CANADIAN JOURNAL OF PLANT SC'E'VCE

Can. J. Plant Sci. Downloaded from www.nrcresearchpress.com by MICHIGAN STATE UNIV on 01/14/17 For personal use only.

m-2, tillers plant-l, and leaves tiller-I.

(Kunelius 1990).

_The appearance of new orchardgrass leave.s and the growttr

l*f

I

lamina are partially offiet leaf lamina senescenoe. Under field conditions, new orchardgrass leaves appearcd at

9f

apprcximately

ll-d

interrr"als

(Iaylor and TempleonlXOy

anO

remained photosyntlretically active for approximarely 30 d (Ryle

D64).

h

perennial ryegnss (Iatiwt pererue L.), trigtrer [af appearanc€ rat€s were recorded under regrorvth interr"als of D !! 13 d, compared with 19 toz3 dand 30-b 34 d (Javiss 972). Netherbage pnodrction is maximizrd by a balance berrcen plant tissue turnwer and herbage utilization W gazinganimah inircham qd Hodgson D83). For a perenniat ryegnss - white dwer (Tiifoliwn repens L.) canopy, net herbage production is maximized when grazing livestock are manaeO to rnainhin suiard L-AI benveen 2.3 and 4J @ircham anA HoOgson UAfy. Thrget I-AI for livestock enterhg and exiting orchard$ass pad_ docls andother productivity indicators twe not Ueen aOequaety researched. Furthermore, raariations among orctrarOgrasjlarijties for productivity indicabrs influenced Uy suar,O managernent have not yet been prwided. This information is essential to desrgn guidelines for identiffing superior orchardgrass varietieg and optimum management of this species. In Canada, orchardgrass cultivars areielected, registered, and recommended on the basis of dry mafter (DM) production under mechanical harvesting. Recent studies have

demonstrated that mechanical harvesting does not assess accurately the herbage yield potential and quality of swards under grazing (Calder et al. 1970; Frame and Hunt l97l: Jones andRoberts 1986; Orr et al. 1988). Grazing trials combined with clipping management schemes maf Ue a more effective procedure for evaluating and identifuing superior -orc-hardgrass dual-purpose (conserved forage and pasture) cultivars. Unless the cultivar management interaition effect is substantial, the additional costs associated with comparing advanced breeding lines and cultivars under grazing cannot be justified. This sody was conducted to deiermine differences in orchardgrass cultivar performance and the significance of cultivar and cultivar maturity type by management interactions. A complementary objective was to charactenze crop growth characteristics of orchardgrass cultivars under different rnanagement systems.

x

MATERIALS AND METHODS

Orchardgrass

tiltering is increased by enhanced light penetration at the base of the sward (Langer 1963). Sward management can alter crop morphology and productivity. Management that promotes long, unintemrpted growth appears to discouiage tillering (Morris 1969; Kunelius 1990). The growth habit of orchardgrass is generally upright but becomes prosrrare when grazed by livestock (Etherington 1984). Tillei number is influenced by the length of rest period (regrowth interval) between grazing (Morris 1969). Shorter regrowth intervali plulled in higher tiller numbers of orchaidgrass swards. Morris (1969) observed that tiller numbers of rotationallv grazed orchardgrass swards ranged from2489 tillers m-t, when grazing livestock were removed at 5.34 mean leaf area index (LAI), to 2858 tillers m-2 when mean LAI was allowed to drop ta2.95. Tiller number is higher under a fourcut hay management system than a three cuq average tiller numbers were 2123 and l9l2 tillers m-2, respectively

Field Design and ltlanagement The study was initiated in the summer of 1988 on a field with a well-drained Orthic Humo-Ferric Podzol (Herbert series; Nowland and MacDougall 1973) gravelly to very gravelly loam soil changing gradually between the blocks to a moderately-drained Rego Gleysol (Chaswood series; Nowland and MacDougall1973) loam at the Experimental Farm of Agriculture and Agri-Food Canada, Nappan, Nova Scotia. Fertility management of the plots throughout the study was based on the province of Nova Scotia soil test

recommendations. Prior to seeding, the field received 4 t ha -r of dolomitic limestone, 68, 68 and 17 kg ha - I of N, P and K, respectively. All plots were harvested rwice in the seeding year and received 60,20 and 60 kg ha-t ofN, p *d 5, respectively, following the first cut, and 135 kg K ha following the second cut. In both production years

-'

(1989 and 1990), the field was fertilized with

p

and K

according to soil test recommendations, and N was applied

at l60kg ha-' over three application dates: 30April, 5 June and 17 July in 1989, and 25 April, 3l May and 25 June in 1990. Lime was surface broadcast at 1000 kg ha-r in the fall of 1989 to all plots. Eight orchardgrass cultivars, representing early- and latematuring types, were selected for having had good DM yield performance under hay management in eastern Canada. Hallmark, Juno, Hercules and Napier were chosen as earlymaturing, while Rideau, Sumas, Kay and Frode were chosen as late-maturing. The trial consisted of_eight orchardgrass cultivars broadcast seeded (10 kg ha-t) in a randomized complete-block design with three replications. plots were

l00mlongand5mwide.

In 1989, the perimeter of each replication was fenced, and the area of each replication was divided into two sections with a permanent page-wire fencing system. The first section (at the back of each replication) was managed as a hay crop system; each replication contained eight cultivar plots, 5 m long and 5 m wide. The second section (95 m long x z$0 m wide) was subdivided into three equal-sized paddocks and managed as a rotationally grazd system within a replication; each paddock contained eight cultivar plots,

30mlongand5mwide.

Dry Matter Production Crossbred ewes (25 ewes ha-l) were introduced to each paddock within a replication when the average sward height was 25 cm. Animals were removed when the sward height averaged 6 cm. Sward height of each cultivar was obtained by recording l0 random measurements treatment plot-l before and after grazing, using a Hill Farming Research Organization sward stick @arthram 1986). Each paddock was grazed seven times throughout each of the two grazing seasons. Under the hay management system, all cultivars were harvested on the same day; the first cut was harvested when the cultivar Kay (check cultivar) reached the early boot stage in 1989 and the late boot stage in 1990. Aftermath cuts (cut 2 and cut 3) were harvested when plant height was 30 cm. A HEGE 125B plot harvester (HEGE E4uipment Inc.

PAPADOPOUTOS ET AL.

_

EFFECTS OF SWARD MANAGEMENT ON ORCHARDGRASS

KS) was used to simulate hay harvesting; sward height was 6 cm after clipping. Before harvesting each cut or grazing each paddock in the 1989 and 1990 growing seasons, three randomly

Can. J. Plant Sci. Downloaded from www.nrcresearchpress.com by MICHIGAN STATE UNIV on 01/14/17 For personal use only.

selected 0.25-m'zareas were clipped from each plot to a height of 6 cm, using hand-held electric sheep shears. Samples were dried at 70"C for at least 48 h prior to DM yield determination, to estimate percentage DM and, consequently, DM yield.

Crop Growth Analysis Tv'o early- and tun late-maturing cultiran (Hallmark and Juno' I(ay and Frode, respectively) were analyzed for crop growth. One paddock replication-t was chosen for measuring crop grcwth data in the rotational grazing section. BefoIe mechanical harvesting or grazing in 1989 and 1990, four sites were chosen at random in each plot, and the number of tillers within a 400-cm2 (20 x 20 cm) quadrat at each site were counted to estimate tiller density (TD). One half of the quadrat area QN cmz; l0 crn x 20 cm) was cut to a height of 4 cm with a scalpel. Sampled material was separated into four com-

ponents: orchardgrass laminae, orchardgrass stems, dead materials, and weeds. Areas of laminae were measured using a LICOR 3100 planimeter (LICOR Inc., Nebraska), and the ratio of laminae area to unit field area (LAI) was calculated. Sample components were dried as previously described, and DM yields of the two orchardgrass components, laminae and stems, were then summed. The ratio of laminae DM / total DM was used to estimate percentage leaf DM yield (DML/DM). CGR was obtained by dividing DM yield per unit area S ttre number of days of regrowth. The growing time for the fint harvest or grazing was calculated from I May in bottr growing seasons.

Statistical Analysis Data were analyzed as a variant of the split-plot design with three replications, as described by Cochran and Cox (1957). Management systems were the whole plot treatments and cultivars were subplot treatments, To accommodate seeding Table

l.

Herbage DM yield (t

PRODUCT'ON 149

of the plots, fencing and animal handling, whole-plot treatments were not randomized but were arranged systematically

(hay at the back of the trial). As a result, themain effect of management system, hay versus grazing, can be assessed, but the signiflrcance probability ofthe differences is in doubt because there is no valid estimate of error (Cochran and Cox 1957). While the statistical significance of the main effect of management system cannot be estimated, it can be estimated for the management x cultivar interaction. Crops from the same plots are repeated measurements and requirb special methods ofanalysis. Snedecor and Cochran (1980) recommend defining comparisons of interest among the cropping years; these contrasts are then assessed through an analysis of variance. For this study, cultivars that deteriorated under grazing will have had lower second cropping year yield compared with that under hay. Such changes highi Ue called cultivar x management interactions and identifred by changes in lo916 (1990 yield) - lo916 (1989 yield; : logls (1990 yield / 1989 yield)' To assess changes in cultivar yields and their interaction with management over the two growing seasons, the mean annual yield and log of the ratio of 1990 to 1989 yields (denoted logls change) were calculated. Logls change data were analyzed following the above split-plot model to determine the change in cultivar in the first year of this study. yield - after grazingbecropping grouped into early- and late-maturity Cultivars could classes for comparison. For this paper we divided them into two groups based on their lo916 change; this division is based on the results of the experiment and could not be specified a priori. The difference between the group means cln be estimated, but the significance probability of the associated t test will not be exact. Since the purpose ofthe study was to determine the magnitude of such group differences. lack of an exact statistical test was not taken as a serious deficiency. Correlation coefficients among parameters measured for each harvest or gtazingdate and season totals or weighted averages were calculated to determine the strength of association.

ha-l) oforchardgrass cultivars under posture and hay

management

Total annual DM yield Pasture (t

Hay (t ha-r)

ha-r) Vo change!

Juno

E

Hallmark

E

Sumas

L L L

Rideau

Iky

E

Napier Hercules Frode Mean

SEM (n

cv

(%)"

E

L

=

3

df:

14)w

8.82 10.09 7.98

9.16 8.38 9.41 7.17 7.97 8.62

0.793 20.2

5.2

8.27 7.94 7.48 6.94 7.21 8.65

13.8 28.8 27.4

6.33 8.22

8.78 8.95 8.29 8.32 8.09 9.08

7.t7\

7.0

7.@

8.

l0

7.@

8.68 8.31

12.o

6.15

8.50 8.52 0.863

7.33

16.4 39.6 26.5

6.72 8.24 6.48 7.51 8.39

7.77

l0.l I

9.76

7.31' 9.39 0.863

2r.l

zThe seasonal rating of the cultivar (E, early; L, la9. vThe back-transformed means of log (ratio of th" ,"rponre, xBased on two out of three replicates. *Standard error of the mean. vCoefficient of variation.

7.23 8.34 8.38

1990

7.76

-r8.2

).J)

-

1.3

-'l .l

0.671 r7.9

in

-23.9

- 18.9 - 18.6

9.r7

8.O2

% change

1989

2n'

6.93 6.67

6.84 o.793 20.2

2t.l

'1.68

0.67r

5l.0 27.r 13.2

tt!

r7.9

to 1989), expressed as Vo change and used to order the cultivars in the

table

150 2'

Table

CANADIAN JOURNAL OF PLANT SC'E'VCE

Analysis of variance for

l9t9 and l99lf total DIII yield and logle yield

change over the two growing seasons among orchardgrass cul-

tivars and management systems

loglq change 1989

df 2

Source

Block

Can. J. Plant Sci. Downloaded from www.nrcresearchpress.com by MICHIGAN STATE UNIV on 01/14/17 For personal use only.

Management system

Error

A

(S)

(Cv) (G)z Error B Cuftivar

Group

SxCv SXG

MS 38.120 0.106 2.867 0.954 1.325

7

2.OrO 3.Ot2

t4

Residual

MATI'RING

prob.

MS

0.10 0.41

0.59 0.29

0.09 0.06

F prob.

8.703 3o.24g

0.56 0.10

10.059

t0.521 58.247

0.71

0.r2

3.047

l,

lmo)

MS

230.577 tS.7Z6

r.g7g E.676

0.40 0.03

9.188

according to negative or positive change.

CT'IJTTVARS 1989

LATE IVIATURINC CUITTWARS 1989 DMIELD (t/hr)

t2

}TALI.MARtr

tuNo l0

HERCULES 8

prob.

3.855

2.953 6.633 3.2e 6.054 t.A3

DMIIEID (r/hr)

l3

F

3.2t5

2.477

zG, Cultivar persistence type based on % change in Table EAXIJY

F

0.515

I 2 7 I t4

(x

1990

RIDEAU

r

SlrlYlAS

trAY

NAPIER

I

SEM

FRODE SEM

6

6

I

4

2

l3s 183 t85 209 n7 243 %t

t30 t3tt l6:t 188 N

%2

Df,TOFT}GTETR

?2t 2+3 26t 2m

D;iTOFTHETEFR

EA$JY M'TTURING CI'LTIVARI! I99O

LATE T{ATURING CUIJTIVARS I99O

DMIELD (t/hd

DMTIELD (Vha)

t2

II

l0 8

IT

l0 8

6

1

4

2

t00 l{9 169 tgo 219 At zsg 278 300

0

r00 l{9 t69 190 3t9 24r 25ii 2t8

Df,TOFTHE IEf,R

Fig. 1. cumulative DM yield of orchardgrass cultivars under pasture management.

Df,'OFTIIE IEIR

3oO

PAPADOPOULOS ET AL.

Table

3.

_

EFFECTS OF SWARD MANAGEMENT ON ORCHARDGRASS

Mean change (% charrye) in DM yield under rotational grazing

and hay management of two cultivar groups Cultivar group

Hay (7o change)

Decreasedz

4 4

Increased*

- 19.9v - 05.7

29.5!

-

23.7

Can. J. Plant Sci. Downloaded from www.nrcresearchpress.com by MICHIGAN STATE UNIV on 01/14/17 For personal use only.

zCultivars with decline in DM yield after 2 yr of grazing (see Table l). YSED between groups 9.83. = xCultivars with change of or higher for DM yield after 2 yr of grazing (see Table

l).

-1.3%

RESULTS AND DISCUSSION Dry Matter Production Although the F probability for the main effect of management system may be in doubt, this study brings forward evidence oflarge orchardgrass yield differences when swards are subjected to hay versus grazing management (Tables I

In 1989, mean total annual DM yield was substantially higher under grazing than under hay management (Table l). However, in 1990 the mean total DM yield under and 2).

hay management was similar to that recorded under grazing, suggesting that yield differences may not be observed in every grazing season. Higher total annual yield was also

reported for perennial ryegrass under grazing than for frequent clipping (Frame and Hunt 1971). In short-lived forage

PRODUCT'ON

151

percentage change under grazing and hay production, relative persistence of each group under the two management systems can be assessed @bles l, 2 and 3). The calculated significance

probability of 0O3 for the system x group interaction is not adjusted because this was a daa-based comparison, but tttis single degree of freedom contrast explains 80Vo of the interaction sums of squares (Thble 2). There was no significant difference between the groups for hay production, but under grazing the difference was significant (P: 0.01). These results demonstrate that cultivar x management interaction was evident in this study and that evaluating the performance oforchardgrass cultivars under a hay management system is not effective in determining their performance under grazing. Evidence of similar cultivar x management interactions were reported for other forage species (Calder et al. 1970; Frame and Hunt 1971; Jones and Roberts 1986; Swift et al. 1992). Dry matter yields of Napier were consistently above the average of all entries in both years and in the two different management systems (Table 1). It appears that this cultivar was not affected by grazing pressure, did not show DM yield decline in the second production year, and, based on mean annual DM yield, ranked first under both management systems. This is an indication that the cultivar Napier has broad adaptations to practical farm management, where a forage sward is commonly subjected to alternating hay and grazing managements within a growing season.

species, such as ltzlian (Inlium multiJlarwn l-am.) and hybrid

When DM yields were analyzed for individual grazing

ryegrass, the reverse trend was found; total annual yield

dates, differences among cultivars were only evident during

under hay management system was always higher than under grazing (Jones and Roberts 1986).

the summer periods of both production years (data not

Dry matter yield of cultivars varied between the two production years (Tables I and 2). However, the range of

among cultivars were observed on 4 July (day 209;

P < 0.01)

DM yield and year-to-year variations of the cultivars were consistent with good orchardgrass performance in Atlantic Canada and other regions of North America (Bryant and Blaser 1961; Mislevy etal. 1977; Kunelius 1990). Early-

or Hallmark (Fig. l).

maturing cultivars tended to produce higher DM yields (8.19 and 8.18 t ha-r for 1989 and 1990, respectively) than late types (7.27 and 8.00 t ha-' for 1989 and 1990, respec-

tively), but differences were significant (P