|
|
||||||||
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
ANIMAL NUTRITION |



* Eastern Oregon Agricultural Research Center, Oregon State University, Burns 97720;
and
Department of Animal Sciences, University of Kentucky, Lexington 40546;
and
College of Veterinary Medicine, Oregon State University, Corvallis 97331; and and
Animal Science Department, University of Tennessee, Knoxville 37996
| Abstract |
|---|
|
|
|---|
Key Words: cattle ergot alkaloid ergovaline prolactin
| INTRODUCTION |
|---|
|
|
|---|
Ergot alkaloids ingested by ruminants have been reported to decrease feed intake, elevate body temperature, lead to excessive salivation, increase respiration rate, decrease reproductive efficiency, and decrease peripheral circulation; these symptoms are collectively known as fescue toxicosis (Rhodes et al., 1991
; Mizinga et al., 1992
; Aldrich et al., 1993
). Other maladies associated with fescue toxicosis include fescue foot, fat necrosis, and agalactia (Hemken et al., 1984
; Paterson et al., 1995
). Paterson et al. (1995)
reported decreased serum prolactin as a consistent measurable result of fescue toxicosis. Research has suggested that a yeast-derived cell wall preparation (YCW) may minimize, or alleviate, negative effects of endophyte toxins on animal performance (Akay et al., 2003a
; Ely et al., 2003
; Aaron et al., 2006
).
Our objectives were to determine the influence of YCW on forage intake and digestibility, ruminal fermentation characteristics, serum prolactin and prolactin stores, incidence of fescue foot, and milk production in beef cattle consuming high-ergot alkaloid tall fescue straw during late fall and winter.
| MATERIALS AND METHODS |
|---|
|
|
|---|
Exp. 1: Digestion and Physiology Study
Sixteen ruminally cannulated Angus x Hereford steers (200 ± 6 kg of BW) were used in a randomized complete block design in December of 2005. Steers were blocked by BW and, within block, were randomly assigned to treatments and housed in individual pens (2 x 4 m) within an enclosed barn, without temperature control and with continuous lighting. Steers had unrestricted access to fresh water. Before straw feeding (0700), a trace mineralized salt mix (22 g/d;
96% NaCl,
0.20% Mn,
0.10% Fe,
0.10% Mg,
0.05% S,
0.025% Cu,
0.01% Co,
0.008% Zn, and
0.007% I) and soybean meal (SBM; 0.068% of BW; CP basis) were supplemented intraruminally via cannula to meet 100% of the estimated degradable intake protein requirement, assuming a microbial efficiency of 11% (NRC, 1996
; model 1). A YCW (MTB-100, Alltech Inc., Nicholasville, KY) was provided to yield the following treatments: 1) 0 g/d of YCW (YCW0), 2) 20 g/d of YCW (YCW20), 3) 40 g/d of YCW (YCW40), and 4) 60 g/d of YCW (YCW60). The appropriate quantity of YCW was added to each steers SBM-trace mineralized salt supplement daily. All steers consumed high-ergot alkaloid (579 µg/kg of ergovaline; DM basis) tall fescue straw that had been processed with a bale chopper (BC-900, Newhouse Manufacturing; Redmond, OR) to yield a particle length of 4 to 8 cm. Tall fescue straw was provided at 0730 at 120% of the previous 5-d average intake, with orts from the previous day determined before feeding. Nutrient content of straw and SBM is provided in Table 1
.
|
At 0700 on d 21, immediately following supplementation, each steer was intraruminally pulse-dosed with 4 g of Co-EDTA in a 150-mL aqueous solution (Uden et al., 1980
) to determine ruminal liquid fill and dilution rate. The Co marker was administered throughout the rumen by using a stainless-steel probe with a perforated tip. Ruminal fluid (approximately 100 mL) was collected by suction strainer (Raun and Burroughs, 1962
; 19-mm diameter, 1.6-mm mesh) prior to SBM supplementation and at 3, 6, 9, 12, 18, and 24 h after Co-EDTA dosing. Samples were immediately analyzed for pH, subsampled by placing 5 mL of ruminal fluid in 1 mL of 25% (wt/vol) meta-phosphoric acid, and stored (–20°C) for later analysis of NH3-N and VFA. In addition, 20 mL of ruminal fluid was stored (–20°C) for later analysis of Co concentration.
Frozen NH3-N and VFA samples were prepared for analysis by thawing, centrifuging (10,000 x g; 20 min), and collecting the supernatant. Volatile fatty acids were analyzed as described by Harmon et al. (1985)
, and NH3-N was determined by a modification (sodium salicylate was substituted for phenol) of the procedure described by Broderick and Kang (1980)
, with a UV-visible light spectrophotometer (UVmini1240, Shimadzu, Columbia, MD). Frozen ruminal fluid samples were prepared for Co analysis by thawing, centrifuging (2,000 x g; 20 min), and collecting the supernatant. Cobalt concentration in ruminal fluid was analyzed by atomic absorption by using an air-acetylene flame (model 351 AA/AE spectrophotometer, Instrumentation Laboratory Inc., Lexington, MA). Ruminal fluid fill and dilution rate were determined by regressing the natural logarithm of Co concentration against sampling time, as described by Warner and Stacey (1968)
.
Spot urine samples were collected at 0700 on d 22 to 27. Urine samples were collected in polyethylene bags (25 x 35 cm) split diagonally and secured with string over the withers and hips of the animal. Bags were left until a sample was collected, which never exceeded 1.5 h. Urine samples were composited by steer and stored (–20°C) for later analysis of creatinine (277–10501, Wako Chemicals US Inc., Richmond, VA) and ergot alkaloids as described by Hovermale and Craig (2001)
and Lodge-Ivey et al. (2006)
. Daily urinary output was calculated based on an estimated excretion rate of 28 mg of creatinine/kg of BW (Whittet et al., 2004
).
On d 22 through 27, fecal grab samples were collected 2 times/d at 12-h intervals, with a 2-h increment added between days to shift the sampling times. This allowed sampling on every even hour of the 24-h day. Fecal subsamples (200 g) were composited by steer, stored (–20°C), lyophilized, and ground in a Wiley mill (1-mm screen).
On d 28, treatment effects on ruminal DM and indigestible ADF (IADF) fill were determined by manually removing the reticuloruminal contents 4 h after feeding. Total ruminal contents were weighed, thoroughly mixed by hand, and subsampled (300 g/subsample, wet weight) in triplicate. The remaining ruminal contents were replaced immediately in the steer. Ruminal samples were weighed, dried in a forced-air oven (55°C; 96 h), reweighed for DM, ground to pass a 1-mm screen in a Wiley mill, and composited by steer.
Ground samples were analyzed for DM and OM (AOAC, 1990
), N (Leco CN-2000, Leco Corp., St. Joseph, MI), and NDF (Robertson and Van Soest, 1981
) and ADF (Goering and Van Soest, 1970
) by using procedures modified for use in an Ankom 200 fiber analyzer (Ankom Co., Fairport, NY). In addition, samples were analyzed for IADF as described by Bohnert et al. (2002)
. The digesta kinetic techniques described by Van Soest (1994)
were used to determine IADF passage by dividing IADF intake by the quantity of IADF in the rumen at 4 h after feeding. Diet digestibility was determined by using IADF fecal concentration in conjunction with nutrient concentration of forage and SBM (Merchen, 1988
). Air-dried straw and orts and lyophilized fecal samples were analyzed for ergovaline and lysergic acid by HPLC, as described by Hovermale and Craig (2001)
and Lodge-Ivey et al. (2006)
, respectively.
Rectal (GLA M500, Agricultural Electronics, San Luis Obisbo, CA) and skin (RayngerST, Raytek, Santa Cruz, CA) temperatures were measured at 1300 on d 1, 8, 15, 22, and 29. Daily ambient temperature was recorded hourly using a weather station (Model 012, Campbell Scientific Inc., Logan, UT), which was located within 300 m of the study location. In addition, 4 h after feeding on d 1, 8, 15, and 22, 10 mL of blood was collected by coccygeal venipuncture into sterile, nonadditive red-topped tubes (Becton Dickinson and Company, Franklin Lakes, NJ) and 20-ga x 2.54-cm blood collection needles (Becton Dickinson and Company, Rutherford, NJ). Blood samples were allowed to clot overnight at 4°C and centrifuged (1,500 x g, 20 min,4°C), and the serum was harvested and stored (–20°C) for prolactin analysis, as described by Hockett et al. (2000)
.
Thyrotropin-Releasing Hormone Challenge.
Steers were subjected to a thyrotropin-releasing hormone (TRH) challenge to measure pituitary prolactin stores on d 29. The afternoon prior to the challenge, steers were catheterized via the jugular vein with Radiopaque FEP Teflon i.v. catheters (16 ga x 14 cm; Abbot Hospitals Inc., North Chicago, IL). On the day of the challenge, TRH (no. P2161, Sigma, St. Louis, MO) was reconstituted by using 0.01 M acetic acid, and was mixed with sterile physiological saline for a final concentration of 30.36 µg of TRH/mL. Each steer was dosed with 1 µg of TRH/kg of BW via the jugular catheter (fitted with a 0.2–µm, low-protein-binding, nonpyrogenic filter; PN 4612, Pall Life Sciences, East Hills, NY), followed by 20 mL of sterile saline to chase the peptide. Blood samples were collected at –30, –15, 0 (before TRH administration), and 5, 10, 15, 20, 30, 45, 60, 90, 120, and 150 min (post-TRH administration). Catheters were kept clot free between sampling times by flushing with 10 mL of physiological saline. Blood samples were immediately transferred to red-topped Vacutainer tubes and allowed to clot overnight at 4°C. Samples were then centrifuged (1,500 x g, 20 min, 4°C) and serum was harvested and stored (–20°C) for prolactin analysis as described previously. The –30-,–15-, and 0-min samples were averaged and used to calculate a basal prolactin concentration (pre-TRH administration; Thompson et al., 1987
). Area under the curve was determined for prolactin by using the trapezoidal summation method.
Statistical Analysis.
Data were analyzed as a randomized complete block by using PROC GLM (SAS Inst. Inc., Cary NC). Treatment and block were included in the model. Contrast statements were 1) linear effect of increasing YCW, and 2) quadratic effect of increasing YCW. Ruminal pH, NH3-N, and VFA, and TRH challenge data were analyzed by using the REPEATED statement with PROC MIXED of SAS. The model included steer, treatment, block, hour, and treatment x hour. In addition, weekly temperature and serum prolactin data were analyzed using the REPEATED statement with PROC MIXED of SAS. The model included steer, treatment, block, day, and treatment x day. The same contrasts described above were used to partition treatment effects for ruminal pH, NH3-N, VFA, temperature, and serum prolactin. An autoregressive covariance structure (AR1 of PROC MIXED of SAS) was determined to be most appropriate based on Akaikes information criterion.
Exp. 2: Cow Performance, Parturition, and Milk Production
Sixty pregnant (approximately 200 d of gestation) Angus x Hereford cows (517 ± 5 kg of BW) were stratified by BCS (5.0 ± 0.1; Herd and Sprott, 1986
) and assigned randomly to 1 of 20 pens and 1 of 5 treatments (3 cows/pen; 4 pens/treatment) in a randomized complete block design in January of 2006. All cows had unrestricted access to fresh water and a loose, trace mineralized salt (
96% NaCl,
0.20% Mn,
0.10% Fe,
0.10% Mg,
0.05% S,
.025% Cu,
0.01% Co,
0.008% Zn, and
0.007% I). A high-ergot alkaloid tall fescue straw and a low-ergot alkaloid tall fescue straw (449 and 147 µg/kg, respectively; Table 1
) were used in formulating the following treatments: 1) ad libitum access to low-ergot alkaloid straw (control), 2) ad libitum access to high-ergot alkaloid straw (YCW0), 3) ad libitum access to high-ergot alkaloid straw plus 20 g/d of YCW (YCW20), 4) ad libitum access to high-ergot alkaloid straw plus 40 g/d of YCW (YCW40), and 5) ad libitum access to high-ergot alkaloid straw plus 60 g/d of YCW (YCW60). Straw was provided to the cows directly from 3-tie bales (46 x 122 x 61 cm) without further processing. In addition, SBM was provided (1 kg of DM) daily at 0700 to all treatments to meet 100% of the estimated degradable intake protein requirement, assuming a microbial efficiency of 11% (NRC, 1996
; model 1). The top-dressed YCW and SBM were provided in a bunk, with approximately 72 cm of bunk space per cow, within each pen. The high-ergot alkaloid straw was used to estimate the degradable protein requirement because of its lower CP concentration. This was to minimize the potential effects caused by differences in forage CP between the low- and high-ergot alkaloid straws (6.5 vs. 5.6%, respectively). The amount of SBM was doubled (2 kg of DM) for all treatments at d 44 of the trial because cow BW and BCS were not being maintained at the previous supplementation level.
Straw samples for alkaloid analysis were obtained weekly, stored (–20°C), and analyzed as described in Exp. 1. In addition, additional samples of low- and high-ergot alkaloid straw and SBM were collected weekly, dried in a forced-air oven (55°C; 48 h), reweighed for calculation of DM, ground in a Wiley mill (1-mm screen), and composited by source and period for analysis of NDF, ADF, N, and OM, as described in Exp. 1.
An evaluation of all cows was conducted daily at 0630. Appraisal included assigning a locomotion score from 1 to 5 (adapted from Sprecher et al., 1997
; Table 2
); a locomotion score of 3 or greater was assumed to be indicative of fescue foot and necessitated removal from the study (1 cow on d 55). Ambient temperature was recorded as described in Exp. 1. The pens used in the current study were located within 200 m of the weather station.
|
Blood samples (approximately 10 mL) were collected by coccygeal venipuncture into sterile, nonadditive, red-topped tubes (Becton Dickinson and Company) and 20-ga x 2.54-cm blood collection needles (Becton Dickinson and Company) on d 1 and within 24 h after parturition. Blood was allowed to clot overnight. Samples were then centrifuged (1,500 x g, 20 min, 4°C) and serum was harvested and stored (–20°C) for prolactin analysis, as described previously (Exp. 1 and 2; interassay CV = 5.0%, and intraassay CV = 8.0%).
After parturition (75 ± 2 d from study initiation), cows and calves were placed in a common pasture (7.3 ha) that had been harvested for hay the previous summer, and the cows and calves were managed as a single group. Cows were provided approximately 11.2 kg/d (DM basis) of meadow hay (6.3% CP, DM basis). One week after the last cow calved, cow-calf pairs were moved to the Northern Great Basin Experimental Range 72 km west-southwest of Burns, Oregon. Cow-calf pairs grazed a sagebrush-bunchgrass range, as described by Ganskopp (2001)
, and were managed according to Northern Great Basin Experimental Range and Eastern Oregon Agricultural Research Center management practices. Postpartum (60 ± 2 d) milk production was estimated using the weigh-suckle-weigh technique (Williams et al., 1979
), with an 8-h separation. Excretory losses were considered minimal and were not collected (Lampkin and Lampkin, 1960
).
Statistical Analyses.
Data were analyzed as a randomized complete block by using PROC GLM (SAS Inst. Inc.). Treatment and block were included in the model. Contrast statements were 1) linear effect of increasing YCW supplementation, 2) quadratic effect of increasing YCW supplementation, 3) control vs. YCW0, and 4) YCW0 vs. YCW20. The fourth contrast was included to evaluate no supplement compared with the maximum currently recommended dose of 20 g/d.
| RESULTS AND DISCUSSION |
|---|
|
|
|---|
|
|
No treatment x day interactions were detected (P > 0.10), so means were averaged across days for rectal, flank, and coronary band temperature. Increasing YCW did not influence rectal or flank temperature (P > 0.17), although a quadratic response (P = 0.06) tended to be observed in coronary band temperature, with YCW40 being the highest (Table 5
). Rectal temperatures in animals consuming high-endophyte tall fescue have been shown to increase in heat-stressed environments but, as in the current study, no differences are seen in moderate to colder ambient temperatures (below 27°C; Hannah et al., 1990
; Rhodes et al., 1991
; Peters et al., 1992
). This is the first data of which we are aware evaluating the effect of YCW on coronary band temperature. We anticipated that with the cold environment of the current study, coronary band temperature would increase with increasing YCW, theoretically because of decreased vasoconstriction and improved blood supply resulting from a decrease in ergot alkaloid metabolism (Oliver et al., 1993
). It is not clear why we observed the quadratic effect in coronary band temperature, but it is possible that our ability to detect potential treatment differences was influenced by the daily variation in fecal and urine contamination of the hair at the coronary band, which may not have allowed for an accurate measurement of temperature.
|
During the TRH challenge, data from 1 steer on YCW0 was excluded because of loss of catheter patency. Prolactin area under the curve for the TRH challenge increased linearly (P = 0.05) with increasing YCW (Table 5
), but the primary source of the linear effect was the increase in prolactin between the 0 and 20 g/d dose, with little difference between YCW20, YCW40, and YCW60. In contrast to our results, Stamm et al. (1994)
reported no difference in serum prolactin in response to a TRH challenge by steers consuming increasing ergovaline levels, but they did not measure the area under the curve and only reported concentration differences. In addition, Stamm et al. (1994)
used tall fescue straw with a lower concentration of ergovaline (475 µg/kg) than in our study (579 µg/kg). Nevertheless, other research has noted a depression in TRH-induced prolactin levels across temperature ranges and seasons in ruminants grazing high-endophyte fescue (Hurley et al., 1980
; Thompson et al., 1987
). Therefore, our results suggest that YCW may be able to correct the prolactin depression normally associated with fescue toxicosis.
No differences (P > 0.25) were detected in intake of ergovaline, lysergic acid, or the combination of ergovaline and lysergic acid (Table 6
). Lysergic acid was included in the analyses because it results from the degradation of ergot alkaloids containing the core ring structure of lysergic acid, such as ergovaline, ergocryptine, ergotamine, and ergonovine, and recent research has suggested it may be a primary toxin in fescue toxicosis (Hill et al., 2003
; De Lorme et al., 2007
). No differences (P > 0.20) were detected in ergovaline or the combined ergovaline and lysergic acid content of feces. However, fecal content of lysergic acid tended to decrease linearly (P = 0.09) as YCW increased, which may be a function of the numerical reduction in ergot alkaloid intake. No differences (P > 0.23) were detected in urine lysergic acid content or total excretion (urine and feces) of ergovaline, lysergic acid, or the combination of ergovaline and lysergic acid. Unlike for ergovaline, urine is the primary route of excretion for lysergic acid.
|
It is possible that we may have underestimated urinary excretion of lysergic acid. This is because we collected spot urine samples daily at 0700. In a recent study, Pearce and Masters (2006)
measured Na excretion in spot samples of urine from wethers by using the creatinine ratio, as we did in the current study. They noted significant diurnal variation in estimated Na excretion depending on the time of spot sample collection. Consequently, by not collecting spot samples throughout the day, we may not have accurately characterized lysergic acid excretion. In addition, based on data from De Lorme et al. (2007)
, the concentration of lysergic acid in ruminal fluid appears to increase until approximately 6 h postfeeding and can be maintained for at least 12 h postfeeding in ruminants consuming endophyte-infected tall fescue. Therefore, given that the rumen is assumed to be the primary site of lysergic acid absorption (Westendorf et al., 1992
; Hill et al., 2001
) and that we provided straw once daily at 0700, we may not have obtained spot urine samples when lysergic acid excretion was greatest, thereby underestimating excretion. However, all treatments were handled the same, and we do not anticipate that our sampling procedure influenced overall interpretation of the data.
Exp. 2: Cow Performance and Production Study
The daily average ambient temperature from study initiation through parturition (approximately 75 d) was 0°C, with the daily average minimum and maximum being –5 and 6°C, respectively. During the course of the experiment, one cow receiving YCW40 was removed on d 55 because of a lameness score of 4 resulting from fescue foot. This animal made a full recovery, calved without difficulty, weaned a healthy calf, and rebred within a 45-d breeding season. In addition, another cow receiving YCW40 was removed because of misdiagnosis of pregnancy at study initiation. Data from these animals were completely removed from the data set.
Intake of straw was not measured in the current study, but we did note the quantity of straw offered from study initiation to calving. There were no treatment differences in the quantity of straw DM offered (P > 0.05), with the control, YCW0, YCW20, YCW40, and YCW60 receiving 13.0, 12.0, 11.9, 12.9, and 11.4 kg/d, respectively (data not shown).
Contrary to Akay et al. (2003a)
, who reported an increase in BW of cows supplemented with YCW at 20 g/d compared with those not supplemented during a 5-mo period (May to October) when grazing tall fescue, we noted no differences (P > 0.20) in pre- or postcalving changes in BW or BCS as YCW increased (Table 7
). These differing results may be a function of heat stress in the Akay study compared with the winter temperatures associated with the current experiment. Nevertheless, the precalving BW change increased (P = 0.02) in control cows compared with YCW0 cows. In agreement, Paterson et al. (1995)
reported lower ADG by cows grazing endophyte-infected fescue compared with cows grazing non-endophyte-infected fescue. Our observed increase in precalving BW change for control cows compared with YCW0 cows may be related to the lower CP content of the high-alkaloid straw, even though we attempted to compensate for the CP difference.
|
Suppression of the periparturient surge of prolactin has been reported in cattle consuming ergot alkaloids and is associated with decreased metabolic activity of the mammary cells (Tucker, 1985
). In addition, suppressed or reduced prolactin concentrations have been reported in numerous studies with animals consuming high-endophyte diets compared with those consuming low- or no-endophyte diets (Schillo et al., 1988
; Stamm et al., 1994
; Samford-Grigsby et al., 1997
). In this study, postcalving prolactin concentration as well as the change from initial to postcalving serum prolactin concentration increased linearly (P = 0.02 and P = 0.06, respectively) with increasing YCW (Table 7
). This coincides with the linear increase detected with the TRH-challenge area under the curve data in Exp. 1. In addition, postcalving serum prolactin was 112 ng/mL for YCW20 compared with 62 ng/mL for YCW0 (P = 0.04), and the change from initial to postcalving prolactin tended (P = 0.08) to be greater with YCW20 compared with YCW0. These data, along with the similar post-calving prolactin concentrations for YCW20, YCW40 and YCW60 (112, 100, and 127 ng/mL, respectively), suggest that YCW20 was sufficient to alleviate the prolactin depression observed with intake of high-ergot alkaloid tall fescue straw. In addition, postcalving serum prolactin and the change from initial to postcalving serum prolactin decreased (P = 0.003 and P = 0.004, respectively) with YCW0 compared with the control.
We are aware of no data evaluating the effects of increasing YCW supplementation on milk production in ruminants consuming high-alkaloid forage. However, Peters et al. (1992)
reported that daily milk production was 25% lower in animals consuming endophyte-infected compared with endophyte-free tall fescue. In our study, milk production of cows that consumed high-ergot alkaloid tall fescue straw during the last third of gestation increased linearly (P = 0.04) as YCW increased (Table 7
). Despite differences in milk production, calf BW at the time of weigh-suckle-weigh was not affected (P > 0.13) by treatment (86, 86, 84, 83, and 95 kg for control, YCW0, YCW20, YCW40, and YCW60, respectively; data not shown). In addition, calf BW gain from birth to weigh-suckle-weigh was not influenced by treatment (P > 0.12) and averaged 0.83 kg/d across all treatments (data not shown). Nevertheless, the YCW-related increase in milk production was similar to the observed increase in cow serum prolactin concentration.
Increasing YCW resulted in greater prolactin stores, alleviated prolactin depression, and increased milk production of beef cattle consuming high-ergot alkaloid tall fescue. Based on our data and other research (Akay et al., 2003a
,b
; Aaron et al., 2006
), 20 g/d of YCW normally appears to be sufficient to reduce the consequences of fescue toxicosis (principally decreased performance and serum prolactin) normally observed with intake of endophyte-infected tall fescue by beef cattle. The mode of action for YCW is not clear, but it has been proposed that YCW binds the toxin(s) causing fescue toxicosis (Akay et al., 2003a
). Nevertheless, ergovaline and lysergic acid excretion in the current study were not influenced by increasing YCW. It is possible that metabolism or excretion of an alkaloid(s) not measured could have been influenced by YCW supplementation and caused the observed effects. Further research is warranted concerning the mechanism by which the YCW reduces the effects of fescue toxicosis.
| Footnotes |
|---|
2 The authors are grateful to Alltech Inc. for partial support, and to Matt Eddie of Quality Trading LLC and Steve Van Mouwerick of Anderson Feed & Grain Co. Inc. for assistance in procurement of tall fescue straw. In addition, special appreciation is expressed to Stephanie Falck, Audrey Carlon, Arthur Nyman, Lyle Black, Abe Clark, Nancy Ramsey, Alma True, Anita Holman, and Martha De-Lorme for their assistance in this project. ![]()
3 Corresponding author: dave.bohnert{at}oregonstate.edu
Received for publication February 1, 2007. Accepted for publication June 13, 2007.
| LITERATURE CITED |
|---|
|
|
|---|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |