|
|
||||||||
a Research Geneticists, USDA-ARS Forage and Range Research Lab., Utah State Univ., Logan, UT 84322-6300
b Supervisory Research Geneticist, and R. Mitchell, Rangeland Scientist, USDA-ARS Grain, Forage, and Bioenergy Research Unit, Univ. Nebraska, Lincoln, NE 68583-0937
c Research Geneticist (Retired), USDA-ARS Northern Great Plains Research Lab., Mandan, ND 58554
d Rangeland Scientist (Retired), USDA-ARS Ft. Keogh Livestock and Range Research Lab., Miles City, MT 59301-4016
e Research Geneticist, USDA-ARS Grazinglands Research Lab., El Reno, OK 73036. Mention of a trademark, proprietary product, or vendor does not constitute a guarantee or warranty of the product by the USDA
* Corresponding author (joseph.robins{at}usu.edu).
| ABSTRACT |
|---|
|
|
|---|
Abbreviations: PAR, plant adaptation region.
| INTRODUCTION |
|---|
|
|
|---|
While some species commonly grown in the Northern Great Plains and Intermountain USA are important only for specific areas, other species are commonly grown throughout these regions. Three of the more important species used in these regions are crested wheatgrass [Agropyron cristatum (L.) Gaertn.; A. desertorum (Fisch. ex Link) Schultes; A. fragile (Roth) Candargy], intermediate wheatgrass [Thinopyrum intermedium (Host) Barkworth & D.R. Dewey], and smooth bromegrass (Bromus inermis Leyss.). Breeding efforts aimed at improving these and other cool-season grass species are ongoing at several locations. Because of the importance of these species in these regions, the identification of best testing locations and environmental groupings of locations would be useful. Similar studies in other crop species, such as wheat (Triticum aestivum L., Trethowan et al., 2003), cotton (Gossypium hirsutum L., Blanche and Myers, 2006), and barley (Hordeum vulgare L., Navabi et al., 2006), have provided detailed information on common testing locations that should be incorporated into future crop improvement efforts.
The objective of this study was to determine locations most appropriate for developing and testing crested wheatgrass, intermediate wheatgrass, and smooth bromegrass for use in the Northern Great Plains and Intermountain regions. Specifically, this study aimed to (i) identify locations best suited for evaluating and testing improved cultivars of these species, (ii) identify useful environmental groupings of the locations included in the study based on entry performance, and (iii) determine whether best testing locations and environmental groupings were the same for each of the three species.
| MATERIALS AND METHODS |
|---|
|
|
|---|
|
|
|
Data Collection and Analysis
Data include initial stand frequency from 2001 and forage production from the 2001, 2002, and 2003 growing seasons at all sites. Initial stand frequency was estimated using the methods described by Vogel and Masters (2001). Briefly, this estimation consisted of the number of squares (15 cm2 in a grid) out of 50 containing rooted, live plant material. The ratio of the number of squares within the grid containing plant material to the total number of squares within the grid was calculated and converted to a percentage.
Forage production was estimated by machine harvesting to a stubble height of
15 cm and measuring plot forage wet weights. After forced-air drying at 60°C, dry weights were determined. These weights were then converted to kg ha1 for the resulting forage production values. At Mandan, Miles City, and both Utah locations, 0.5 m was trimmed from the ends of each plot before harvest to minimize border effects, and then only the two middle rows were harvested. At the Nebraska locations, plots were trimmed to uniform 3-m lengths before harvest, and a 0.91-m swath was harvested from the center of each plot. Forage was harvested once per year at all locations, except the Nebraska locations, where it was harvested twice per year. All values were converted to yearly forage totals in kg ha1.
Data were analyzed using the MIXED procedure of SAS (Littell et al., 1996; SAS Institute, 2006). The statistical model considered the main effect due to entries as fixed and all remaining main effects (locations, years, and blocks) and interactions as random. Further examination of the entries and the entry x location interaction occurred with the GGEbiplot software (Yan, 2001; Yan and Kang, 2003; Yan and Tinker, 2005). The GGEbiplot model employed was the tester (location)-centered model based on singular value decomposition of the untransformed data standardized by the within-tester (location) standard error.
| RESULTS AND DISCUSSION |
|---|
|
|
|---|
A key difference between the locations was precipitation. Thirty-year mean precipitation levels range from
300 mm yr1 at Miles City to
700 mm yr1 at Ithaca (Fig. 2). Based on 30-yr site averages and with some yearly deviations, the general trend throughout the study was normal to near-normal precipitation levels for the four eastern locations (Fig. 2). The Intermountain USA, including the Blue Creek and North Logan locations, experienced an extended drought during the period of this study. Blue Creek had below-normal precipitation each year except 1999, and North Logan had below-normal precipitation from 2000 to 2002.
Mixed Model Analysis
Population means for initial stand frequency and forage production differed among the crested wheatgrass and smooth bromegrass populations, but there were no differences among intermediate wheatgrass population means for either trait (Table 2). Although no differences occurred among intermediate wheatgrass entries analyzed on total yearly production across all locations, within the Nebraska locations there were differences among the entries when analyzed on an individual harvest basis (data not shown). Variation due to location was not significant for either trait for any of the species, but entry x location interaction was significant for each trait and for each species (Table 2). Because entry x location interaction was significant for each trait and species, analysis of the resulting data with biplot techniques was an appropriate method of interpreting the data.
|
|
Environmental Grouping and Testing Ability of Locations
On a biplot display, the cosine of the angle between the vectors (i.e., lines that connect the locations to the biplot origin) of two locations approximates the correlation between the two locations in ranking the entries: the smaller the resulting angle, the more highly correlated the locations (Yan and Kang, 2003; Yan and Tinker, 2006) (Fig. 3). Correlation coefficients between each set of locations were also calculated (Table 3). Based on the biplot analysis and correlation values, environmental groupings were identified, which represented groupings of locations within the target region where tested plant materials behaved similarly (Gauch and Zobel, 1997). This concept of identifying similar testing locations has been used for a number of species (Yan et al., 2000; Trethowan et al., 2003; Navabi et al., 2006).
|
Crested Wheatgrass
For crested wheatgrass stand frequency and forage production, there was good separation of locations into environmental groups. Group 1 included Ithaca, Mandan, and Miles City; and Group 2 included Blue Creek, North Logan, and Sidney (Fig. 3A, B). There was also high correlation among most of the locations for forage production (Table 3; Fig. 3B), although this may be due to the inclusion of universally poor performing entries that if removed might loosen the correlations (W. Yan, personal communication, 2006). Sidney was an intermediate location between the two groups, but its inclusion with Blue Creek and North Logan was most appropriate because of consistency of entry performance. Group 1 consisted mostly of the more eastern and northern sites with little respect to precipitation. Ithaca had the highest precipitation level (based on 30-yr means) of all locations (Fig. 2), but Mandan had only intermediate precipitation, and Miles City had the lowest precipitation of all locations. Although Mandan lies on the border between PARs 331.3 and 331.4 (Fig. 1) (Vogel et al., 2005b), it shares Ecoregion 331, and likely the PAR 331.3, with Miles City. None of the sites in Group 2 share PARs (Fig. 1) (Vogel et al., 2005b) but are the more western and southern locations. Additionally, North Logan and Sidney had similar precipitation levels (Fig. 2), and Blue Creek and North Logan represented the Intermountain region. For the crested wheatgrass traits, separation of locations into environmental groupings appeared to be based on geography rather than other traits, although the role of precipitation cannot be ruled out.
For crested wheatgrass stand frequency, Mandan was the most discriminating location (Fig. 3A). Blue Creek, Miles City, North Logan, and Sidney had roughly equivalent, yet intermediate, discriminating ability, and Ithaca was the least discriminating location (Fig. 3A). The most representative locations for crested wheatgrass stand frequency were Blue Creek and Miles City (Fig. 3A), although their representativeness was not high. The remaining locations were not representative. Due to the lack of representativeness of the locations, it was difficult to identify a best testing location for Group 1. However, the excellent discriminating ability of Mandan suggested it as a good choice. Additional work on stand frequency at each location would add clarification to these results. Among the Group 2 locations, Blue Creek and North Logan had roughly equivalent discriminating ability, but Blue Creek was more representative.
For crested wheatgrass forage production, the most discriminating locations were Ithaca followed by Blue Creek, Mandan, and Miles City (Fig. 3B). While North Logan and Sidney were both representative, their representativeness was similar to that of Ithaca, Mandan, and Miles City. Blue Creek was less representative. Overall, Ithaca was apparently the best location for testing crested wheatgrass forage production. Ithaca would also be the best location for testing crested wheatgrass forage production in environmental Group 1. The best location for testing crested wheatgrass forage production in Group 2 was unclear due to the lack of a location with both high discriminating ability and good representativeness, but Blue Creek might be the most promising because of its discriminating ability.
The identification of the best testing locations for crested wheatgrass was tenuous. However, due to the consistent environmental groupings of locations for the crested wheatgrass traits, efforts could be focused on developing crested wheatgrass varieties that are environmental group specific.
Intermediate Wheatgrass
There were three environmental groupings for intermediate wheatgrass stand frequency. Group 1 included Ithaca and Sidney, Group 2 Mandan and Miles City, and Group 3 Blue Creek and North Logan (Fig. 3C). The groupings shared obvious traits. Group 1 comprised the two Nebraska locations, which were the most central and among the higher-rainfall areas. Group 2 consisted of the Northern areas that also shared a PAR (Fig. 1) (Vogel et al., 2005b). Group 3 consisted of the two Intermountain locations. The groupings for intermediate wheatgrass forage production were almost identical with the exception of the Sidney location, which grouped more closely to the Mandan and Miles City locations than to the Ithaca location. Sidney shared an ecoregion with the two northern sites (Fig. 1).
For intermediate wheatgrass stand frequency, North Logan, Mandan, and Miles City were the most discriminating locations (Fig. 3C). The most representative locations were Blue Creek and North Logan. North Logan appeared to be a good location for testing intermediate wheatgrass stand frequency due to its excellent discriminating ability and representativeness. Within the identified environmental groupings, the best testing locations were North Logan and Mandan, due to better representativeness than Miles City, and Sidney. Intermediate wheatgrass forage production was best discriminated at Ithaca, followed by Blue Creek and North Logan (Fig. 3D). Ithaca was the only location exhibiting representativeness for intermediate wheatgrass forage production, making it the best location for testing this trait. In the other two intermediate wheatgrass forage production environmental groupings, Mandan was the best location for intermediate wheatgrass forage production in Group 2, and there was no substantial difference between Blue Creek and North Logan in Group 1.
The identification of best testing locations within environmental groupings for intermediate wheatgrass traits was clearer and more consistent than for crested wheatgrass. Environmental groupings, with the exception of Sidney grouping with the northern locations rather than Ithaca for forage production, were consistent for both traits. Additionally, North Logan was an excellent location for testing both traits in the Intermountain region, and Mandan appeared to be the best location for testing both traits among the northern locations. Ithaca would also be a testing location for both traits, but more due to the fact that it did not group well with the other locations with the exception of Sidney for stand frequency. Thus, targeting intermediate wheatgrass improvement to environmental groupings would be a good tactic and should result in improvements for both traits simultaneously.
Smooth Bromegrass
Three environmental groupings were identified for smooth bromegrass stand frequency. Group 1 included Blue Creek; Group 2 Ithaca, Miles City, North Logan, and Sidney; and Group 3 Mandan (Fig. 3E). Other than containing both Nebraska locations, there did not appear to be any discernible connection between the sites in Group 2. Ithaca, North Logan, and Sidney were also three of the higher-rainfall locations. It appeared that Blue Creek and Mandan form vastly different regions than do the other locations. Two environmental groupings were identified for smooth bromegrass forage production. Group 1 consisted of Mandan, North Logan, and Sidney; and Group 2 Blue Creek, Ithaca, and Miles City (Fig. 3F). Sidney was again an intermediate location, but appeared to fit better with Mandan and North Logan than the other locations. The connection between Group 1 locations was likely due to the already mentioned shared PAR between Mandan and Sidney (Fig. 1) (Vogel et al., 2005b). Additionally the three Group 1 locations were all intermediate in their precipitation levels (Fig. 2) and are sites typically classified as adapted for smooth bromegrass production. Group 2 is less clear because of the differences in precipitation levels and geographic location between Ithaca and the other two sites.
Mandan, followed by Sidney, was the most discriminating location for smooth bromegrass stand frequency (Fig. 3E). Ithaca, Miles City, North Logan, and Sidney were the most representative smooth bromegrass stand frequency locations (Fig. 3E). For smooth bromegrass forage production, Ithaca was the most discriminating location, with little difference among the remaining locations. Ithaca was also very representative (Fig. 3F), making it the best location for testing smooth bromegrass forage production.
Smooth bromegrass improvement did not appear to lend itself to targeted regions, particularly when attempting improvement of both traits simultaneously. Due to the ambiguity associated with the grouping of locations for stand frequency, none of the locations stood out as being best. For forage production, Ithaca was the best location. However, its connection to Blue Creek and Miles City in a grouping might be unrealistic due to vastly differing environmental conditions. There did seem to be good evidence for grouping Mandan and North Logan and then Blue Creek and Miles City for forage production. However, within these two groupings differences between the locations were minor, making recommendations of best testing locations difficult.
Across Traits and Species
Across both traits (stand frequency and forage production), there was some consistency in the clustering of locations into groupings. With the exception of the smooth bromegrass traits, Blue Creek consistently grouped with North Logan, representing the Intermountain locations, and Mandan consistently grouped with Miles City, representing the northern locations (Fig. 3). However across each of the species, with the above-mentioned exceptions, groupings were not consistent. The overall lack of common groupings and best testing locations across species suggested the need to approach each species individually, at different locations, or with trade-offs between discriminating ability and representativeness. From a practical standpoint the trade-offs approach is most feasible. Thus, a location like Blue Creek might be a good choice for testing crested wheatgrass due to its good discriminating ability and reasonable representativeness for both traits (Fig. 3A, B).
One of the more interesting and consistent findings was the value of Ithaca as a testing location. Ithaca was likely the worst location for testing stand frequency. It was one of the least discriminating and, generally, least representative locations for each species. However, for forage production, Ithaca was the most discriminating location and one of the most representative locations for each of the species. This result was most likely due to Ithaca's place as the location with the highest precipitation. The high precipitation likely made evaluation of stand frequency difficult because there was sufficient soil moisture and precipitation to ensure good stands of each species. However, the same precipitation levels made Ithaca a good location for testing forage production because precipitation was not limiting and the species were able to maximize their forage potential. Other consistencies were not as strong as those of Ithaca, and in general, best testing locations were both species and trait dependent.
| CONCLUSIONS |
|---|
|
|
|---|
| NOTES |
|---|
|
|
|---|
Received for publication September 26, 2006.
| REFERENCES |
|---|
|
|
|---|
This article has been cited by other articles:
![]() |
L. Gutierrez, J. D. Nason, and J.-L. Jannink Diversity and Mega-Targets of Selection from the Characterization of a Barley Collection Crop Sci., March 17, 2009; 49(2): 483 - 497. [Abstract] [Full Text] [PDF] |
||||
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
| The SCI Journals | Agronomy Journal | Vadose Zone Journal | |||
| Journal of Natural Resources and Life Sciences Education |
Soil Science Society of America Journal | ||||
| Journal of Plant Registrations | Journal of Environmental Quality |
The Plant Genome | |||