Five Decades of Alfalfa Cultivar Improvement

Published online February 24, 2006

Five Decades of Alfalfa Cultivar Improvement: Impact on Forage Yield, Persistence, and Nutritive Value JoAnn F. S. Lamb,* Craig C. Sheaffer, Landon H. Rhodes, R. Mark Sulc, Daniel J. Undersander, and E. Charles Brummer

J.F.S. Lamb, USDA-ARS Plant Science Research Unit and Dep. of Agronomy and Plant Genetics, and C.C. Sheaffer, Dep. of Agronomy and Plant Genetics, Univ. of Minnesota, 411 Borlaug Hall, 1991 Upper Buford Circle, Saint Paul, MN 55108; L.H. Rhodes, Dep. of Plant Pathology, and R.M. Sulc, Dep. of Horticulture and Crop Sci., Ohio State Univ., 2021 Coffey Road, Columbus, OH 43210-1086; D.J. Undersander, Dep. of Agronomy, Univ. of Wisconsin-Madison, 1575 Linden Drive, Madison, WI 53706; E.C. Brummer, Raymond F. Baker Center for Plant Breeding, Dep. of Agronomy, Iowa State Univ., 1204 Agronomy Hall, Ames, IA 50010. *Corresponding author ([email protected]).

(1988) used data reported by the CAIC and determined that average forage yield for improved alfalfa cultivars released between 1971 and 1988 was 9% greater than the yield of Vernal. This contrasted with a report by these same authors (Hill and Rosenberger, 1985), using a different set of data, that attributed a yearly yield increase to cultivars released between 1971 and 1981 to ,1%. Other researchers reported a 3% gain in forage yield between the mid-1950s and the early 1970s (Hill and Kalton, 1976; Elliott et al., 1972). Holland and Bingham (1994) grew alfalfas from three eras: 1898 to 1910, 1940 to 1953, and 1979 to 1985 at one location in Wisconsin. They concluded that mean forage yields increased between each era with a greater rate of increase in yield occurring between the second and third era than between the first and second era. However, data reported in their research showed that forage yield from some individual cultivars from the second and third eras were similar. In contrast to the aforementioned results, Wiersma et al. (1997) used CAIC yield data to report a slight but statistically significant decline in forage yield for cultivars released between 1978 and 1996. Volenec et al. (2002) used an updated version of the database created by Wiersma et al. (1997) to compare first harvest yields of alfalfas released between 1907 and 1998. They evaluated both the highest and average yield of landraces or cultivars released in each year and reported no improvement or a small decline in forage yield for more recently released cultivars. Conflicting reports on changes in forage yield of cultivars released during the past century cause speculation on the reality of yield improvement in alfalfa. Yield improvement in forage crops during the past century has lagged far behind that of annual grain crops (Brummer, 1999). Some factors contributing to the slower rate of progress for alfalfa include the dynamics related to survival and overwintering of a perennial growth habit; multiple years of evaluation before selection decisions can be made; and inability to repartition plant assimilates to maximize harvest index, as has been done in annual grain crops, since the entire aboveground portion of the alfalfa plant is harvested (Hill et al., 1988). Another potential reason for reduced yield improvements in alfalfa has been the major focus of most U.S. alfalfa breeding programs on disease and pest resistance. Breeding efforts during the past five decades have resulted in nearly all currently marketed alfalfa cultivars having high levels

Published in Crop Sci. 46:902–909 (2006). Crop Breeding, Genetics & Cytology doi:10.2135/cropsci2005.08-0236 ª Crop Science Society of America 677 S. Segoe Rd., Madison, WI 53711 USA

Abbreviations: CAIC, Central Alfalfa Improvement Conference; CP, crude protein; IVTD, in vitro true digestibility; NDF, neutral detergent fiber; NDFD, neutral detergent fiber digestibility; NIRS, near infrared reflectance spectroscopy.

Reproduced from Crop Science. Published by Crop Science Society of America. All copyrights reserved.

ABSTRACT Previous research has implied that forage yield in released alfalfa (Medicago sativa L.) cultivars declined slightly between 1978 and 1996. Our objective was to compare alfalfa cultivars released during the past five decades side by side in replicated yield trials to test for any changes in forage yield across time. Ten cultivars, two from each of the five decades, four recently released cultivars, and two check cultivars were compared for forage yield, persistence, and nutritive value at four locations. Cultivars were established in May 1999 at Iowa, Wisconsin, Ohio, and Minnesota. Forage was harvested three to four times in each of four production years depending on location. Plots were subsampled for nutritive value analyses for the first and third harvests in 2000 and 2001. Year 3 location 3 cultivar-release date interactions demonstrated that forage yield and final stand densities differed among the cultivars in each year of the experiment at each location. Nutritive value traits were similar among all cultivars. Evidence for changes in forage yield for cultivars released between 1940 and 1995 was environmentally dependent. In environments where conditions lead to plant stand losses, recently released cultivars with multiple disease resistance had a yield advantage over older cultivars, but in environments where no differences in plant density occurred across time, older cultivars yielded the same as recent cultivars.

P

to genetically improve alfalfa in the USA began in 1901 with the field evaluations of ‘Grimm’ alfalfa, a landrace of alfalfa introduced by immigrants from Germany. Several public breeding programs were initiated at agricultural experiment stations and the USDA-ARS throughout the USA during the first half of the 1900s, which led to the release of ‘Atlantic’ in New Jersey in 1940, ‘Ranger’ in Nebraska in 1942, and ‘Vernal’ in Wisconsin in 1953 (Barnes et al., 1977). Private companies began releasing and marketing cultivars in the 1960s, and by the 1970s the majority of the marketed cultivars were proprietary (Barnes et al., 1988). Starting in 1956, forage yield trial results from across the upper Midwest were published annually by the Central Alfalfa Improvement Conference (CAIC). Hill et al. LANT BREEDING PROGRAMS

902

903

Reproduced from Crop Science. Published by Crop Science Society of America. All copyrights reserved.

LAMB ET AL.: 50 YEARS OF ALFALFA IMPROVEMENT

of resistance to a wide variety of important diseases and insects (AOSCA, 2002). Consequently, resistant cultivars should have superior yields compared with susceptible cultivars when grown in environments that favor pest development. While breeding for pest resistance can play a role in realizing yield potential, it likely does not increase the frequency of genes for yield per se in populations under improvement. A possible explanation for the previously mentioned conflicting reports on changes in yield for new alfalfa cultivars is that yield totals were compiled from trials conducted at numerous locations across several years ignoring germplasm by year (age of the stand) and germplasm 3 environment (location) interactions. Hill and Baylor (1983) reported that response patterns related to age of the stand and disease infestation would occur for total yield in a perennial forage like alfalfa, and that these age and disease resistance effects would not be reflected in the cultivar mean across all locations. Increasing forage yield has once again become a focus for discussion among alfalfa researchers (Brummer, 1999; Volenec et al., 2002). In an attempt evaluate genetic improvement in alfalfa during the past 60 yr, while accounting for genotype 3 environment and age interactions, our objective was to compare the forage yield, and assess persistence and nutritive value of alfalfa cultivars released from 1940 to 1995, side by side in replicated trials for 5 yr at four upper Midwest locations. MATERIALS AND METHODS Selection of Cultivars The CAIC Annual Variety Tests publications were used to identify high-yielding alfalfa cultivars that were typically grown across the upper Midwest during the past five decades. Forage yield results from the mid-1950s, 1960s, 1970s, 1980s, and 1990s were considered at Rosemount, MN; Ames, IA; and Arlington, WI. At least three consecutive years of forage yield averages (excluding the establishment year) were used to choose two cultivars representative of each of the past five decades (Table 1). The cultivars Wintergreen, DK 127, WL 324, and Jade II were also included in the study to represent the most recent cultivars available that had at least three con-

secutive years of forage yield data available by the spring of 1999. Two more cultivars were included as forage yield checks, Vernal released in 1953, and 5454 released in 1991. All cultivars released before 1988 underwent seed increase under cage isolations in the summer of 1998 to alleviate any concerns of seed age on comparisons of these cultivars for forage yield potential. However, one seed increase was not successful and seed of ‘520’ used in this experiment was produced in 1970. Germination tests were conducted and seeding rates for all cultivars were adjusted to equal numbers of pure live seed m22. Untreated seed was used for all entries.

Experimental Design Experiments were established at four locations across the upper Midwest in four replications of a randomized complete block design in May 1999. The four locations were Arlington, WI (438329 N, 898369 W; 800 mm average annual rainfall), on a Plano silt loam (fine-silty, mixed, superactive, mesic Typic Argiudolls); Columbus, OH (398599 N, 838019 W; 1073 mm average annual rainfall), on a Miami silt loam (fine-loamy, mixed, active, mesic Oxyaquic Hapludalfs); Rosemount, MN (448439 N, 938069 W; 744 mm average annual rainfall), on a Tallula silt loam (coarse silty, mixed, superactive, mesic Typic Hapludoll); and Ames, IA, (428599 N, 938559 W, 792 mm average annual rainfall) on a Nicolett loam (fine-loamy, mixed, superactive, mesic Aquic Hapludolls). All cultivars were sown at a rate of 690 live seed m22 in five rows 15 cm apart in 0.9- by 6.1-m plots. A 0.9-m border of alfalfa was established around the experiment. At all locations, soil test levels of P and K were maintained at recommended levels for high yields of perennial forages (Rehm et al., 2001). All plots were scouted and monitored for weeds and insects, and herbicides and insecticides were applied as needed. All plots were mown in July of 1999 and subsequent forage yields were estimated from a 0.9- by 5.2-m swath cut through each plot at a stubble height of 5 to 7.5 cm. At all locations, the target maturity for harvest was bud to early flower, with the last harvest in each growing season taken in early to midSeptember. The numbers of harvests taken annually to evaluate forage yield differed at the four locations. At Iowa, two, three (third harvest in 2000 not recorded because of harvest error), four, four, and four harvests were taken in 1999, 2000, 2001, 2002, and 2003 respectively, for a total of 17 harvests. At Minnesota, one harvest was taken in 1999, three harvests each in 2000 to 2002, and four in 2003, for a total of 14 harvests. One harvest was taken in the establishment year, and four harvests

Table 1. Alfalfa cultivars evaluated and their developer, release dates, decade designation, fall dormancy, and disease ratings.†‡ Cultivar

Developer or marketer

Release date

Decade

FD

FW

AN

BW

PRR

VW

APH

Atlantic Ranger Vernal Cayuga 525 Saranac 520 Blazer 526 Magnum III Garst 645 5454 Wintergreen DK 127 WL 324 Jade II

NJ AES USDA-ARS, NE AES USDA-ARS, WI AES Cornell Univ. Pioneer Hi-Bred Int. Cornell Univ. Pioneer Hi-Bred Int. Land O’ Lakes Pioneer Hi-Bred Int. Dairyland Seed Co. ICI Seeds Pioneer Hi-Bred Int. ABI Alfalfa DEKALB W-L Research NC1 Hybrids

1940 1942 1953 1961 1962 1965 1969 1978 1981 1988 1990 1991 1994 1994 1994 1995

1955 1955 Check 1965 1965 1975 1975 1985 1985 1995 1995 Check n/a n/a n/a n/a

4 3 2 4 3 4 3 3 2 4 3 4 3 3 3 4

–

LR S S S S S LR LR LR MR HR HR HR HR HR R

S

S

– – R – – S – LR – MR R MR HR R R R

– – – – – – – – – MR MR LR R HR HR MR

S MR – R R MR R MR R HR HR HR R HR HR

– R R R R HR HR HR R HR R HR HR HR HR

– S S S S LR MR LR R HR HR HR HR HR HR

† FD, fall dormancy; FW, Fusarium wilt; AN, anthracnose Race 1; BW, bacterial wilt; PRR, Phytophthora root rot; VW, Verticillium wilt; APH, Aphanomyces. ‡ S, susceptible (0–5% of the plants resistant); LR, low resistance (6–14% of the plants resistant); MR, moderate resistance (15–30% of the plants resistant); R, Resistant (31–50% of the plants resistant); HR, high resistance (.50% of the plants resistant); – , not tested.

904

CROP SCIENCE, VOL. 46, MARCH–APRIL 2006

Table 2. Analysis of variance of forage yield across four locations and 5 yr for cultivar-release date.

Reproduced from Crop Science. Published by Crop Science Society of America. All copyrights reserved.

Source Year (Y) Location (L) Y3L Rep (Location) Error A Release date (D) Y3D L3D Y3L3D Error B R2 CV

df

Release date MS

4 3 11 12 45 13 52 39 143 893

3759.80*** 754.44*** 127.92*** 11.95*** 5.35 39.09*** 4.35*** 5.52*** 2.02*** 1.07 0.96 9.43

*** Significant at the 0.001 level of probability.

were taken each in 2000 to 2003, for a total of 17 harvests at Ohio. Two harvests were taken in 1999, and four harvests each year for 2000 to 2002, for a total of 14 harvests at Wisconsin. Hand grab samples (minimum 300 g wet weight) were taken from harvested materials in at least 10% of the plots at each harvest, dried at 558C, reweighed and used to calculate dry matter yields. All yields were recorded on a dry matter basis. Final stand densities of all plots were estimated as percentage ground cover m22 (at Iowa, Minnesota, and Ohio in October 2003) or number of stems m22 (at Wisconsin in October 2002).

Nutritive Value Comparisons Subsamples were taken from the first and third harvests in 2000 and 2001 at Minnesota, Ohio, and Wisconsin to assess any changes in nutritive value among these cultivars released during the past five decades. Plots were hand clipped to a stubble height of 5 cm, in a minimum of three random, nonborder 1-ft2 areas to produce a minimum sample size of 300 g wet weight. These subsamples were dried at 558C in forced-air ovens, weighed, and ground through a 1-mm screen in preparation for nutritive value analysis. Crude protein (CP), neutral detergent fiber (NDF), and in vitro true digestibility (IVTD) were determined via near infrared reflectance spectroscopy (NIRS) analysis (model 6500, NIRSystems, Silver Springs, MD)1 using NIRS equations developed for alfalfa in Minnesota. Equations for NIRS were developed using the software program Calibrate (NIRS 3 version 4.0, Infrasoft International, Port Matilda, PA) with modified partial least squares regression option (Shenk and Westerhaus, 1991a, 1991b). Random subsets of 10 samples were chosen and subjected to conventional chemical analysis for CP (Kjeldahl N 3 6.25), NDF, and IVTD (Goering and Van Soest, 1970); and used as monitoring sets. Predicted values for CP, NDF, and IVTD were adjusted for bias based on conventional analysis results from the monitoring sets. Neutral detergent fiber digestibility (NDFD) for each sample was calculated from NDF and IVTD using equations developed by Hoffman et al. (2001).

Statistical Analyses Analysis of variance was conducted to estimate the effects of year, location, and cultivar or cultivar-release date on forage yield and quality (PROC GLM; SAS Institute, 2001). Loca-

1 Mention of a proprietary product does not constitute a recommendation or warranty of the product by the USDA-ARS or the University of Minnesota and does not imply approval to the exclusion of other suitable products.

tions were considered random and years and cultivars were fixed. Means and standard errors were calculated for forage yield for each cultivar-release date. When cultivars differed for forage yield or final stand score, regression analyses were conducted to describe the response in relation to their release dates (P $ 0.01) (PROC REG; SAS Institute, 2001). Means and LSD (0.05) were calculated for nutritive value traits for each cultivar.

RESULTS AND DISCUSSION Most cultivars had unique release dates, but three of the four cultivars selected to represent the most recent alfalfas available on the market (Wintergreen, DK 127, and WL 324) were released in the same year, 1994. Analyses of variance among all 16 cultivars showed that these three cultivars essentially yielded the same in all environments (data not shown). Therefore, all three cultivars represented the 1994 release date in the ANOVAs and regression analyses. Years, locations, and cultivar-release date and all twoand three-way interactions among these main effects impacted forage yield (Table 2). The year 3 location 3 cultivar-release date interaction demonstrated that forage yields among the cultivars differed in each year within each location. Therefore, the ANOVA for forage yield among the cultivar-release dates was conducted separately for each year within each location (Table 3). Regression analyses of forage yield vs. cultivar-release date were conducted for those years within each location where the cultivars differed. No difference in forage yield was detected among the cultivars in the establishment year (1999) at any of the four locations (Table 3 and Fig. 1). From the first production year through the final year of the experiment, the annual forage yield response among the cultivars was different at each location. For all three production years at Wisconsin (2000 to 2002), cultivars differed in forage yield (Table 3). Regression analyses showed an increase in forage yield for more recently

Table 3. Analysis of variance for annual forage yields at each of the four locations for cultivar-release date. Mean Squares Source Wisconsin Rep Cultivar-release Error Ohio Rep Cultivar-release Error Minnesota Rep Cultivar-release Error Iowa Rep Cultivar-release Error

df

1999

2000

2001

2002

date

3 13 47

0.1 0.4 0.3

1.9 3.8*** 0.3

1.8 7.3*** 0.8

3.6** 22.6*** 1.1

n/a n/a n/a

date

3 13 47

1.1 0.6 0.5

6.8** 2.8* 1.3

5.1** 9.2*** 1.0

35.2*** 13.6*** 0.8

13.3*** 13.2*** 0.9

date

3 13 47

0.1 0.1 0.1

0.2 0.6* 0.3

3.4*** 1.9*** 0.3

6.6*** 0.2 1.0

9.6*** 4.3** 1.3

date

3 13 47

0.3 0.4 0.5

0.1 1.5 1.1

1.2 5.6 4.1

0.2 1.9 1.2

3.6 3.0 3.2

* Significant at the 0.05 level of probability. ** Significant at the 0.01 level of probability. *** Significant at the 0.001 level of probability.

2003

905

Reproduced from Crop Science. Published by Crop Science Society of America. All copyrights reserved.

LAMB ET AL.: 50 YEARS OF ALFALFA IMPROVEMENT

Fig. 1. Regression analyses of annual forage yield vs. cultivar-release date at (A) Wisconsin, (B) Ohio, (C) Minnesota, and (D) Iowa. Regressions were calculated only for years where forage yield differed among the cultivars.

Reproduced from Crop Science. Published by Crop Science Society of America. All copyrights reserved.

906

CROP SCIENCE, VOL. 46, MARCH–APRIL 2006

released cultivars (Fig. 1A). The slope of the line increased and the equations explained more variation in forage yield among the cultivars with each successive year. By 2002, .60% of the variability in forage yield among the cultivars was explained by cultivar-release date. In the first production year, Ranger (1942) and ‘Cayuga’ (1961) had the lowest forage yields, followed by Atlantic (1940), Vernal (1953), and 525 (1962). The highest-yielding group included cultivars released in the 1990s, as well as ‘Saranac’ (1965), ‘526’ (1981), and ‘Magnum III’ (1988). For the last two production years at Wisconsin, the highest-yielding cultivars were released in either the late 1980s or in the 1990s. The lowestyielding cultivars in both years were 525 (1962), Atlantic (1940), Saranac (1965), Ranger (1942), and Vernal (1953). The next two lowest-yielding cultivars in both years were Cayuga (1961) and Blazer (1978). The positive linear relationship between forage yield and cultivarrelease date at Wisconsin was strongly impacted by the fact that the five oldest cultivars were among the lowestyielding cultivars for all three production years. In the first production year (2000) at Ohio, differences in forage yield among the cultivars were identified (Table 3). Regression analysis indicated that relationship between forage yield and cultivar-release date was small and the equation explained only a small portion of the variation among the cultivars (Fig. 1B). Blazer (1978) had the highest yield, while Ranger (1942) and Cayuga (1961) were lower in forage yield compared with the rest of the cultivars. In 2001, 2002, and 2003, regression equations explained more of the variation in forage yield among the cultivars with each successive year of the experiment. By 2003, more than half of the

variation in forage yield among the cultivars was explained by cultivar-release date. Starting in 2001 and continuing through 2003, 5454 (1991) was consistently the highest in forage yield, while Atlantic (1940), Ranger (1942), and Saranac (1965) were lowest. The remaining cultivars shifted in rank, with several yielding the same as 5454 in some years, but all were greater in forage yield compared with the low-yielding cultivar group. In the first production year (2000) at Minnesota, cultivars differed in forage yield (Table 3). Regression analysis demonstrated that the relationship between forage yield and cultivar-release date was minimal (Fig. 1C). Vernal (1953) and 525 (1962) yielded less than the other cultivars in the first production year. The older cultivars released in the 1940s yielded the same as cultivars released during the latter three decades. A slight increase in forage yield for more recently released cultivars occurred again in the second production year (2001). In 2001, 520 (1969) had the highest yield, while Atlantic (1940), 525 (1962), and Vernal (1953) were lower in forage yield compared with the rest of the cultivars. In 2002, no differences in forage yield among the cultivars occurred. Difference in forage yield was found among the cultivars in 2003, but no relationship with release date was demonstrated. In the final 2 yr of the experiment, cultivar-release date appeared to have little influence on forage yield at Minnesota. At Ames, no differences in forage yield among cultivars released between 1940 through 1995 were shown (Table 3 and Fig. 1D). The oldest cultivars, Atlantic (1940) and Ranger (1942), had comparable forage yields to the rest of the cultivars, including Jade II (1995), in every year of the experiment.

Fig. 2. Regression analysis of final stand estimates vs. cultivar-release date at all four locations. Regressions were calculated only for locations where final stand densities differed among the cultivars.

907

LAMB ET AL.: 50 YEARS OF ALFALFA IMPROVEMENT

Table 4. Analysis of variance for crude protein (CP), in vitro true digestibility (IVTD), neutral detergent fiber (NDF), and neutral detergent fiber digestibility (NDFD), for spring and summer harvests across 2 yr and three locations. MS CP

Reproduced from Crop Science. Published by Crop Science Society of America. All copyrights reserved.

Source Year (Y) Location (L) Rep (Loc) Y3L Error A Release date (D) Y3D L3D Y3L3D Error B

IVTD

NDF

NDFD

df

Harvest 1

Harvest 3

Harvest 1

Harvest 3

Harvest 1

Harvest 3

Harvest 1

Harvest 3

1 2 9 2 9 13 13 26 26 282

2 820* 1 129 218 27 913*** 550 236** 50 178** 121 97

5 804*** 18 910*** 362 634 156 316*** 126 136 100 112

160 998*** 134 006*** 554 180 511*** 2 543 391 218 365 495 356

3 404 241 135*** 1 556 32 856*** 1 022 783 458 424 477 465

90 061*** 96 194*** 511 173 079*** 1761 413 213 418 450 313

1 585 217 802*** 1 356 18 562*** 893 709* 438 452 381 400

201 361*** 151 213*** 537 130 894*** 3 160 740* 192 323 448 352

50 226*** 214 833*** 1 470 51 514*** 932 1 100** 397 392 525 453

* Significant at the 0.05 level of probability. ** Significant at the 0.01 level of probability. *** Significant at the 0.001 level of probability.

Final Plant Stand Density Final stand density estimates of the cultivars were different at each of the four locations (Fig. 2). As with forage yield, final stand density was similar (average 5 90%) for all cultivars at Iowa. Stand survival among the cultivars was different at the three remaining locations. At Minnesota, final stand density ranged from 73 to 95% ground cover with Atlantic (1940), Saranac (1965), 525 (1962), Blazer (1978), and 520 (1969) having the greatest stand loss. Ranger (1942), Vernal (1953), Cayuga (1961), and Magnum III (1988) were intermediate, while 526 (1981) and all cultivars released between 1990 and 1995 had the greatest stand survival. The range in final stand densities among the cultivars at Ohio was much wider (33–81%) compared with Minnesota. Saranac (1965), Atlantic (1940), and Ranger (1942) had the most profound plant stand losses. The remaining cultivars released between 1953 and 1981 were intermediate in plant stand, and all cultivars released in 1988 or later had the greatest plant stand density. At Wisconsin, final stand estimates ranged from 277 to 519 stems m22. DK 127

(1994), Wintergreen (1994), ‘Garst 645’ (1990), WL 324 (1994), and Magnum III (1988) had the highest stem counts, while Saranac (1965), 525 (1962), Ranger (1942), and Atlantic (1940) had the lowest stem counts.

Forage Nutritive Value The environmental effects of year and location, and the interactions between year and location accounted for the major portion of variation among the cultivars for the four nutritive value traits (Table 4.) No relationship between cultivar-release date and CP, IVTD, NDF, or NDFD was demonstrated by regression analyses. A location 3 cultivar-release date interaction was found for CP for the first harvest (Table 4). Differences among the cultivars were found at Ohio and Wisconsin, but not at Minnesota. The range in CP among the cultivars was nearly 50% greater at Wisconsin compared with Ohio (Table 5). DK 127 (1994) had the highest CP at both locations, while Magnum III (1988) had the lowest at Ohio and 5454 (1991) had the lowest at Wisconsin. Differences in CP were also found among the cultivars

Table 5. Means of nutritive value traits by cultivar-release date. CP Harvest 1 Cultivar

Release date

MN

OH

NDFD WI

Harvest 3

NDF Harvest 3 g kg

Atlantic Ranger Vernal Cayuga 525 Saranac 520 Blazer 526 Magnum III Garst 645 5454 Wintergreen DK 127 WL 324 Jade II LSD (0.05) † NS, not significant.

1940 1942 1953 1961 1962 1965 1969 1978 1981 1988 1990 1991 1994 1994 1994 1995

206 217 212 207 205 210 211 207 208 209 214 212 206 202 210 216 NS†

216 216 212 214 216 217 220 212 213 209 218 211 221 226 214 216 8

221 214 212 219 207 223 213 219 214 210 222 204 216 230 223 217 9

219 219 218 218 219 218 216 218 223 212 226 213 220 224 217 216 7

Harvest 1

Harvest 3

339 333 330 327 327 329 330 320 325 322 321 322 317 331 321 323 13

333 326 327 314 332 329 321 324 337 317 336 321 326 329 326 321 15

21

418 422 423 422 421 422 427 423 413 429 409 426 416 413 423 422 14

Reproduced from Crop Science. Published by Crop Science Society of America. All copyrights reserved.

908

CROP SCIENCE, VOL. 46, MARCH–APRIL 2006

for the third cut, but no cultivar was consistently high or low for CP and no trend with cultivar-release date was demonstrated. Cultivars did not differ in IVTD or in NDF in the first (spring) harvest. Cultivars did differ in NDF in the third (late summer) harvest where Garst 645 (1990) had the lowest fiber content and Magnum III (1988), 5454 (1991), and 520 (1969) had the highest (Table 5). Cultivars differed in NDFD in the first and third harvests, but ranking was not consistent across harvests. Sheaffer et al. (1998) also showed differences in nutritive value among alfalfa cultivars released in the 1990s, but the nutritive value of the entries were unstable across environments. Essentially no changes in nutritive value (CP, IVTD, NDF, or NDFD) were found among the cultivars released between 1940 and 1995 in our study.

Implications The environmental factors of year and location had a profound effect on the interpretation of forage yield differences among cultivars released between 1940 and 1995. If this experiment had been conducted only at Iowa, we would conclude that there have been no changes in forage yield among alfalfa cultivars released during the past five-and-a-half decades. The Wisconsin and Ohio data would lead us to conclude that there has been phenomenal improvement in forage yield for more recently released cultivars, and these yield advantages increased with the age of the stand. At Minnesota, a small yield advantage for more recently released cultivars was shown for the first two production years, but as the stand aged, these differences disappeared. These results emulate the conflicting information on alfalfa yield across time that has been reported by others (Hill et al., 1988; Hill and Rosenberger, 1985; Hill and Kalton, 1976; Elliott et al., 1972; Holland and Bingham, 1994; Wiersma et al., 1997; Volenec et al., 2002). Hill and Baylor (1983) also reported large environment by cultivar interactions and questioned the validity of analyzing and comparing alfalfa cultivar yields across years and diverse locations. The conflicting yield responses across locations and years imply that factors other than yield per se influenced our results. Simple correlations between total yield per plot and final stand density per plot were greater at Wisconsin (r 5 0.73, P $ 0.001) and Ohio (r 5 0.65, P $ 0.001) compared with Minnesota (r = 0.48, P $ 0.001) and Iowa (not significant). The disparity in the relationship between yield and final stand density were likely influenced by biotic and environmental stresses unique to each location including differences in soil type, temperature, and moisture regimes proscribed by climate and incidence of disease or pests. At Ohio, the plots were evaluated annually for incidence of disease. At Ohio, Ariss (2005) found significant negative relationships between the total number of diseased stems [primarily Fusarium wilt, Fusarium oxysporum Schlecht f. sp. medicaginis (Weimer) Snyd. and Hans., and anthracnose, Colletotrichum trifolii Bain and Essary] per plot and total yield (r = 20.65, P $ 0.01), final stand

density (r = 20.71, P $ 0.01), and release date (r = –0.65, P $ 0.01). Results at Ohio indicated that recently released cultivars (1988–1995) had fewer diseased stems, denser final stands, and higher forage yield compared with older cultivars. Multiple disease resistance in marketed alfalfa has increased with time as newer cultivars were released (Table 1). Cultivars rated as the most susceptible to disease (Atlantic, Ranger, and Saranac) had the greatest stand losses and the lowest yields at Ohio. We speculated that the increased disease resistance advantage of recently released cultivars improved stand survival and protected yield potential at Ohio. Disease infestation assessments were not conducted at the other three locations. However, at Wisconsin, results were similar to those found at Ohio; cultivars with higher levels of multiple pest resistance (released between 1988 and 1995) had higher yields and greater final stem counts compared with the rest of the cultivars. By fall 2002, stand scores (stems m22) of the more susceptible cultivars were so low that the decision was made to terminate the study at Wisconsin. At Iowa and Minnesota, where differences in plant stand did not occur or were minimal, forage yields were similar for all cultivars regardless of release date. We concluded that evidence for changes in forage yield for cultivars released between 1940 and 1995 was environmentally dependent. In environments where conditions lead to plant stand losses, recently released cultivars with multiple disease resistance had a yield advantage over older cultivars, but in environments where no differences in plant density occurred, older cultivars yielded the same as the improved cultivars. ACKNOWLEDGMENTS The authors would like to thank ABI Alfalfa, Cal/West Seeds, Dairyland Seed Company, Pioneer Hi-Bred International, Inc., and W-L Research, Inc. for providing seed increases of the alfalfa cultivars evaluated in this research.

REFERENCES Ariss, J.J. 2005. Pathological factors affecting persistence in alfalfa with emphasis on diseases caused by Fusarium and Colletotrichum species. Ph.D. thesis. The Ohio State Univ., Columbus. Association of Official Seed Certifying Agencies. 2002. Report of the National Alfalfa and Miscellaneous Legumes Variety Review Board. No. 191. AOCSA, Meridian, ID. Barnes, D.K., E.T. Bingham, R.P. Murphy, O.J. Hunt, D.F. Beard, W.H. Skrdla, and L.R. Teuber. 1977. Alfalfa germplasm in the United States: Genetic vulnerability, use, improvement, and maintenance. Tech. Bull. 1571. USDA, Washington, DC. Barnes, D.K., B.P. Goplen, and J.E. Baylor. 1988. Highlights in the USA and Canada. p. 1–24. In A.A. Hanson et al. (ed.) Alfalfa and alfalfa improvement. Agron. Monogr. 29. ASA, CSSA, and SSSA, Madison, WI. Brummer, E.C. 1999. Capturing heterosis in forage crop cultivar development. Crop Sci. 39:943–954. Elliott, F.C., I.J. Johnson, and M.H. Schonhorst. 1972. Breeding for forage yield and quality. p. 319–333. In C.H. Hanson (ed.) Alfalfa science and technology. Agron. Monogr. 15. ASA, Madison, WI. Goering, H.K., and P.J. Van Soest. 1970. Forage fiber analysis: Apparatus, reagents, procedures and some applications. USDA Agric. Handb. 379. U.S. Gov. Print. Office, Washington, DC. Hill, R.R., Jr., and J.E. Baylor. 1983. Genotype 3 environment interaction analysis for yield in alfalfa. Crop Sci. 23:811–815.

Reproduced from Crop Science. Published by Crop Science Society of America. All copyrights reserved.

LAMB ET AL.: 50 YEARS OF ALFALFA IMPROVEMENT

Hill, R.R., Jr., and R.R. Kalton. 1976. Current philosophies in breeding for yield. p. 51. In Rep. 15th Alfalfa Improvement Conf., Ithaca, NY. 13–15 July 1976. USDA-SEA, Peoria, IL. Hill, R.R., Jr., and J.L. Rosenberger. 1985. Methods of combining data from germplasm evaluation trials. Crop Sci. 25:467–470. Hill, R.R., Jr., J.S. Shenk, and R.F. Barnes. 1988. Breeding for yield and quality. p. 809–825. In A.A. Hanson et al. (ed.) Alfalfa and alfalfa improvement. Agron. Monogr. 29. ASA, CSSA, and SSSA, Madison, WI. Hoffman, P.C., R.D. Shaver, D.K. Combs, D.J. Undersander, L.M. Bauman, and T.K. Seeger. 2001. Understanding NDF digestibility of forages. Focus on forage. Univ. of Wisconsin Ext. Bull. 3:10. Holland, J.B., and E.T. Bingham. 1994. Genetic improvement for yield and fertility of alfalfa cultivars representing different eras of breeding. Crop Sci. 34:953–957. Rehm, G., M. Schmitt, J. Lamb, and R. Eliason. 2001. Fertilizer recommendations for agronomic crops in Minnesota [Online]. Univ. of Minnesota Ext. Serv. BU-06240-S. Available at www.soils.umn. edu/extension/extension_publications.php [accessed 3 Sept. 2004; verified 29 Nov. 2005].

View publication stats

909

Sheaffer, C.C., D. Cash, N.J. Ehlke, J.C. Henning, J.G. Jewett, K.D. Johnson, M.A. Peterson, M. Smith, J.L. Hansen, and D.R. Viands. 1998. Entry 3 environment interactions for alfalfa forage quality. Agron. J. 90:774–780. Shenk, J.S., and M.O. Westerhaus. 1991a. Population definition, sample selection, and calibration procedures for near infrared reflectance spectroscopy. Crop Sci. 31:469–474. Shenk, J.S., and M.O. Westerhaus. 1991b. Population structuring of near infrared spectra and modified least squares regression. Crop Sci. 31:1548–1555. SAS Institute. 2001. The SAS system for Windows. v. 8.02. SAS Inst., Cary, NC. Volenec, J.J., S.M. Cunngingham, D.M. Haagenson, W.K. Berg, B.C. Joern, and D.W. Wiersma. 2002. Physiological genetics of alfalfa improvement: Past failures, future prospects. Field Crops Res. 75: 97–110. Wiersma, D.W., D.J. Undersander, and J.G. Lauer. 1997. Lack of alfalfa yield progress in the Midwest. p. 36. In Central Alfalfa Improvement Conf. Abstracts. LaCrosse, WI. 16–18 July 1997.