Agronomy Journa l • Volume 102 , I s sue 2 • 2010 667

Fora

ges

Published in Agron. J. 102:667–674 (2010)Published online 26 Jan. 2010doi:10.2134/agronj2009.0376Copyright © 2010 by the American Society of Agronomy, 5585 Guilford Road, Madison, WI 53711. All rights re-served. No part of this periodical may be reproduced or transmitted in any form or by any means, electronic or mechanical, including photocopying, recording, or any information storage and retrieval system, without permission in writing from the publisher.

Temperate, perennial grasses are the principal component of pasture and hay lands throughout the

United States (Burns and Bagley, 1996). Because ruminant livestock agriculture depends on productive, persistent grass-lands, their response to frequency and extent of defoliation has been the focus of numerous studies from the perspective of yield and quality (Burns et al., 2002; Holden et al., 2000; Volesky and Anderson, 2007), or crop morphology and persistence (Brummer and Moore, 2000; Dobson et al., 1978; Hart et al., 1971; Horrocks and Washko, 1971). Although these studies have typically used clipping to impose treatments, which may infl ate yield response compared to grazing (Bryant and Blaser, 1968), the results provide important information about fundamental growth processes and species diff erences.

Despite the broad range of cultural and climatic conditions under which these studies were conducted, some general conclu-sions can be drawn from the results. Increasing the time interval between harvests has been shown to increase the annual DM yield of tall fescue (Burns et al., 2002), orchardgrass (Hart et al., 1971), and reed canarygrass (Phalaris arundinacea L.; Geber,

2002), but reduce herbage nutritive value at each harvest (Burns et al., 2002; Holden et al., 2000, Marten and Hovin, 1980). Th e eff ect of harvest height, however, on herbage production and nutritive value is less consistent, and is likely species-dependent. Dobson et al. (1978) found that over 4 yr, lowering the clipping height of tall fescue from 10 to 5 cm increased mean annual DM yield by 22%. In contrast, Volesky and Anderson (2007) found that annual yield of irrigated orchardgrass and smooth brome-grass (Bromus inermis Leyss.) harvested over four time periods at a 7-cm RSH was less than that harvested at a 14- or 21-cm RSH. Harvest height (5, 12.5, and 20 cm) did not infl uence annual yield of reed canarygrass when cut two or three times per year, but annual yield increased as harvest height increased when cut four times per year (Geber, 2002). Unlike yield, the eff ects of harvest height on herbage nutritive value have been relatively small (Burns et al., 2002; Geber, 2002; Volesky and Anderson, 2007).

Tall fescue and orchardgrass comprise a signifi cant propor-tion of the grasses grown in pastures and hay lands within their area of adaptation (van Santen and Sleper, 1996; Sleper and West, 1996). Compared with meadow fescue, tall fescue and orchardgrass have been the subject of considerably more research. A search of articles published in Agronomy Journal and Crop Science since their inception in 1907 and 1961, respectively, to the present listed approximately 700 studies of tall fescue and 450 studies of orchardgrass culture, physiol-ogy, breeding, endophyte association, and animal performance relationships. In contrast, meadow fescue has been the subject of only 15 investigations reported in those journals during the

ABSTRACTMeadow fescue [Schedonorus pratensis (Huds.) P. Beauv.] represents an alternative to temperate grasses typically used in forage-livestock systems. Our objective was to compare the productivity, nutritive value, and persistence of diverse meadow fescue cultivars with that of tall fescue [Lolium arundinaceum (Schreb.) Darbysh.] and orchardgrass (Dactylis glomerata L.) when harvested by regimes representing lax and severe hay production and rotational grazing. ‘Azov’ (plant introduction strain cross), ‘Bartura’ (commercial cultivar), and ‘Hidden Valley’ (naturalized population) meadow fescue, ‘Barolex’ tall fes-cue, and ‘Bronc’ orchardgrass were harvested infrequently (40- to 65-d harvest interval) or frequently (when plants reached 25-cm sward height) to a 5- or 10-cm residual sward height (RSH) at two Wisconsin locations in 2 yr. Annual dry matter (DM) yield of all grasses was greater when harvested infrequently (7.46 Mg ha−1) than frequently (5.92 Mg ha−1), or at 5 cm (7.52 Mg ha−1) than 10-cm RSH (5.88 Mg ha−1). Tall fescue and orchardgrass annual yield was greater than that of all meadow fes-cues when harvested infrequently, but diff erences among grasses were relatively small when harvested frequently, particularly at 10-cm RSH. Neutral detergent fi ber digestibility (NDFD) of meadow fescue was 30 to 80 g kg−1 neutral detergent fi ber (NDF) greater than that of tall fescue or orchardgrass at every harvest in all environments. Meadow fescue cultivars were less persistent than tall fescue aft er 2 yr, but were usually equal to or more persistent than orchardgrass. Meadow fescue should be considered as a viable alternative to tall fescue and orchardgrass in temperate, managed intensive rotational grazing systems due to its comparable yield and superior digestibility.

USDA-ARS, U.S. Dairy Forage Research Center, 1925 Linden Drive West, Madison, WI 53706. Received 28 Sept. 2009. *Corresponding author (Geoff rey.Brink@ars.usda.gov).

Abbreviations: CP, crude protein; DM, dry matter; NDF, neutral detergent fi ber; NDFD, neutral detergent fi ber digestibility; RSH, residual sward height.

Meadow Fescue, Tall Fescue, and Orchardgrass Response to Defoliation Management

G. E. Brink,* M. D. Casler, and N. P. Martin

668 Agronomy Journa l • Volume 102, Issue 2 • 2010

same period. As the subject of research eff orts, its absence in the literature is largely the result of a decline in meadow fescue use in the United States, beginning in the early 1900s. Due to its susceptibility to crown rust (Puccinia coronata Corda) and the discovery that tall fescue had superior vigor and productiv-ity (Buckner et al., 1979), meadow fescue has not been grown on signifi cant acreage since then.

Within the last decade, there has been a resurgence in meadow fescue utilization in temperate regions of the United States due to the release of improved cultivars from European breeding programs and research that documented benefi cial traits compared with tall fescue from a livestock perspective. Casler and co-workers (Casler et al., 1998, Casler and van Santen, 2001) reported that meadow fescue had superior appar-ent DM intake by grazing cattle compared with tall fescue, and concluded that the greatest potential utility for this grass is in managed intensive grazing systems. An additional advantage is that the nutritive value of meadow fescue is typically greater than that of most temperate grasses (Brink et al., 2007). Infor-mation is lacking, however, on the response of meadow fescue to defoliation variables in comparison with typical pasture grasses such as tall fescue and orchardgrass. Th e objective of this study was to compare the productivity, nutritive value, and persistence of diverse meadow fescue cultivars with those of tall fescue and orchardgrass under defoliation regimes represent-ing lax and severe hay production and rotational grazing in contrasting environments.

MATERIALS AND METHODSTh e experiment was conducted in 2005 and 2006 at the

University of Wisconsin Lancaster Agricultural Research Station (42°51́ 0˝ N, 90°42´36˝ W) on a Rozetta silt loam (fi ne-silty, mixed, superactive, mesic Typic Hapludalf) and at the University of Wisconsin Marshfi eld Agricultural Research Station (44°38´60˝ N, 90°7 4́8˝ W) on a Withee silt loam (fi ne-loamy, mixed, superactive, frigid Aquic Glossudalf). Th e soils at both sites had pH 6.5 to 6.8, 52 to 75 ppm P (Bray P1), and 117 to 138 ppm K. Monthly historical and actual pre-cipitation is presented in Fig. 1. In April 2004, seed of Azov, Bartura, and Hidden Valley meadow fescue, Barolex tall fescue, and Bronc orchardgrass were broadcast at 11.2 kg pure live seed ha−1 in 1.22- by 2.74-m plots on a prepared seedbed. Azov is a strain cross of 11 Russian or Ukranian plant introductions (PI numbers 314525, 314528, 314570, 315435, 315436, 315438, 324122, 325525, 440357, 440358, and 502375) selected for high forage yield under managed intensive rotational grazing (Casler and van Santen, 2001). Bartura is a commercial cultivar developed in the Netherlands by the Royal Barenbrug Group. Hidden Valley is a naturalized population of meadow fescue found on a dairy farm near Mineral Point, WI (42°51́ 36˝ N, 90°12´36˝ W), where it is thought to have originated sometime shortly aft er European colonization of the region in the 1850s. Plot size was dictated by the limited availability of Azov and Hidden Valley meadow fescue seed.

Th e experimental design was a split-plot arrangement of a randomized complete block design with harvest management (four combinations of two frequencies and two RSHs) as the whole plot and grass cultivar as the subplot in four replicates. A 1-m border of ‘Phoenix’ turf-type tall fescue surrounded each whole plot. Plots were clipped to a 10-cm RSH and fertilized with 44.8 kg N ha−1 as NH4NO3 in June and August of the seeding year.

Before grass growth began in early April 2005 and 2006, plots were clipped to a 5-cm RSH to remove winter-killed herbage. Plots were fertilized in mid-April with 67.2 kg N ha−1 as NH4NO3. Plots were harvested from May to September of each year (Table 1) at a frequency typical for either hay manage-ment (infrequent; 40- to 65-d harvest interval) or rotational grazing (frequent; when sward height reached approximately 25 cm) and at either a 5- or 10-cm RSH. Based on 10 random tillers sampled before harvest, maturity of grasses cut infre-quently was R1 to R3 (Moore and Moser, 1995) at fi rst harvest, and stage V4 to V5 at the second and third harvests, while maturity of grasses cut frequently was stage V3 to V4 at all harvests. Plots retained the same treatment assignment in 2005 and 2006. Forage DM yield was determined by cutting a 50-cm swath through the center of each plot using a rotary mower equipped with a catch basket. A 600- to 800-g subsample was taken from each yield sample, dried at 65°C for 48 h, and weighed to determine DM content. Aft er the fi rst and second infrequent harvests and aft er the second and fourth frequent harvests, appropriate plots were fertilized with 67.2 kg N ha−1 as NH4NO3 for an annual total of 201.6 kg N ha−1.

Dry matter samples from each plot were ground to pass a 1-mm Wiley mill screen, and 50-g subsamples of ground forage were stored in plastic bottles. Ground samples were analyzed for N, neutral detergent fi ber (NDF), and in vitro neutral

Fig. 1. Monthly average historical (1970–2000) and actual precipitation during 2 yr at Lancaster and Marshfield, WI.

Agronomy Journa l • Volume 102, Issue 2 • 2010 669

detergent fi ber digestibility (NDFD) by calibrated near-infra-red refl ectance spectroscopy. Herbage N [crude protein (CP) = N × 6.25] was measured by the Dumas method (Bremner, 1996), NDF by the method of Mertens (2002), and NDFD by the method of Goering and Van Soest (1970). Calibration sta-tistics were the following: N, standard error of prediction cor-rected for bias [SEP(C)] = 0.47 and R2 = 0.98; NDF, SEP(C) = 0.83 and R2 = 0.98; NDFD, SEP(C) = 2.45 and R2 = 0.92.

Treatment eff ects on grass persistence were measured in 2007 on 27 April at Lancaster and on 17 May at Marshfi eld using a point intercept method (Jonasson, 1983) to estimate ground cover. A 2-m tape was placed diagonally across each

plot and the presence or absence of a tiller of the seeded grass was noted at 10-cm intervals. Persistence was calculated as the proportion of total points (20) intercepting a tiller of the seeded grass and was expressed as a percentage.

Data were analyzed by the Mixed Models procedure of SAS, with block considered a random eff ect, and year, location, har-vest management, and grass cultivar considered fi xed eff ects. Nutritive value data were analyzed by harvest frequency (infre-quent and frequent) due to the unequal number of harvests for the two treatments. Main and interaction eff ects for nutritive value were evaluated according to P values and percentage of sums of squares; eff ects were ignored if their contribution

Table 1. Dates of harvest of fi ve grass cultivars harvested infrequently and frequently at two Wisconsin locations over 2 yr.

CutLancaster Marshfi eld

Infrequent Frequent Infrequent Frequent2005 2006 2005 2006 2005 2006 2005 2006

1 26 May 1 June 3 May 10 May 10 June 8 June 23 May 23 May2 8 July 24 July 23 May 1 June 26 July 9 Aug. 10 June 12 June3 30 Aug. 26 Sept. 16 June 22 June 21 Sept. 29 Sept. 28 June 3 July4 5 July 19 July 18 July 3 Aug.5 28 July 16 Aug. 25 Aug. 30 Aug.6 26 Aug. 21 Sept. 16 Sept. 22 Sept.

Table 2. Signifi cance (P > F) of main effects and their interactions for annual dry matter (DM) yield and ground cover, and herbage crude protein (CP), neutral detergent fi ber (NDF), and neutral detergent fi ber digestibility (NDFD) of meadow fescue, tall fescue, and orchardgrass.

Effect DM yield

Ground cover

CP NDF NDFDInfreq. Freq. Infreq. Freq. Infreq. Freq.

P > FLocation (L) 0.005 0.008 0.014 0.194 0.009 0.004 <0.001 0.936Harvest mgt. (H)† <0.001 <0.001 0.003 0.332 0.444 0.154 0.438 0.832L × H 0.864 0.074 0.896 0.047 0.045 0.078 0.029 0.030Cultivar (Cv) <0.001 <0.001 <0.001 <0.001 <0.001 <0.001 <0.001 <0.001Cv × L 0.013 0.037 0.032 0.112 <0.001 <0.001 0.001 0.048Cv × H 0.007 0.049 0.147 0.040 0.244 0.161 0.103 0.035Cv × L × H 0.071 0.160 0.030 0.642 0.562 0.484 0.290 0.244Year (Y) 0.887 <0.001 <0.001 <0.001 <0.001 0.018 0.500Y × L 0.197 0.656 <0.001 <0.001 <0.001 <0.001 <0.001Y × Cv <0.001 0.102 0.044 <0.001 0.002 <0.001 <0.001Y × L × Cv 0.470 0.026 0.217 0.018 <0.001 0.067 0.003Y × H 0.360 0.123 0.915 0.067 <0.001 <0.001 <0.001Y × L × H <0.001 0.672 0.001 0.420 0.078 0.628 0.233Y × H × Cv 0.116 0.192 0.046 0.591 0.144 0.534 <0.001Y × L × H × Cv 0.508 0.606 0.018 0.529 0.068 0.248 0.665Cut (C) <0.001 <0.001 <0.001 <0.001 <0.001 <0.001C × L <0.001 <0.001 <0.001 <0.001 <0.001 <0.001C × H <0.001 <0.001 <0.001 <0.001 <0.001 <0.001C × L × H 0.097 <0.001 <0.001 0.003 <0.001 0.070C × Cv <0.001 <0.001 <0.001 <0.001 <0.001 <0.001C × L × Cv <0.001 <0.001 <0.001 <0.001 <0.001 <0.001C × H × Cv <0.001 0.001 <0.001 <0.001 <0.001 0.002C × L × H × Cv 0.888 0.074 0.526 0.584 0.360 0.141C × Y <0.001 <0.001 <0.001 <0.001 <0.001 <0.001C × L × Y <0.001 <0.001 <0.001 <0.001 <0.001 <0.001C × Cv × Y <0.001 <0.001 <0.001 <0.001 <0.001 <0.001C × L × Cv × Y <0.001 <0.001 0.110 <0.001 0.820 <0.001C × H × Y 0.392 <0.001 0.153 <0.001 0.002 <0.001C × L × H × Y 0.020 <0.001 0.009 <0.001 0.080 0.061C × H × Cv × Y 0.666 0.010 0.466 0.003 0.150 0.388C × L × H × Cv × Y 0.413 0.052 0.141 0.286 0.063 0.513

† Harvest management (H) effect refers to the combination of harvest frequency and residual sward height for DM yield and ground cover, and to residual sward height for CP, NDF, and NDFD.

670 Agronomy Journa l • Volume 102, Issue 2 • 2010

to total treatment sums of squares was relatively low (<10%). Harvest management eff ects were evaluated using orthogonal, single degree-of-freedom contrasts and grass cultivars were compared by Fisher’s LSD (P ≤ 0.05).

RESULTS AND DISCUSSIONAnnual Dry Matter Yield

Th e eff ect of harvest management on annual DM yield of all grass cultivars was relatively consistent despite a signifi cant (P < 0.001) year × location by harvest management interac-tion (Table 2). At Lancaster and Marshfi eld in 2005 and 2006, grasses harvested infrequently (three times per year) produced greater annual DM yield than those harvested frequently (six times per year) at the same RSH, and grasses cut at a 5-cm RSH produced greater annual yield than those cut at a 10-cm RSH within either harvest frequency (Table 3). Th e results are in general agreement with those of previous studies of the eff ects of harvest frequency (Geber, 2002; Hart et al., 1971) and har-vest height (Dobson et al., 1978) on temperate grass yield. Th e results also corroborate the fi ndings of Parsons et al. (1988). Using a grass growth model, these authors found that severe defoliation, coupled with a long regrowth period, analogous to the infrequent harvest/5-cm RSH treatment employed here, maximized yield over a growing season because maximum leaf area was maintained over an extended time period. Although frequent, severe defoliation (5-cm RSH) of the sward increased annual yield by an average of 1500 kg DM ha−1 compared with a 10-cm RSH (Table 3), cattle grazing the lower portion of the canopy (5–10 cm) would be forced to consume herbage of lower nutritive value (Brink et al., 2007).

Typical of temperate grasses, a large proportion of the annual DM yield was produced in the spring (Fales, 1986); 48 and 60% of the annual DM was produced at the fi rst cut at

Lancaster and Marshfi eld, respectively, by grasses har-vested infrequently, and 46 and 52% of the annual DM was produced at the fi rst and second cuts at Lancaster and Marshfi eld, respectively, by grasses harvested frequently (mean of 2 yr; data not shown). For grasses harvested infrequently, the remaining proportion of annual yield was evenly distributed between the second and third cuts. For grasses harvested frequently, however, individual-harvest DM yield followed a production trend typical for this region, declining through the summer and increasing in the fall.

Diff erences in annual DM yield among the grass cultivars were dependent on harvest management. Th e signifi cant grass cultivar × harvest management interac-tion (Table 2) was in part due to the fact that annual DM

yield of all meadow fescue cultivars was less (P ≤ 0.05) than that of tall fescue when harvested either infrequently or fre-quently at a 5-cm RSH, but when harvested at a 10-cm RSH, yield of Azov and Hidden Valley meadow fescue was similar to that of tall fescue (Table 4). For producers, these results suggest that yield diff erences between meadow fescue and tall fescue will be less when clipped or grazed according to recommended guidelines for rotational grazing (Undersander et al., 2002). Of the meadow fescues, only Azov was productive as orchardgrass when harvested by any combination of harvest frequency and height (Table 4).

Signifi cant cultivar × location and cultivar × year interac-tions for annual DM yield were also found (Table 2). Aver-aged across all harvest management combinations and years, annual yield of tall fescue and orchardgrass was greater than that of all the meadow fescue cultivars at Lancaster (Fig. 2A) in southwestern Wisconsin. At Marshfi eld in north central Wisconsin, however, yield of Azov meadow fescue was similar to that of Barolex tall fescue and Bronc orchardgrass, and yield of Hidden Valley meadow fescue was similar to that of Bronc orchardgrass. Tall fescue has its origin in Europe and North Africa (Sleper and West, 1996), and orchardgrass has its origin in the Mediterranean region, western Europe, and Asia (van Santen and Sleper, 1996). In contrast, meadow fescue is a major constituent of grasslands in the Nordic and Baltic countries due to its superior winter-hardiness (Fjellheim et al., 2009). Th e comparative yield of Azov and Hidden Valley meadow fescue relative to tall fescue and orchardgrass may be attributed to preferential adaptation to the cooler climate of Marshfi eld (1420 growing degree days 10base) compared with Lancaster (1540 growing degree days 10base).

Averaged across harvest management and location, cultivar productivity also diff ered between years. With the exception of

Table 4. Annual dry matter (DM) yield of fi ve grass cultivars harvested by four combinations of frequency and residual sward height (RSH; mean of 2 yr and two locations).

CultivarFrequency–RSH

Infrequent–5 cm Infrequent–10 cm Frequent–5 cm Frequent–10 cmkg DM ha–1

Azov meadow fescue 8480 6680 6740 5220Bartura meadow fescue 7900 6040 6250 4850Hidden Valley meadow fescue 7790 6490 6420 5030Barolex tall fescue 9200 6650 7020 5280Bronc orchardgrass 8460 7000 6920 5500LSD (0.05) 540 460 270 360

Table 3. Annual dry matter (DM) yield of grasses harvested by four combinations of frequency and residual sward height (RSH) at two Wisconsin locations in 2 yr (mean of fi ve grass cultivars).

Frequency – RSHLancaster Marshfi eld

2005 2006 2005 2006kg DM ha–1

Infrequent–5 cm 8330 9300 8180 7640Infrequent–10 cm 6830 7150 6400 5900Frequent–5 cm 7060 7060 6300 6250Frequent–10 cm 5720 5600 4610 4770

Contrast P > F Infreq.–5 cm vs. Infreq.–10 cm <0.001 <0.001 <0.001 <0.001 Freq.–5 cm vs. Freq.–10 cm <0.001 <0.001 <0.001 <0.001 Infreq.–5 cm vs. Freq.–5 cm <0.001 <0.001 <0.001 <0.001 Infreq.–10 cm vs. Freq.–10 cm <0.001 <0.001 <0.001 <0.001

Agronomy Journa l • Volume 102, Issue 2 • 2010 671

Azov meadow fescue, which produced the most annual yield, no yield diff erences were measured among the grasses in 2005. In 2006, however, mean yield of tall fescue and orchardgrass was greater than that of all meadow fescue cultivars (Fig. 2B). Th is result suggests a possible reduction in yield potential over time for meadow fescue relative to tall fescue and orchardgrass, but longer-term studies will be required to verify and quantify this possible occurrence and to determine its cause.

Nutritive Value

For herbage CP concentration, the main eff ect of cultivar and 8 of 15 cultivar × environment (location, harvest manage-ment, year, cut) interactions for infrequent harvesting, plus the main eff ect of cultivar and 10 of 15 cultivar × environment interactions for frequent harvesting were signifi cant (P ≤ 0.05; Table 2). Of those eff ects, the cultivar main eff ect and the cultivar × cut interaction accounted for 73% of all the vari-ability associated with cultivar for infrequent harvesting and 65% of all the variability associated with cultivar for frequent harvesting.

Herbage CP concentration of grasses harvested frequently was infl uenced largely by timing of N application (data not shown). Mean CP of herbage produced aft er N was applied (176, 193, and 204 mg kg−1 for the fi rst, third, and fi ft h cut, respectively) was 40 mg kg−1 higher than herbage produced for the second, fourth, and sixth cut. Th ere were few diff erences among cultivars for CP concentration for any cut, except in the late summer and fall when tall fescue CP was 20 to 30 mg kg−1 less than that of the other grasses (data not shown). Similarly, few diff erences in herbage CP existed among cultivars har-vested infrequently (mean CP of 111, 137, and 139 mg kg−1 for the fi rst, second, and third cut, respectively).

Th e main eff ect of cultivar and 7 of 15 cultivar × environ-ment (location, harvest management, year, cut) interactions on NDF concentration of herbage harvested infrequently, plus the main eff ect of cultivar and 10 of 15 cultivar × environ-ment interactions on NDF concentration of herbage harvested frequently were signifi cant (P ≤ 0.05; Table 2). Of those eff ects, the cultivar main eff ect and the cultivar × cut, cultivar × loca-tion, and cultivar × cut × location interactions accounted for 79% of all variability associated with cultivar for infrequent harvesting and 73% of all variability associated with cultivar for frequent harvesting.

As the interactions above indicate, location had a signifi cant eff ect on NDF concentration diff erences among the grasses at each cut. When harvested infrequently (Table 5) or frequently (Table 6) at Lancaster, NDF concentra-tion of the meadow fescue cultivars was oft en less than those of tall fescue or orchardgrass. In contrast, this diff er-ence in NDF between meadow fescue and tall fescue or orchardgrass was less consistent at Marshfi eld. Th ese results were likely due to diff erences in growth rate among the grasses at Lancaster and Marshfi eld. Although all grasses were observed to be within a relatively nar-row range of maturity at harvest, tall fescue and orchardgrass were probably

more mature than meadow fescue under the warmer growing conditions at Lancaster compared with Marshfi eld (Brink and Casler, 2005, unpublished data).

For herbage NDFD concentration, the main eff ect of cultivar and six of 15 cultivar × environment (location, harvest management, year, cut) interactions for infrequent harvesting plus the main eff ect of cultivar and 10 of 15 cultivar × environ-ment interactions for frequent harvesting were signifi cant (P ≤ 0.05; Table 2). Of those eff ects, the cultivar main eff ect and the cultivar × cut interaction accounted for 76% of all the variabil-ity associated with cultivar for both infrequent and frequent harvesting.

Unlike CP and NDF concentration, large and consistent diff erences in NDFD were measured among the grass cultivars

Fig. 2. Annual dry matter (DM) yield of five temperate grass cultivars (MF, meadow fescue; TF, tall fescue; OG, orchardgrass) at two locations (A, mean of four harvest management combinations and 2 yr) and for 2 yr (B, mean of four harvest management combinations and two locations). LSD bars apply to comparisons among cultivars for separate (A) locations or (B) years.

Table 5. Neutral detergent fi ber concentration at each cut of fi ve grass cultivars harvest-ed infrequently at two Wisconsin locations (mean of two residual sward heights and 2 yr).

CultivarLancaster Marshfi eld

Cut†1 2 3 1 2 3

g kg–1Azov meadow fescue 543 526 532 620 528 485Bartura meadow fescue 536 524 510 605 532 483Hidden Valley meadow fescue 515 526 501 600 526 476Barolex tall fescue 523 537 528 573 549 503Bronc orchardgrass 564 551 556 590 521 497LSD (0.05) 12 6 9 10 9 6

† See Table 1 for harvest date of each cut.

672 Agronomy Journa l • Volume 102, Issue 2 • 2010

throughout the growing season at both loca-tions in both years. At each cut, meadow fescue cultivars had greater NDFD than tall fescue or orchardgrass when harvested infrequently (Table 7) or frequently (Table 8). Oba and Allen (1999) summarized the relationship between NDFD and dairy cow performance reported in the literature, and reported that while marginal increases in NDFD might not be linearly related to animal response, enhanced NDF digestibility of forage improves DM intake and milk yield of dairy cows. Th us, whether harvested for hay or grazed rotation-ally, the results suggest that this nutritive value parameter represents the greatest advantage meadow fescue has over tall fescue and orchardgrass.

Persistence

A signifi cant (P ≤ 0.05) grass cultivar × harvest management interaction was

observed for persistence as measured by ground cover (Table 2). Within every harvest management combination, however, tall fescue had the greatest ground cover of all grasses aft er 2 yr (Table 9). Brummer and Moore (2000) reported that tall fescue, with or without the fungal endophyte Neoty-phodium coenophialum, had greater percentage stand than orchardgrass, prairie bromegrass (Bromus catharticus Vahl [syn. B. willdenowii Kunth]), reed canarygrass, and smooth bromegrass when subject to continuous stocking to maintain a RSH < 5 cm. In addition to being endophyte-free, the tall

Table 7. Neutral detergent fi ber digestibility at each cut of fi ve grass cultivars harvested infrequently (mean of two residual sward heights, two locations, and 2 yr).

CultivarCut†

1 2 3g kg–1 NDF

Azov meadow fescue 653 662 659Bartura meadow fescue 657 660 686Hidden Valley meadow fescue 684 677 700Barolex tall fescue 614 590 642Bronc orchardgrass 598 604 613LSD (0.05) 12 8 10

† See Table 1 for harvest date of each cut.

Table 6. Neutral detergent fi ber concentration at each cut of fi ve grass cultivars harvested frequently at two Wisconsin locations (mean of two residual sward heights and 2 yr).

CultivarCut†

1 2 3 4 5 6g kg–1

LancasterAzov meadow fescue 375 467 461 495 496 495Bartura meadow fescue 358 473 471 502 495 482Hidden Valley meadow fescue 360 478 479 503 488 470Barolex tall fescue 404 497 496 525 522 505Bronc orchardgrass 400 476 487 519 508 530LSD (0.05) 9 9 6 6 4 7

Marshfi eldAzov meadow fescue 441 522 471 520 467 448Bartura meadow fescue 430 523 475 512 476 444Hidden Valley meadow fescue 420 558 473 505 462 441Barolex tall fescue 450 546 487 528 478 466Bronc orchardgrass 437 521 491 512 483 451LSD (0.05) 9 10 6 6 9 5

† See Table 1 for harvest date of each cut.

Table 8. Neutral detergent fi ber digestibility at each cut of fi ve grass cultivars harvested frequently (mean of two residual sward heights, two locations, and 2 yr).

CultivarCut†

1 2 3 4 5 6g kg–1 NDF

Azov meadow fescue 801 752 772 703 755 726Bartura meadow fescue 809 754 776 723 766 742Hidden Valley meadow fescue 810 745 792 725 778 752Barolex tall fescue 748 695 723 658 725 702Bronc orchardgrass 772 710 741 660 719 682LSD (0.05) 9 8 9 8 8 7

† See Table 1 for harvest date of each cut

Table 9. Persistence (% ground cover) after 2 yr of fi ve grass cultivars harvested by four management combinations of frequency and residual sward height (RSH; mean of two locations).

Frequency–RSH

CultivarAzov

meadow fescue

Bartura meadow fescue

Hid. Val. meadow fescue

Barolex tall

fescue

Bronc orchard-

grass

LSD (0.05)

% ground coverInfrequent–5 cm 52 72 67 81 61 6Infrequent–10 cm 75 79 86 94 69 8Frequent–5 cm 75 82 81 100 79 6Frequent–10 cm 84 87 91 99 88 5Contrast P > F Infreq.–5 cm vs. Infreq.–10 cm 0.002 0.047 <0.001 <0.001 0.035 Freq.–5 cm vs. Freq.–10 cm 0.170 0.266 0.016 0.470 0.119 Infreq.–5 cm vs. Freq.–5 cm 0.002 0.025 0.003 <0.001 0.001 Infreq.–10 cm vs. Freq.–10 cm 0.170 0.079 0.178 0.008 0.002

Agronomy Journa l • Volume 102, Issue 2 • 2010 673

fescue cultivar used here was selected for improved palatabil-ity and livestock acceptance (the soft -leaf trait). Even with the incorporation of this trait, tall fescue retained the productiv-ity and persistence advantage others have observed in tall fescue compared with other temperate grasses in a variety of environments (Brummer and Moore, 2000; Coblentz et al., 2006; Denison and Perry, 1990; Hart et al., 1971).

An important cause of the grass cultivar × harvest manage-ment interaction for ground cover was that when harvesting occurred frequently at either a 5- or 10-cm RSH, ground cover was similar for the meadow fescue cultivars and orchardgrass (Table 9). However, when harvesting occurred infrequently at either RSH, orchardgrass ground cover was less than that of Bartura and Hidden Valley meadow fescue, suggesting that the low-light environment at the base of the plant resulting from infrequent defoliation had a more detrimental eff ect on tiller production (Nelson, 1996) in orchardgrass than in the two meadow fescue cultivars.

The effect of RSH on ground cover was generally con-sistent with those of previous reports when defoliation occurred infrequently (Volesky and Anderson; 2007) or frequently (Carlassare and Karsten, 2003). When harvested infrequently, ground cover of all grasses was reduced by lowering RSH from 10 to 5 cm (Table 9). The stem bases of these grasses serves as a temporary storage of carbohydrates, which contributes to growth of new leaves following defolia-tion (Nelson, 1996). Defoliation at the lower RSH not only removed additional leaf area, but also likely reduced the quantity of carbohydrates available for growth and survival of the plant. With the exception of Hidden Valley meadow fescue, ground cover was not affected by RSH when harvest occurred frequently (Table 9). A possible explanation for this result is that during the two decades in which Hidden Valley meadow fescue was grown on the southwestern Wis-consin farm on which it was collected, the grazing pressure imposed by the producer resulted in an ecotype not adapted to a lower RSH.

Th e eff ect of harvest frequency on ground cover of each grass was dependent on RSH. Harvesting infrequently at a 5-cm RSH reduced ground cover of all cultivars compared with frequent harvesting at a 5-cm RSH (Table 9) due likely to the inhibition of tiller production by shading at the stem base (Nelson, 1996). When cut at a 10-cm RSH, however, infre-quent harvesting reduced (P ≤ 0.05) ground cover of tall fescue and orchardgrass but not that of any meadow fescue cultivar. Langer et al. (1964) reported similar results for tiller density of ‘Aberystwyth S215’ meadow fescue when defoliated either three or seven times during the growing season.

The grass cultivar × location interaction for ground cover was significant (P ≤ 0.05; Table 2) due primarily to a dif-ference in survival of two meadow fescue cultivars. Ground cover was similar at Lancaster and Marshfield for tall fescue (92 and 94%, respectively), orchardgrass (78 and 70%, respectively), and Azov meadow fescue (73 and 70%, respec-tively). In contrast, Hidden Valley and Bartura meadow fescue had greater survival at Lancaster (86 and 85%, respectively) than at Marshfield (77 and 76%, respectively), indicating that possibly due to their genetic background,

these two cultivars are better adapted to the more moderate climate at Lancaster.

CONCLUSIONSPrevious studies reported the potential value of meadow

fescue in rotational grazing systems due to its preference by livestock over other temperate grasses (Casler et al., 1998). Our results indicate that, while meadow fescue cultivars of diverse genetic background respond to defoliation frequency and height in a manner similar to tall fescue and orchard-grass, differences in annual productivity and persistence among the grasses were relatively small when managed according to recommended grazing guidelines for frequency (25–30 cm sward height) and RSH (10 cm). The major agronomic advantage that meadow fescue has over tall fescue and orchardgrass that favors its inclusion in pasture-based livestock systems is its superior fiber digestibility, an aspect of nutritive value that is positively associated with livestock intake and performance (Oba and Allen, 1999). Unlike nutritive value parameters that were more subject to the inf luence of management and location (CP and NDF), meadow fescue had superior NDFD under all harvest fre-quency and height combinations, and in all environments. From a producer’s perspective, persistence is also a crucial factor differentiating perennial grass performance. While persistence of meadow fescue was usually less than that of tall fescue, it was equivalent or superior to that of orchard-grass. Differences in persistence among the cultivars tested suggest that genetic improvement of persistence should be possible in meadow fescue.

REFERENCES

Bremner, J.M. 1996. Nitrogen- total. p. 1085–1121. In D.L. Sparks et al (ed.) Methods of soil analysis. Part 3. Chemical Methods. Book series no. 5. Soil Science Society of America and American Society of Agronomy, Madison, WI.

Brink, G.E., M.D. Casler, and M.B. Hall. 2007. Canopy structure and neu-tral detergent fiber differences among temperate perennial grasses. Crop Sci. 47:2182–2189.

Brummer, C.E., and K.J. Moore. 2000. Persistence of perennial cool-season grass and legume cultivars under continuous grazing by beef cattle. Agron. J. 92:466–471.

Bryant, H.T., and R.E. Blaser. 1968. Effects of clipping compared to graz-ing of ladino clover-orchardgrass and alfalfa-orchardgrass mixtures. Agron. J. 60:165–166.

Buckner, R.C., J.B. Powell, and R.V. Frakes. 1979. Historical development. p. 1–8. In R.C. Buckner and L.P. Bush (ed.) Tall Fescue. Agron. Monogr. 20. ASA, CSSA, and SSSA, Madison, WI.

Burns, J.C., and C.P. Bagley. 1996. Cool-season grasses for pastures. p. 321–356. In L.E. Moser et al (ed.) Cool-season forage grasses. Agron. Monogr. 34. ASA, CSSA, and SSSA, Madison, WI.

Burns, J.C., D.S. Chamblee, and F.G. Giesbrecht. 2002. Defoliation inten-sity effects on season-long dry matter distribution and nutritive value of tall fescue. Crop Sci. 42:1274–1284.

Carlassare, M., and H.D. Karsten. 2003. Species population dynamics in a mixed pasture under two rotational sward grazing height regimes. Agron. J. 95:844–854.

Casler, M.D., D.J. Undersander, C. Fredericks, D.K. Combs, and J.D. Reed. 1998. An on-farm test of perennial forage grass varieties under management intensive grazing. J. Prod. Agric. 11:92–99.

Casler, M.D., and E. van Santen. 2001. Performance of meadow fescue acces-sions under management-intensive grazing. Crop Sci. 41:1946–1953.

674 Agronomy Journa l • Volume 102, Issue 2 • 2010

Coblentz, W.K., K.P. Coff ey, T.F. Smith, D.S. Hubbell, III, D.A. Scarbrough, J.B. Humphry, B.C. McGinley, J.E. Turner, J.A. Jennings, C.P. West, M.P. Popp, D.H. Hellwig, D.L. Kreider, and C.F. Rosenkrans, Jr. 2006. Using orchardgrass and endophyte-free fescue versus endophyte-infected fescue overseeded on bermudagrass for cow herds: I. Four-year summary of forage characteristics. Crop Sci. 46:1919–1928.

Denison, R.F., and H.D. Perry. 1990. Seasonal growth rate patterns for orchard-grass and tall fescue on the Appalachian plateau. Agron. J. 82:869–873.

Dobson, J.W., E.R. Beaty, and C.D. Fisher. 1978. Tall fescue yield, tillering, and invaders as related to management. Agron. J. 70:662–666.

Fales, S.L. 1986. Eff ects of temperature on fi ber concentration, composition, and in vitro digestion kinetics of tall fescue. Agron. J. 78:963–966.

Fjellheim, S., I. PaŠakinskien’e, S. Grønnerød, V. Paplauskien’e, and O.-A. Rognli. 2009. Genetic Structure of Local Populations and Cultivars of Meadow Fescue from the Nordic and Baltic Regions. Crop Sci. 49:200–210.

Geber, U. 2002. Cutting frequency and stubble height of reed canary grass (Phalaris arundinacea L.): Infl uence on quality and quantity of biomass for biogas production. Grass Forage Sci. 57:389–394.

Goering, H.K., and P.J. Van Soest. 1970. Forage fi ber analyses (apparatus, reagents, procedures, and some applications). Agric. Handbook 379. ARS, USDA, Washington, DC.

Hart, R.H., G.E. Carlson, and D.E. McCloud. 1971. Cumulative eff ects of cutting management on forage yields and tiller densities of tall fescue and orchardgrass. Agron. J. 63:895–898.

Holden, L.A., G.A. Varga, G.A. Jung, and J.A. Shaff er. 2000. Comparison of ‘Grasslands Puna’ chicory and orchardgrass for multiple harvests at dif-ferent management levels. Agron. J. 92:191–194.

Horrocks, R.D., and J.B. Washko. 1971. Studies of tiller formation in reed canarygrass (Phalaris arundinacea L.) and ‘Climax’ timothy (Phleum pretense L.). Crop Sci. 11:41–45.

Jonasson, S. 1983. Th e point intercept method for non-destructive estimation of biomass. Phytocoenologia 11:385–388.

Langer, R.H.M., S.M. Ryle, and O.R. Jewiss. 1964. Th e changing plant and tiller populations of timothy and meadow fescue swards. I. Plant survival and the pattern of tillering. J. Appl. Ecol. 1:197–208.

Marten, G.C., and A.W. Hovin. 1980. Harvest schedule, persistence, yield, and quality interactions among four perennial grasses. Agron. J. 72:378–387.

Mertens, D.R. 2002. Gravimetric determination of amylase-treated neutral detergent fi ber in feeds with refl uxing in beakers or crucibles: Collabora-tive study. J. AOAC Int. 85:1217–1240.

Moore, K.J., and L.E. Moser. 1995. Quantifying developmental morphology of perennial grasses. Crop Sci. 35:37–43.

Nelson, C.J. 1996. Physiology and developmental morphology. p. 87–126. In L.E. Moser et al (ed.) Cool-season forage grasses. Agron. Monogr. 34. ASA, CSSA, and SSSA, Madison, WI.

Oba, M., and M.S. Allen. 1999. Evaluation of the importance of the digest-ibility of neutral detergent fi ber from forage: Eff ects on dry matter intake and milk yield of dairy cows. J. Dairy Sci. 82:589–596.

Parsons, A.J., I.R. Johnson, and A. Harvey. 1988. Use of a model to optimize the interaction between frequency and severity of intermittent defo-liation and to provide a fundamental comparison of the continuous and intermittent defoliation of grass. Grass Forage Sci. 43:49–59.

van Santen, E., and D.A. Sleper. 1996. Orchardgrass. p. 503–534. In L.E. Moser et al (ed.) Cool-season forage grasses. Agron. Monogr. 34. ASA, CSSA, and SSSA, Madison, WI.

Sleper, D.A., and C.P. West. 1996. Tall fescue. p. 471–502. In L.E. Moser et al (ed.) Cool-season forage grasses. Agron. Monogr. 34. ASA, CSSA, and SSSA, Madison, WI.

Undersander, D.J., E.C. Albert, D.R. Cosgrove, D.G. Johnson, and P.R. Peter-son. 2002. Pastures for profi t: A guide to rotational grazing. Univ. of Wisconsin Coop. Ext. Publ. A3529. Univ. of Wisconsin, Madison.

Volesky, J.D., and B.E. Anderson. 2007. Defoliation eff ects on production and nutri-tive value of four irrigated cool-season perennial grasses. Agron. J. 99:494–500.