B-1695 L January 1991 x) Productivity of 10 Warm - Season Perennial Grasses Over Several Years a in Central Texas The Texas Agricultural Experiment Station / Charles J. Arntzen, Director The Texas A&M University System / College Station, Texas g [Blank Page in Originfl Bulletin] ‘ v1. § 1 .25’ ' a‘; a .55. m ' '3' '5‘:- Productivity of 10 Warm-Season Perennial Grasses Over Several Years in Central Texas Matt A. Sanderson, Ronald M. Jones, and James S. Newman Assistant Professor, Research Scientist and Resident Director of Research, respectively, Texas A&M Agricul- tural Research and Extension Center, Rt. 2 Box O0, Stephenville, TX 76401. \ [Blank Page in Origiafi Bulletin] V ‘a. 2‘ , .1.‘ ‘ , ,.,. 1 Paiv‘ ‘M: ABSTRACT Perennial warm-season grasses provide valu- able forage during summer months in the southern Great Plains. Objectives of two 5-year studies were to compare 10 warm-season grasses for productivity and response to nitrogen (N) fertilizer over many en- vironments. In Experiment 1, the yield response of ‘Morpa’ and ‘Renner’ lovegrass [Eragrostis curvula (Schrad.) Nees], ‘Selection-75’ kleingrass (Panicum coloratum L.), buffelgrass (Cenchrus ciliaris L., strain 18-35) and ‘Coastal’, ‘Coastcross-l’, and ‘Alecia’ ber- mudagrass [Cynodon dactylon (L.) Pers.] to four levels of N (0 to 500 lb per acre) was measured during 1972 and from 1974 to 1977. Plots were harvested three or four times each year. In Experiment 2, Selection-75 kleingrass, ‘Llano’ buffelgrass, PMT- 587 b1uestem(Dicanthium spp.), ‘Pretoria-90’ bluestem (Dicanthium annulatum Stapf.) and caucasian bluestem [Bothriochloa caucasica (Trin.) C.E. Hubb.] were evaluated for herbage yield from 1975 to 1979. Plots were harvested one to three times each year. In Experiment 1, bermudagrasses exhibited the great- est response to applied N and outyielded other spe- cies when rainfall was average or above average and N rate was high. Buffelgrass, kleingrass, and love- grass produced well with limited N and rainfall; however, buffelgrass winterkilled after 3 years. In Experiment 2, Pretoria-90 bluestem produced very high yields from 1975 to 1978 before being severely winter damaged in the winter of 1978-1979. Buffelgrass yield was much reduced by winter injury during the winter of 1976-1977 and suffered complete stand loss the following winter. Caucasian bluestem, klein- grass, and lovegrass produced reliable yields over 5 years. INTRODUCTION Perennial warm-season grasses can meet a sig- nificant forage need during the summer months in the southern Great Plains and are well suited for revegetation of marginal lands. The establishment of the Conservation Reserve Program (CRP) in the 1985 Food Security Act (P.L. 99-198) has resulted in 30 million acres of highly erodible cropland being set aside, much of which has been planted to warm- season perennial grasses (Heimlich and Kula, 1989). In 1996, the first of the 10-year contracts will expire. This has prompted natural resource managers and farmers to consider the fate of these grasslands (Goetz, 1989; Heimlich and Kula, 1989). Thus, it is imperative to gather management information so that CRP lands can be integrated successfully into existing range and pasture land when the program expires. Producers desire information on persistence, productivity, nitrogen (N) requirements, and selec- tion of species or varieties of warm-season grasses to design effective forage production systems. Revenues from forage-rangeland animal pro- duction systems comprise more than one-half of the annual $12 billion of agricultural income in Texas in alwilrilaalfi- Wm ‘jan- 1989 (Anderson and Summerour, 1990). Livestock operations located in the 20- to 30-inch rainfall zone of the United States often combine rangeland with cropland and tame pastures. Much of the tame pasture and hayland areas are planted to warm- season perennial grasses. The objectives of the 5-year studies reported here were to: (1) compare warm-season grasses for their productivity over several years in this region, and (2) to determine the response to N fertilizer over several years under the temperature and rainfall conditions of central Texas. MATERIALS AND METHODS Experiment 1 The response of six warm-season grasses to four levels of N was evaluated during 1972, and 1974 to 1977. Strain 18-35 buffelgrass, Morpa lovegrass, Selection-75 kleingrass, and Coastal, Coastcross-1, and Alecia bermudagrass were planted in a Windthorst fine sandy loam (fine, mixed, thermic, Udic Paleu- stalf) soil at Stephenville, Texas, in spring of 1972. Renner lovegrass was seeded in spring of 1973. Lovegrass, buffelgrass, and kleingrass were seeded at 2 lb of pure live seed (PLS) per acre in rows 3 ft apart. Bermudagrasses were sprigged in rows 3 ft apart. Species whole plots were 17 ft by 40 ft in four randomized blocks. Four N levels—none, low, medium, and high— were imposed on subplots of each species. In 1972, N was applied at 50, 100, and 2001b/acre in March and after the first harvest for the low (100 lb), medium (200 lb), and high (400 lb) treatments, respectively. In 1974, 50 lb ofN was applied on 22 March and 25 lb after Harvests 2 and 3 for the low (100 lb) rate. Medium (313 lb) and high (500 lb) rates of N were applied at 62.5 and 100 lb/acre, respectively, on 22 March and after each of the four harvests. From 1975 to 1977, one-third of the total N was applied in mid- April and after the first and second harvests for a total of 75, 188, and 300 lb N/acre for low, medium, and high treatments. ~ The experimental design was a randomized complete block with a split-plot arrangement of N treatments in four blocks. Grass entries were whole plots and N rates were subplots. Nitrogen effects were partitioned into linear, quadratic, and lack-of- fit terms. Because of different N rates in different years, years were analyzed separately. Experiment 2 Selection-75 kleingrass, Llano buffelgrass, PMT- 587 bluestem, Pretoria-90 bluestem, and caucasian bluestem were evaluated for herbage yield for 5 years. Plants of kleingrass, buffelgrass, and Preto- ria-90 bluestem were started in pots in a glasshouse and then transplanted on 2-ft centers into 6 ft by 12 ft plots (12 plants per plot) on 27 May and 2 June, 1975. PMT-587 and caucasian bluestem were broad- cast seeded at 2 lb PLS per acre in early June, 1975. All plots received 0.75 inches of water by solid-set irrigation sprinklers on 1 June, 30 June, 11 July, and 7 Aug., 1975 to ensure a stand. Rates of N, P205, and K20 applied to plots were 50-40-0 lb/acre, respec- tively, in 1975, 150-80-160 in 1976, 300-40-0 in 1977, 200-80-160 in 1978, and 150-0-0 in 1979. Herbage was cut 2 inches aboveground from a 3 ft by 12 ft section in each plot, weighed, and a subsample dried at 158°F to determine dry matter (DM) yield. Plants were harvested at boot or early flower stage. One harvest was made in 1975 and 1979, three in 1976 and 1977, and two in 1978. In 1976 and 1977, herbage samples from each harvest of each species were analyzed for total N concentration via standard macro-Kjeldahl proce- dures. Crude protein was calculated as total N x 6.25. The experimental design was a randomized complete block with four blocks. Grass entry differ- ences were determined by a protected least signifi- cant difference test. Years were analyzed separately because of different entries in each year. RESULTS AND DISCUSSION Experiment 1 Entry differences Analysis of variance indicated significant spe- cies, N level, and species x N level effects on DM yield in each year except 1977 when species were not sig- nificantly different in yield (Table 3). Yield of all entries varied with year, primarily because of cli- matic differences (Tables 1 and 2). Buffelgrass yields declined progressively from 1972 to 1975 because of cumulative winter injury, and died out completely by 1976. Buffelgrass is not adapted to winter in central and north central Texas (Holt and Bashaw, 1976). Coastal bermudagrass outyielded other bermu- dagrass cultivars except in 1977 when no differences among cultivars were observed. Alecia bermudagrass outyielded Coastcross-1 4 out of 5 years. Holt et al. (1978) also reported that Coastal bermudagrass gen- erally outyielded other bermudagrass cultivars. Kleingrass outyielded all entries in 1976 when rain- Table 1. Rainfall (inches) during spring (March to May), summer (June to Aug.), tall (Sept. to Nov.), and winter (Dec. to Feb.) from 1972 to 1979 at Stephenville, Texas, compared with 40-year average. Season Average 1972 1974 1975 1976 1977 1978 1979 Spring 9.7 a1 2.4 7.4 11.5 13.3 5.9161 Summer as 4.9 a.o 3311.4 4.1 3.7 7.7 Fall 8.0 6.2 14.8 2.9 10.0 2.9 8.1 2.8 Winter 5.0 1.9 3.0 4.1 2.8 2.2 3.4 8.3 Total 29.5 21.1 28.2 18.2 35.7 22.6 21.1 34.9 Table 2. Average daily temperature (°F) during spring (March to May), summer (June to August), tall (Sept. to Nov.), and winter (Dec. to Feb.) from 1972 to 1979 at Stephenville, Texas, com- pared with 40-year average. Season Average 1972 1974 1975 1976 1977 1978 1979 Spring 63.0 67.6 68.2 61.9 62.6 56.8 62.9 61.5 Summer 81.2 79.7 80.4 81.9 79.3 82.7 81.9 78.0 Fall 65.2 65.1 64.7 64.7 61.8 68.4 64.4 63.8 Winter 45.9 47.9 47.7 43.9 49.8 44.4 38.6 39.0 Table 3. Analysis oi variance and significance oi F values for dry matter yields from 1972 to 1977 in Experiment 1. Source of Significance level‘ variation dt“ 1972 1974 1975 1976 1977 species (s) 5 ... ... ... ... NS N rate (NR) 3 ... ... ... ... ... 1 it! tit iii it! tit Quadratic 1 NS NS NS NS "' Lack-of-fit 1 NS "' "' ' "' S x NR 15 ... ... ... ... ... S X Linear 5 .. ... ... .. .. S x Quad 5 NS NS "' NS NS S x LOF 5 NS " NS NS NS CV % 21.5 20.3 17.6 20.0 17.0 *"' P<0.001; " P<0.01; ' P<0.05; NS = not significant. "Degrees of freedom for 1974 were, S .—. 6, NR = 3, S X NR = 18 S X linear = 6, S X quadratic = 6, S X lack-of-fit (LOF) = 6. fall was much above average. Renner lovegrass generally outyielded Morpa lovegrass in each year. Entry response to N Species DM yield response to N level was linear in each year except in 1975 when there was a quad- ratic response by Renner and Morpa lovegrass (Tables 3, 4, 5, and 6). The species x linear and species x quadratic interaction indicated that species differed in their yield response to N application. Buffelgrass, kleingrass, and Morpa lovegrass did not show a yield response to N in 1972, a dry year (Tables 1 and 5). Bunchgrasses increased in N responsiveness with age of stand with peak N re- sponse occurring in 1976 when excess rainfall oc- curred during the entire growing season (Tables 1 and 5). Although 1974 was near normal in total rainfall, spring rainfall was much reduced and may have limited bunchgrass and bermudagrass response to N. In 1975, however, adequate spring rain may have allowed an early N response even though total rainfall was 11 inches below normal. Generally, the bermudagrasses showed the great- est response (larger regression slope) to applied N in each year with the greatest response occurring dur- ing years of average or above average rainfall (e.g., 1976). The bermudagrasses each responded to in- creased N application with Coastcross-1 and Alecia having a greater response (average slope = 16.98, 22.79, and 21.31 lb ofDM/lb of N applied for Coastal, Coastcross-1, and Alecia, respectively, over 5 years). The bunchgrasses typically yielded more at zero- N (larger regression intercepts) than did bermu- Table 5. Flegression equations describing response of bunchgrasses to applied nitrogen in Experiment 1. Equation form is Yield = A + b (Nitrogen rate)‘. r’ = coefficient of simple or multiple determination. REMS=root error mean square. P=probability level. Table 4. Total season dry matter yields of bunchgrasses and bermudagrasses in response to nitrogen fertilizer over 5 years at Stephenville, Texas. Experiment 1. Nitrogen Rate BG KG MLG RLG CBG ABG CCBG* Dry matter yield (lb/acre) (lb/acre) 1972 0 4425 4715 3003 3547 2994 1454 100 4614 5319 3157 4956 3629 3866 200 4810 5282 31 46 7646 6338 4699 400 4582 5932 3445 8813 6207 541 5 Mean 4608 5312 3188 6240 4792 3858 1974 0 513 1 1 27 1635 1894 721 272 446 100 1472 1610 2374 2909 1573 992 1460 313 3253 2512 2192 3221 4239 4134 3858 500 3316 2528 2644 3589 5390 5431 4032 Mean 2138 1944 2211 2903 2981 2707 2449 1975 0 656 1564 2122 3191 645 371 456 75 919 2257 3671 561 2 2062 1934 2023 188 2709 4423 6351 8325 5912 6069 4837 300 3138 5919 6645 7966 9370 7462 61 16 Mean 1856 3540 4697 6273 4497 3959 3358 1976 0 3171 2020 3485 2013 1251 1499 75 6910 4986 6716 4441 3625 5609 188 12519 6264 9072 10598 8255 8899 300 12971 7130 11153 12758 12185 11615 Mean 8893 5100 7606 7452 6329 6905 1977 0 1593 2834 1955 842 728 1282 75 3382 3057 3566 3014 3091 4104 188 5845 5104 5202 6364 6254 6074 300 5620 5909 5706 7537 7996 6275 Nban 4110 4107 4226 4439 4517 4434 *BG = Strain 18-35 bufielgrass; KG = Selection-75 kleingrass; MLG = Morpa lovegrass; RLG = Renner lovegrass; CBG = Coastal ber- mudagrass; ABG = Alecia bermudagrass; CCBG = Coastcross-1 bermudagrass. Year Equation r’ REMS P Buffelgrass 1972 4524+ 0.44 x 0.01 602 NS 1974 810+ 5.82 x 0.76 683 0.001 1975 563+ 9.18 x 0.59 928 0.005 Kleingrass 1972 4820+ 2.62 x 0.12 1126 NS 1974 1279+ 2.92 x 0.53 568 0.001 1975 1410+ 15.14 x 0.85 777 0.001 1976 4113+ 33.96 x 0.78 2202 0.001 1977 2139+ 14.00 x 0.79 867 0.001 Morpa lovegrass 1972 3005+ 0.97 X 0.05 675 NS 1974 1866+ 1.51 X 0.29 489 0.03 1975 1956+ 31.68 X +0.05 X2 0.87 796 0.025 1976 2870+ 15.84 X 0.75 1097 0.001 1977 2641 +11.26 X 0.69 924 0.001 Renner lovegrass 1974 2228+ 2.96 X 0.55 555 0.001 1975 3076+ 44.14 X -0.09 X2 0.83 1033 0.005 1976 4143+ 24.60 X 0.78 1591 0.001 1977 2352+ 12.47X 0.81 729 0.001 ‘The parameter “A” is the intercept of the equation and estimates the average yield response to zero N. The parameter “b" is the slope of the line or the change in yield per unit of N applied. dagrasses. During 1975, a very dry year, lovegrass and kleingrass yields at zero-N were two to three times greater than those of buffelgrass and bermu- dagrass. Lovegrasses outyielded other grasses in 1974, 1975, and 1976 at low N inputs (75 to 100 lb/ acre). Only at N fertilizer levels greater than 200- to 300-lb/acre did bermudagrass outyield bunchgrasses. Renner lovegrass responded more to applied N than did Morpa in 1974, 1976, and 1977. The yield response of kleingrass in 1976, a wet year, was nearly as great as bermudagrass responses. In our study, environment and species as well as cultivar influenced yield-N relationships in warm- season grasses. Read (1982) reported no differences among three warm-season bunchgrasses [Selection- 75 kleingrass, plains bluestem (B. ischaemum L. King)] and ‘Alamo’ switchgrass (Panicum virgatum L.) in response to various N levels. Jung et al. (1990), however, reported year, cultivar, and N rate interac- Table 6. Regression equations describing response of three bermudagrasses to applied nitrogen in Experiment 1. Equation form is Yield = A + b (Nitrogen rate). r2 = coefficient oi simple or multiple determination. REMS = root error mean square. P = probability level. Year Equation r’ FiEMS P Coastal 1972 2158+ 9.07 x 0.68 1017 0.001 1974 719+ 7.58 x 0.88 589 0.001 1975 622+ 19.43 x 0.92 712 0.001 1976 2306+ 32.68 x 0.77 2160 0.001 1977 2158 +16.16 x 0.70 1280 0.002 . Coastcross-1 1972 3686+ 13.62 x 0.62 1712 0.003 1974 757+ 9.74 x 0.91 647 o.oo1 1975 293+ 29.87 x 0.94 929 0.001 1976 2127+37.84 x 0.89 1613 0.001 1977 1217+ 22.89 x 0.86 1145 0.001 Alecia 1972 3085+ 9.10 x 0.48 1549 0.0031 1974 210+ 10.94 x 0.91 719 0.001 1975 433+ 25.05 x 0.93 857 0.001 1976 1115+37.04 x 0.95 1024 0.001 1977 1080+ 24.42 x 0.93 820 0.001 tions in various warm-season bunchgrasses in the Northeast. They grouped warm-season grasses into three categories: (i) those that responded early to applied N, (ii) those that showed a delayed response (2 to 3 years), and (iii) those with little or no response. Thus, the warm-season bunchgrasses used in our study would fit in Category 2, delayed response types, because the greatest N response occured after 2 to 3 years, whereas bermudagrasses, which re- sponded in each year, would fit in Category 1. Kleingrass and lovegrass should be excellent grasses to grow in dry areas with limited N fertilizer inputs. Our data indicate that the N requirements of kleingrass and lovegrass could be met by including an adapted warm-season legume with moderate or high N-fixing potential at low stands or establishing a good stand of a legume with a low N-fixing poten- tial. Thus, legume and grass management to enable such a legume stand to persist may not be critical. Experiment 2 There were no differences in DM yield among species in 1975 (Fig. 1a) probably because of limited rainfall (Table 1). The greatest yields occurred in 1976 when rainfall was 6 inches above normal. Pretoria-90 bluestem outyielded other species from 1976 to 1978 (Fig. 1b-d). Buflelgrass yields were very high in 1976, then declined greatly in 1977. Buffelgrass eventually died out during the winter of 1977-1978. y‘ Pretoria-9O bluestem was weakened during the winter of 1978-1979 as evidenced by the low yield in 1979 confirming its lack of winterhardiness in northcen- tral Texas. There were 1, 9, and 8 days in the winters of 1975-1976, 1976-1977, and 1977-1978, respec- tively, during which minimum temperatures of 7 to 15°F were recorded. Caucasian and PMT-587 bluestems and klein- grass yields were similar during 1976 to 1978. In 1979, caucasian bluestem yielded the most, whereas PMT-587 bluestem and kleingrass were similar in yield. Yield for each species was distributed simi- larly throughout each season (data not shown). Most production occurred in late spring to early summer. Yields were least in summer and moderate in late summer to early fall. Pretoria-90 bluestem yield was distributed mostly toward the fall. Although stand density data were not collected, yield data and visual observations made at the end of the trial indicated little loss of stand among surviving bunchgrasses. Kleingrass and buffelgrass had the highest crude protein concentration in 1976, whereas there was no clear pattern of crude protein differences among species in 1977 (Table 7). Lowest crude protein concentrations occurred in the last harvest, likely because the forage was more mature. Table 7. Crude protein concentrations (percent ofdry matter) of six warm-season perennial grasses at three harvest dates in 1976 and 1977. Data are the means oi four replicates. Experi- ment 2. 1976 1977 Species 1 2 3 1 2 3 Caucasian bluestem 9.3 9.3 6.1 9.7 7.6 5.8 Buffelgrass 11.1 11.6 8.0 8.6 8.4 PMT-587 bluestem 7.8 7.9 6.0 11.2 9.1 6.8 Pretoria-90 bluestem 7.8 7.9 6.0 9.9 7.0 5.5 Kleingrass 11.0 10.9 6.4 10.6 11.2 7.8 L.S.D‘ 1.6 1.6 1.1 0.9 1.3 0.8 ‘Least significant difference at the 0.05 probability level. INTERPRETIVE SUMMARY Federal conservation programs mandated by the 1985 Food Security Act have as their goal removing highly erodible land from cultivated crop production and changing its use to include permanent cover. Perennial grasses requjringlimited inputs work well as permanent cover and, in the southern Great Plains, warm-season grasses would work well be- cause they are adapted to the hot, dry summers typical of the region. Ten warm-season grasses were “" evaluated in two 5-year experiments in our study. L___._.____._______ vfl W" _ W _ 1975 YIELDS DRY MATTER YlELDIlbI CBS BG PMT-587 P-90 BS KG SPECIES I I HARVEST1 I Figure 1a 1977 YIELDS DRY MATTER YIELD (1000 lb) LSD -1991 — css BG PMT-587 P-90 as KG SPECIES I I HARVEST 1 V////¢ HARVEST 2 [:1 HARVEST 3 Figure 1c Figures 1a - e. Yields of caucasian bluestem (CBS), buffelgrass (BG), PMT-587 bluestem (PMT-587), Pretoria-90 bluestem (P-90 BS), and kleingrass (KG) over 5 years at Stephenville, Texas. Experiment 2. L.S.D. = least significant difference at the 0.05 probability level. NS = not significant. Buffelgrass winterkilled during 1977-1978 1976 YIELDS DRY MATTER YIELD (1000 lbI 20 " Ls0— 250s 16.455 1 5.576 PMT-587 P-90 BS KG SPECIES CBS BG I I HARVEST 1 V/fl HARVEST 2 C} HARVEST 3 Figure 1b 1978 YIELDS DRY MATTER YIELD (1000 lbI LSD = 1596 % /// % CBS BG PMT-587 P-90 as KG SPECIES I - HARVEST 1 V/fl HARVEST 2 Figure 1d 1979 YIELDS DRY MATTER YIELD (1000 lb) LSD=624 CB BG PMT-587 P-90 as KG SPECIES 1 HARVEST 1 i Figure 1e Bermudagrasses were most productive under high N input and average or above average years of rainfall. Buffelgrass and Pretoria-90 bluestem produced high yields but winterkilled after 3 to 4 years. Kleingrass, lovegrass, caucasian bluestem, and PMT-587 bluestem produced well under low N inputs and dry years. These grasses should provide permanent cover and reliable forage during the summer months on highly erodible lands with low inputs. Nitrogen inputs could be further reduced by the inclusion of an adapted, warm-season legume. Because relatively small amounts of N were required for yield responses, either a low stand of a moderate to high N-fixing legume or a good stand of a poor N -fixing legume could provide the N needed. The excellent persis- tence of some of the bunchgrasses coupled with their low N requirement provide additional benefits in potentially reducing nonpoint source pollution. Additionally, species such as kleingrass provide excellent habitat and food (shattered seed) for wild- life. ‘ REFERENCES Anderson, C.G., and B. Summerour. 1990. Estimated value of agricultural production and related items, Texas, 1986- 1989. Texas Agricultural Extension Service, College Station. Goetz, H. 1989. The Conservation Reserve Program—Where are we heading? Rangelands 11:251-252. Heimlich, R.E., and O.E. Kula. 1989. Grazing lands: How much CRP land will remain in grass? Rangelands 11:253-256. Holt, E.C., and E.C. Bashaw. 1976. Developing improved grasses and legumes. pp. 7-9. In: E.C. Holt and R.D. Lewis (ed.) Grasses and legumes in Texas: Development, production, and utilization. Texas Agric. Exp. Stn. Res. Monograph 6C. Holt, E.C., B.E. Conrad, W.C. Ellis, RM. Jones, D.A. Lovelace, and Q W.J. Norris. 1978. Bemudagrass research in central and g north Texas. Texas Agric. Exp. Stn. Prog. Rpt. 3482. Jung, G.A, J.A. Schaffer, W.L. Stout, and M.T. Panciera. 1990. Warm-season grass diversity in yield, plant morphology, and nitrogen concentration and removal in Northeastern USA. Agron. J . 82:21-26. Read, J .C. 1982. Response of kleingrass, switchgrass, and plains bluestem to nitrogen fertility. Agron. Abstr. p. 127. ‘w. [Blank Page in Original Bulletin] ' ‘k4 Mention of a trademark or a proprietary product does not constitute a guarantee or a warranty of the product by The Texas Agricultural Experiment Station and does not imply its approval to the exclusion of other products that also may be suitable. All programs and information of The Texas Agricultural Experiment Station are available to everyone without regard to race, color, religion, sex, age, national origin, or handicap. 1K 1-91 a