P1: GFU/GVM Journal of Chemical Ecology [joec]
pp438-joec-369022
April 5, 2002
8:41
Style file version Nov. 19th, 1999
C 2002) Journal of Chemical Ecology, Vol. 28, No. 4, April 2002 (°
DISTRIBUTION OF LOCOWEED TOXIN SWAINSONINE IN POPULATIONS OF Oxytropis lambertii
M. H. RALPHS,1,∗ S. L. WELSH,2 and D. R. GARDNER1 1 USDA/ARS
Poisonous Plant Laboratory 1150 E. 1400 N., Logan, Utah 84341 2 Bean
Natural History Museum Brigham Young University Provo, Utah 84602
(Received October 10, 2001; accepted December 17, 2001)
Abstract—Oxytropis lambertii has been considered to be one of the major locoweeds responsible for livestock poisoning on rangelands, but there has been much confusion as to its taxonomic identity. The objective of this study was to conduct a field survey of several populations of each of the three varieties [var. lambertii Pursh; var bigelovii A. Gray; var. articulata (E. Greene) Barneby] to document the presence or absence of the locoweed toxin, swainsonine. Swainsonine was found at detectable levels (>0.001% dry weight) in only five populations of var. bigelovii in the southwest portion of its distribution in southern Utah, Arizona, and southwestern New Mexico, USA. No swainsonine was detected in populations in the northeast areas of its distribution (eastern Utah, Colorado, northeastern New Mexico, USA). The other varieties, articulata and lambertii, also did not contain swainsonine. It is suspected that a plant fungal endophyte may be responsible for the high variability in swainsonine content in populations of O. lambertii. Key Words—Lambert locoweed, crazyweed, Oxytropis lambertii, swainsonine, livestock poisoning, toxic plant, poisonous plant.
INTRODUCTION
Locoweeds (Astragalus and Oxytropis spp.) cause widespread poisoning of livestock in the western United States (Kingsbury, 1964). Much of the early toxicological research was conducted on the complex of species then called Aragallus lambertii (Marsh, 1909), which included both the white-flowered and the lavenderflowered groups of plants. Later Marsh (1919) referred to the white locoweed as ∗
To whom correspondence should be addressed. E-mail:
[email protected]
701 C 2002 Plenum Publishing Corporation 0098-0331/02/0400-0701/0 °
P1: GFU/GVM Journal of Chemical Ecology [joec]
702
pp438-joec-369022
April 5, 2002
8:41
Style file version Nov. 19th, 1999
RALPHS, WELSH, AND GARDNER
Astragalus lambertii. Butcher et al. (1953) were first to point out that the plant with which Marsh conducted most of his early toxicology research was indeed the white flowered O. sericea. Taxonomists have separated the two species based on the color of their flowers (O. sericea is white to slightly yellow; O. lambertii is purple violet or lavender) and the dolabriform (ox head-like) pubescence that is unique to O. lambertii (Barneby, 1952; Welsh, 2001). The confusion in taxonomy complicates the question as to whether O. lambertii is really a locoweed and whether it has been responsible for poisoning livestock. Literature reviews of the locoweeds all include O. lambertii as one of the major locoweeds (Marsh, 1919; Kingsbury, 1964; James et al., 1981; Ralphs and James, 1999). However, there is no empirical evidence of poisoning attributable to O. lambertii. The objective of this research was to document the presence of the locoweed toxin swainsonine in O. lambertii. METHODS AND MATERIALS
A literature review was conducted of the toxicology and chemistry literature for any reference to swainsonine in O. lambertii. A widespread field survey of O. lambertii populations was conducted and laboratory analysis performed to determine the presence or absence of swainsonine. Voucher specimens are held in the S. L. Welsh Herbarium in the Bean Natural History Museum, Brigham Young University, Provo, Utah. Known populations of each of three varieties representing the entire geographical distribution of that variety were sampled in June 2000 (Table 1). Ten individual plants were collected at each site. Half of each plant was pressed for identification. The other half was air dried in a forced air oven for 48 hr at 60◦ C and then ground through a cyclone grinder. Samples were analyzed for the locoweed toxin swainsonine by the procedures described by Gardner et al. (2001). Detection limit of swainsonine was 0.001% of dry weight. Swainsonine concentration was compared by one-way analysis of variance to determine differences between locations. RESULTS AND DISCUSSION
There are few reports in the literature of O. lambertii containing swainsonine (Table 2). Molyneux et al. (1989, 1991) reported the presence of swainsonine in O. lambertii by using a thin-layer chromotography technique. Later samples from the Henry Mountains in southeast Utah, and from Livermore, Colorado, showed swainsonine levels of 0.031 and 0.007%, respectively (Molyneux, unpublished data). Fox et al. (1998) reported O. lambertii from western and northern New Mexico contained swainsonine. Braun (1999) also reported swainsonine in
P1: GFU/GVM Journal of Chemical Ecology [joec]
pp438-joec-369022
April 5, 2002
8:41
Style file version Nov. 19th, 1999
703
SWAINSONINE IN LOCOWEED
TABLE 1. VARIETIES OF Oxytropis lambertii VAR. bigelovii, STAGE OF GROWTH AND SAMPLING SITES Variety/ voucher
Stage
bigelovii 27661a
Seed
Location
37◦ 060 1900 N 111◦ 510 2800 W Flagstaff, AZ 37◦ 060 4100 N 111◦ 510 2800 W Springerville, AZ 34◦ 000 4900 N 111◦ 100 4800 W Kingston, NM 32◦ 520 5100 N 107◦ 510 5500 W Winston, NM 33◦ 210 4300 N 107◦ 340 4100 W Ocate, NM 36◦ 150 1100 N 105◦ 020 3200 W Capulin, NM 30◦ 410 2500 N 104◦ 080 3500 W Sophia, NM 36◦ 280 0600 N 103◦ 590 5400 W Ft. Collins, CO 40◦ 560 3900 N 105◦ 150 3300 W Ferron, UT 39◦ 060 5700 N 111◦ 170 3600 W Kanab, UT
27665a
Veg
27667a
Veg
27668a
Veg
27669a
Bud
27672
Flo
440981
Flo
440982
Flo
440980
Flo
440983
Flo
articulata 27689
Pod
Knowles, OK
27697
Pod
Buffalo, OK
27698
Pod
Meade, KA
lambertii 27704
Flo
Sidney, NB
27717
Flo
Hot Spring, SD
27721
Flo
Lusk, WY
a
GPS
Soil/Site
Sandy hills
Plant community
Desert shrub
Sandy loam foothills Gravely hills
Ponderosa pine openings Pinyon/juniper woodland Rocky, mountain Ponderosa pine/alligator slopes juniper forest l´oamy valley Blue grama grassland Gravely hills Ponderosa pine Rocky loam sidehill L´oamy flats
Shortgrass prairie Shortgrass prairie
Gravely foothills Mixed grass prairie Sandy mountain bench
36◦ 550 1500 N Rolling hills 100◦ 180 2300 W 36◦ 480 4400 N Rolling hills 99◦ 460 2200 W 37◦ 100 0900 N Calcareous 100◦ 230 0300 W breaks 41◦ 090 1800 N Limestone bluff 103◦ 050 1800 W 43◦ 240 3500 N Gravely hills 103◦ 200 2300 W 43◦ 050 1200 N Sandy/gravel 104◦ 190 3600 W hills
Low sage, bluebunch wheatgrass Tall grass prairie Tall grass prairie Tall grass prairie
Mixed grass prairie Ponderosa pine, mixed grass prairie Mixed grass prairies
Populations containing swainsonine.
O. lambertii in Grant County, New Mexico, that was associated with a fungal endophyte. A recent grazing study comparing the relative palatability of O. lambertii with O. sericea growing together on short-grass prairie in north-central Colorado, showed that swainsonine was not present in detectable amounts in samples of
P1: GFU/GVM Journal of Chemical Ecology [joec]
pp438-joec-369022
April 5, 2002
704
8:41
Style file version Nov. 19th, 1999
RALPHS, WELSH, AND GARDNER TABLE 2. REPORTED ACCOUNTS OF O. lambertii CONTAINING SWAINSONINE Date
Location
Reference
Swainsonine (%)
1986 n.d. 1989 1987 1991 1991 1999 1999
Henry Mt., Wayne Co., UT Livermore, Larimer Co., CO no location given Winston, Sierra Co., NM Winston, Sierra Co., NM Ocate, Moro Co., NM Grant Co., NM Colfax Co., NM
Molyneux (unpubl.) Molyneux (unpubl.) Molyneux et al. (1991) Fox et al. (1998) Fox et al. (1998) Fox et al. (1998) Braun (1999) Braun (1999)
0.031 0.007 >0.01a 0.099 0.021 0.026 0.576 0.0006
a
Presence detected by thin-layer chromotography.
O. lambertii taken from different pastures in 1998 and 1999 (Ralphs et al., 2001). Samples of O. lambertii collected from another research location in northeastern New Mexico also did not contain swainsonine (Ralphs, unpublished data). There are three varieties of O. lambertii. Variety bigelovii A. Gray is separated from the other two by having stipitate pods (slender stalklike base). It occurs from southeastern Wyoming, southward through the foothills and mountains of Colorado and New Mexico, and westward into eastern Utah and Arizona. Variety lambertii Pursh is separated from var. articulata (E. Greene) Barneby by its longer calyx teeth, about one third as long as the tube and shorter flowers,