| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
a Texas Agric. Res. Cent., Route 2, Box 00, Stephenville, TX 76401-9698
b Rosepine Res. Stn., P.O. Box 26, Rosepine, LA 70659
c Dean Lee Res. Stn., 8105 Tom Bowman Ave., Alexandria, LA 71302-9306
* Corresponding author (jmuir{at}tamu.edu)
Received for publication December 11, 2000.
 |
ABSTRACT |
|---|
 TOP ABSTRACT INTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
Abbreviations: DM, dry matter
|
INTRODUCTION |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
Two aspects of initial establishment of burr medic were identified as potentially critical over a range of environments. These were seeding rate and P fertilization. Early evaluations of burr medic led to a range in recommended seeding rates from 11 (Lancaster et al., 1949) to 22 kg ha-1 (Donahue et al., 1956). For an unpredictable environment with highly erratic rainfall, Cocks (1988) suggested that even higher rates of hard seed could be beneficial. The recommended rate by the supplier of the Armadillo cultivar in the area of its development is at the low end of the above range (Pogue Seed Co., 1997). Naturally occurring stands, especially on upland sites in Louisiana, are not uniform and appear to regenerate primarily where competition is limited and heavy seed crops have been produced. Appropriate seeding rates could readily differ throughout the south-central region within the moisture range from humid Louisiana conditions to semiarid West Texas.
Although soil tests would not necessarily indicate P deficiency for many soils within the region, P can still be the most limiting nutrient for medic seedlings. Dahmane and Graham (1981) in Australia found that while maximum growth of an associated grass occurred with 10 mg P kg-1 soil, an annual medic (Medicago sp.) required 160 mg P kg-1 soil for maximum growth. Buxton (1989) noted that legumes in general require higher soil concentrations of nonmobile nutrients such as P for maximum growth than do grasses. In addition to the need for high concentrations of soil P by temperate legumes in general, responses of seedlings to banded application of P illustrate the potential added benefits to establishment from higher concentrations of P than typically recommended. Vough et al. (1995) suggested that four times as much broadcast P is required for a response equal to that from banded applications. Such responses of alfalfa (M. sativa L.) seedlings to high concentrations of banded P have been obtained on both low- and high-P soils (Sheard, 1980).
At a range of locations in Texas, various annual legumes have responded to P fertilization (Evers and Pennington, 1991; Haby, 1988; Read, 1980), with the greatest response obtained from the initial increment of P, which was generally near 25 kg P ha-1. The response of the annual medic buttonclover [M. orbicularis (L.) Bartal.] at Dallas was several times greater than that of arrowleaf clover (Trifolium vesiculosum Savi) or alfalfa (Read, 1980). Yield of established alfalfa stands responded to P fertilization on only the two lowest P sites of eight sites in an East Texas evaluation (Beedy et al., 1994). However, fertilization recommendations for pasture production of warm-season grasses in East Texas have included 17 kg P ha-1, even for soils testing high in P (Pratt et al., 1971). At Stephenville, TX, on a Windthorst fine sandy loam soil testing very low in soil P, added P at rates <15 kg ha-1 provided little yield response by alfalfa (Sanderson and Jones, 1993). Rates 
29 kg P ha-1 produced large yield responses. On a coastal plain soil in Louisiana, the greatest response of establishing tall fescue (Festuca arundinacea Schreb.) to soil amendment in the field was from P (Pitman, 2000). In an accompanying greenhouse evaluation on the Louisiana soil, linear response to soil P was obtained up to the highest level of 142 mg P kg-1 soil (Pitman, 2000).
Responses of Armadillo burr medic to seeding rate and P fertilization were evaluated at three diverse sites in Texas and Louisiana. The specific objectives were to assess the potential usefulness of this cultivar in cooler and more-humid environments than its area of development and to evaluate responses to the potentially critical factors of initial seeding rate and P fertilization at the diverse sites.
|
MATERIALS AND METHODS |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
Factorial arrangement of two seeding rates (11 and 22 kg ha-1; approximately 370 and 740 seeds m-2) and two rates of P fertilizer (0 and 25 kg P ha-1) were evaluated in a randomized complete block design in three replications at each location. All sites were disked, and P was incorporated into individual 1.5- by 6.0-m plots before seeding. Dehulled, scarified seeds (328000 seeds kg-1) inoculated with Rhizobium spp. specific for Medicago spp. (Urbana Laboratories, St. Joseph, MO) were broadcast on the surface of well-prepared seedbeds and firmly packed with a roller in early November 1998. Laboratory germination tests indicated 90% pure live seed, including 9% hard seed.
Seedling counts were taken in January 1999 in a 1-m2 area of each plot at each location. At Stephenville, forage dry matter (DM) accumulation above a 1.0-cm stubble height was determined in March in two 1-m2 areas of each plot. Regrowth following March harvest was measured in May. Carbaryl (1-naphthyl N-methylcarbamate) at 1.5 l a.i. ha-1 was applied to plots at Stephenville during March and April of both 1999 and 2000 due to heavy infestations of alfalfa weevil (Hypera postica) larvae and twelve-spotted cucumber weevil (Diabrotica unclecimpunctata howardi). At the other two locations, forage accumulation was assessed with a single harvest in April. Total N and P concentrations were determined in forage from both Stephenville and Rosepine using a modification of the Al block digestion procedure of Gallaher et al. (1975). Sample weight was 1.0 g, digestate used was 5 g of 33: 1:1 ratio of potassium sulfate (K2SO4)/copper sulfate (CuSO4)/titanium dioxide (TiO2), and digestion was conducted for 2 h at 400°C using 17 mL of sulfuric acid (H2SO4). Phosphorus and N in the digestate were determined by semiautomated colorimetry (Hambleton, 1977) using a Technicon Autoanalyzer II (Technicon Ind. Syst., Tarrytown, NY).
Number of seedpods, seed number, and total seed weight were determined at all three locations in May 1999 from samples collected in a previously unharvested 1-m2 area of each plot. Seed number and individual seed weight were estimated by removing seeds from a 50-pod subsample using a Forsberg drum scarifier (Forsberg, Thief River Falls, MN). At Stephenville, similar seed data were also collected for plants that had been harvested in March. The soil of individual plots was sampled to a 150-mm depth near the end of the initial growing season and analyzed for P, K, Ca, Mg, and pH (1:1 H2O). These analyses were conducted according to procedures described by Dunigan (1989). Phosphorus was extracted using Bray no. 2 solution, and ammonium acetate was used to extract the other nutrients.
Seedling counts were taken at all three locations in November 1999 and at Stephenville and Alexandria in November 2000 from 0.25-m2 quadrats following autumn germination. Forage accumulation was determined at Stephenville and Rosepine in March 2000 and at Stephenville and Alexandria in March 2001 using the procedures described for 1999 forage sampling. Nongerminated seeds (one season old) were counted from samples collected on 20 May 2000 at all three sites. This sampling took place before maturity of the 2000 seed crop. The top 2 cm of soil in a 1-m2 area was collected from plots using trowels and hand-sifted in the laboratory to recover not only pods, but also individual seeds.
Responses were analyzed using analysis of variance with a model including main effects of location and treatments and the interaction effects of location x P treatment x seeding rate, location x P treatment, location x seeding rate, and P treatment x seeding rate. The perceived likelihood of interactions, especially involving location, and their potential impact on treatment responses indicated that an experimental approach, such as this, to assess possible variations in responses should precede comprehensive assessment of responses to P rates at individual locations. When interactions occurred, responses were assessed within interacting factors. When only main effects of treatments were significant, multiple mean separations were determined using least significant difference (LSD). Only standard deviations (SDs) were reported where two means were different.
|
RESULTS |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
|
|
|
Plant tissue N and P concentrations were assessed only at Stephenville and Rosepine. Neither seeding rate nor P fertilization had any effect, but differences (P = 0.0001) were apparent between locations. The means (SD = 0.5) were 20.5 and 14.9 g N kg-1 at Stephenville and Rosepine, respectively, with even lower concentration at a sufficient level for plant growth and nutrition of grazing ruminants. The tissue P (SD = 0.14) means of 4.11 and 3.29 g kg-1 at Stephenville and Rosepine, respectively, suggest adequate P at Stephenville but less than the minimum critical concentration of 4 g kg-1 suggested by Denton et al. (1997) at Rosepine. Despite this apparent P-tissue deficiency, lack of response to added P indicates that P was not the first limiting factor at Rosepine.
Pod yield was not affected by treatments (P > 0.10) but differed by location (P = 0.0001; LSD0.10 = 22.8). Pod yields were 199 g m-2 at Stephenville, 89 at Alexandria, and 34 at Rosepine. Pod numbers increased with P application only at Stephenville. Pod numbers were greater at Stephenville than at Alexandria where pod numbers were in turn higher than those at Rosepine (location x P fertilization interaction, P = 0.001; LSD0.10 = 999; Fig. 1) . The higher seeding rate increased (P = 0.05; SD = 337) pod numbers 30% from 3211 to 4286 pods m-2 over all locations and P applications.
|
|
As with stands in the year of planting, naturally reseeded stands in the following year were affected (P = 0.04; LSD0.10 = 1263) by a location x P fertilization x seeding rate interaction. Seedling numbers were greatest at Stephenville for each treatment combination (Table 4). The higher seeding rate resulted in more seedlings at both P levels only at this location. The 1999 seedling counts reflect both the large amount of seeds present and the earlier evaluation date (before substantial self thinning) than the previous year. Autumn 2000 seedling counts were again greater (P = 0.001) at Stephenville than at Alexandria (5458 and 1655 seedlings m-2, respectively).
|
Soil seed (1999 seeds only) numbers in May 2000 were 69% lower than the June 1999 seed production. Treatment effects were not significant, but location differences existed (P = 0.001). Seed survival was 44% (4650 m-2) at Stephenville, 18% (1487 seeds m-2) at Alexandria, and 19% (476 seeds m-2) at Rosepine, based on seed production numbers from the previous season. The wetter, warmer climates at Alexandria and Rosepine probably contributed to the greater loss of hard seed than at Stephenville.
|
DISCUSSION |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
200 plants m-2 is probably sufficient for maximum DM yield under favorable growing conditions. Seedling numbers in November of the reseeding year (Table 4) were undoubtedly excessive and subject to substantial reduction through self thinning with development of individual plants. Cumulative emergence approaching 1000 seedlings m-2 was obtained in an Australian evaluation where seed reserves exceeded 10000 seeds m-2 (Carter and Lake, 1985). Derkaoui et al. (1991) concluded that the optimum seeding rate ranged from 465 to 930 seeds m-2 for a semiarid location in Morocco where both initial stand establishment and reserves for a subsequent year could be needed. Although this seed number is substantially greater than the plant population required, seed reserves within this range (870 seeds m-2) resulted in emergence of only 80 plants m-2 in an Australian experiment (Carter and Lake, 1985). Climate, proportion of hard seed and extent of scarification at planting substantially affect the relationship between seed number and subsequent plant populations. Seed production and seed survival through the initial summer from undefoliated stands of Armadillo burr medic were more than adequate for natural stand regeneration in the following growing season at all three locations. At Stephenville, seed yields were reduced 34% by forage harvest in March. Considering the large seedling populations in the second and third seasons, sufficient number of seeds can be expected from moderately grazed stands for development of soil seed reserves needed to reseed stands. The high seedling numbers observed in the reseeded stand may indicate insufficient hard seed for long-term stands unless soil seed reserves are replenished each year. Seedlings present in November 1999 as a percent of the number of seeds produced in the preceding growing season were 54% at Stephenville, 55% at Rosepine, and 27% at Alexandria. These high germination rates of new-crop seeds contrast with results from locations with drier, cooler summer seasons where autumn seedling recruitment arises primarily from seeds produced in earlier years (Cocks, 1993; Wagner and Spira, 1994). Only a small portion of our 1999 seedlings could have arisen from the 9% hard seed in the seeds originally planted in 1998. Also, in contrast to our results, only 5 to 9% of available seeds produced plants (cumulative emergence) in an Australian evaluation although seed reserves declined by 16 to 41% (Carter and Lake, 1985). The 44% soil seed survival at Stephenville in May 2000 indicates that very few seeds germinated during the winter and spring months at this location. Because only 18.5% of soil seed reserves were still present at the two Louisiana sites 11 mo after seed set, loss of seed reserves through germination during the cool season, seed deterioration, or consumption by predators was substantial at these sites.
Cocks (1988) reported that high seeding rates from 20 to 40 kg ha-1 resulted in greater seed production despite seedling competition. We obtained similar seed yield response only with P fertilization. Competitive abilities and effects of crowding were suggested to be important in devising management strategies for medics in Australia (Stern, 1985) where mixtures of annual medic species are often used. Initial seeding rate must include consideration that persisting seed reserves of at least 200 kg ha-1 are critical for long-term stand maintenance in some environments (Carter and Lake, 1985).
The distinct contrast of superior DM accumulation at one location and superior seed production at another may be due to climate and environment because freeze damage in December 1998 and insect herbivory in March 1999 contributed to the limited forage accumulation at Stephenville. Seed production is recognized as particularly responsive to environmental conditions (Reed et al., 1989). Burr medic generally produces well and persists in fertile soils (Archer and Bunch, 1953; McKee, 1934). Annual medics are known to flower in response to daylength and temperature, but factors such as grazing can also influence date of flowering (Reed et al., 1989). Additionally, seed yield can be reduced by excessive grazing, frost, and drought (Reed et al., 1989). Decreasing soil moisture following floral initiation may contribute to partitioning of photosynthates from vegetative growth to reproductive development. Such a response is consistent with adaptation of annual medics to environments with low summer rainfall (Cocks, 1995) or heavy harvest pressure (Shrestha et al., 1998; Zhu et al., 1996) and is consistent with responses of many other legumes. Continuing rain and sustained soil moisture late in the growing season at Alexandria may have contributed to sustained partitioning of photosynthate to vegetative growth while the lower rainfall following partial removal of leaves by insects at Stephenville may have triggered greater seed production, utilizing essentially all available photosynthates.
Previous reports of forage production responses to P fertilization by annual medics have been inconsistent. High, not low, rates of P increased productivity in low-P soils (Pozo et al., 1994; Materon and Ryan, 1995). Bolland (1997), however, concluded that burr medic required very low levels of P in alkaline soils to attain maximum yields. Bolland (1992) and Shukla and Lal (1991) reported that adequate soil moisture was critical for yield responses of burr medic to P application. The extremely low forage yields in the first 2 yr, and perhaps the lack of response to P at Stephenville, could have been partially due to adequate P levels, low soil moisture, and the observed freeze damage at that study site. The early harvest at this location was made to salvage remaining forage because insects continued to damage the crop despite monthly applications of insecticide. The previous year, in contrast, forage DM production of 2230 kg ha-1 was obtained from Armadillo burr medic in other research at Stephenville, TX (Muir and Reed, 1999). Rainfall of 350 mm during the period of December 1997 to March 1998 produced the 2230 kg ha-1 yield while only 159 mm of rain was received from December 1998 to March 1999 when forage yields were 260 kg ha-1 at Stephenville and insect populations had less forage to consume. Also, there was negligible frost damage in the 1997–1998 winter (Muir and Reed, 1999) while freeze damage was severe during the present study in December 1998 and January 1999.
Bolland (1985) reported that higher P resulted in increased seed yield through increased number of pods. This is in agreement with our results only at the high seeding rate. Bolland and Paynter (1990) indicated that seed size and P concentration are important for subsequent stand productivity. We did not obtain differences in seed size (3.3 and 2.7 mg seed-1 at Stephenville and Alexandria, respectively) of sufficient magnitude to affect stands.
A key factor in the commercial acceptance of annual medics in Australia has been the low levels of inputs required to produce high quality, profitable pastures (Sheaffer and Lake, 1997). Although the average yields of 3330 kg ha-1 in the initial year at Alexandria were the greatest among the three locations, annual ryegrass (Lolium multiflorum Lam.) yields at this location generally exceed 10000 kg ha-1 (Alison et al., 1991). Naturalized populations of annual medics throughout Louisiana typically occur on disturbed sites. Dense stands of sod-forming warm season grasses generally limit re-establishment of medics under pasture conditions in this high-rainfall environment. Additionally, DM yields were low at Rosepine. Limitations at this location probably include low soil fertility and biological constraints such as diseases and pests, as reported for Australia (Denton et al., 1997; Reed et al., 1989; Sheaffer and Lake, 1997). Foliar disease symptoms were observed, but not identified, in the fall growth at Rosepine in 1999 and in spring growth in March 2000 at Alexandria. Thus, the relatively low production potential, lack of forage DM response to moderate P applications, and constraints to natural stand regeneration in pastures of warm-season perennial grasses limit the usefulness of burr medic for livestock enterprises in these high-rainfall environments. Its use for strategic, early cool-season grazing or in wildlife food plots, however, may become important. Even comparatively low-yielding species can provide useful forage for wildlife when stands regenerate from year to year without additional inputs. Also, the early winter growth can fill an acute nutritional gap at a time when other cool-season pasture species typically produce minimal growth. Treatments applied failed to enhance forage production of this legume at Louisiana sites because the low seeding rate provided sufficient number of seeds and the low forage yield was apparently not due to soil-P levels as first-limiting factors surmountable by modest fertilization rates.
Both amount and distribution of rainfall at Stephenville, where low rainfall in summer is typical, appear to be more conducive to sustained stands of annual medic. Somewhat less-dense growth of predominantly bunch-type warm season grasses with limited summer rainfall enhances opportunity for high seedling densities in autumn. The sod-forming grasses typical of Louisiana pastures compete more with seedlings of cool-season species in autumn and with reproductive burr medic in the spring. The typical extensive pasture management, especially the acceptance of forage production levels requiring lower stocking rates, provide realistic potential for use of Armadillo burr medic as pasture in north-central Texas. Yield stability needs assessment, particularly as affected by typical variation in rainfall, early defoliation, and insect damage in commercial-scale plantings. Although seed production was more than adequate in all treatments, the positive response in seed production to the combination of high seeding rate and P fertilization and the response in seedling numbers to increased seeding rate in both the establishment year and the subsequent growing season at Stephenville are noteworthy. Such responses to initial seeding rate could justify the added input because heavier grazing pressure and risk of adverse weather conditions are typical. The inclusion of some unscarified seeds in commercial plantings would also bolster soil seed banks and stand regeneration in subsequent years. Seed yield increase and response in initial seedling numbers to P with the highest seeding rate at Stephenville illustrate that maintenance of adequate soil P levels can contribute to establishment and natural regeneration of burr medic in this environment.
|
REFERENCES |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
This article has been cited by other articles:
|
|
 |
|
  J. P. Muir and J. R. Bow Herbage, Phosphorus, and Nitrogen Yields of Winter-Season Forages on High-Phosphorus Soil Agron. J., June 2, 2009; 101(4): 764 - 768. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 J. R. Bow, J. P. Muir, D. C. Weindorf, R. E. Rosiere, and T. J. Butler Integration of Cool-Season Annual Legumes and Dairy Manure Compost with Switchgrass Crop Sci., July 1, 2008; 48(4): 1621 - 1628. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
J. P. Muir, W. R. Ocumpaugh, and T. J. Butler Trade-Offs in Forage and Seed Parameters of Annual Medicago and Trifolium Species in North-Central Texas as Affected by Harvest Intensity Agron. J., January 1, 2005; 97(1): 118 - 124. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
W.R. Ocumpaugh, M.A. Hussey, W.J. Grichar Jr., J.C. Read, D.H. Bade, W.E. Pinchak, G.R. Smith, R.A. Lane, W.D. Pitman, J.P. Muir, et al. Registration of 'Armadillo' Burr Medic Crop Sci., May 1, 2004; 44(3): 1023 - 1024. [Full Text] [PDF]
|
|
|
|
|
|
J. P. Muir, W. R. Ocumpaugh, and J. C. Read Spring Forage Yield and Nutritive Value of Texas Black Medic Accessions Agron. J., July 1, 2003; 95(4): 908 - 912. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
| The SCI Journals | Crop Science | 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 | |||
