Invasion of Bromus tectorum L. into Western North America: An ecological chronicle

Agro-Ecosystems 7 (1981) 145--165

145

Elsevier Scientific Publishing Company, Amsterdam -- Printed in The Netherlands

INVASION OF B R O M U S TECTOR UM L. INTO WESTERN NORTH AMERICA: AN ECOLOGICAL CHRONICLE

RICHARD N. MACK

Department of Botany, Washington State University, Pullman, WA 99164 (U.S.A.) (Accepted 12 February 1981)

ABSTRACT Mack, R. N., 1981. Invasion of Bromus tectorum L. into Western North America: an ecological chronicle. Agro-Ecosystems, 7: 145--165.

Bromus tectorum L., the most ubiquitous alien in steppe vegetation in the intermountain West of North America, entered British Columbia, Washington, and Utah ca. 1889--1894. By ca. 1928 the grass had reached its present distribution occupying much of the perennial grasslands in Washington, Idaho, Oregon, Nevada, Utah and British Columbia as native grasses dwindled with overgrazing and cultivation. In the process this cleistogamous winter annual may have competitively displaced both native colonizers (including cleistogamous annual grasses) as well as the dominants of climax stands. The spread of B. tectorum demonstrates the degree of success an alien may achieve when preadaptation, habitat alteration simultaneous with entry, unwitting conformation of agricultural practices to the plant's ecology and apparent susceptibility of the native flora to invasion, are all in phase.

INTRODUCTION

To Darwin one group of examples which best illustrated the inherent "geometrical powers of increase" to be seen in all species involved the rapid spread of some transoceanic immigrants in a new range (Darwin, 1956). Through such transit a species may arrive in a new homeland: (a) to which it is preadapted (Grant, 1977); (b) where habitat modification has occurred simultaneously with alien entry (Harper, 1977); (c) in which its potential competitiors may be inferior (Harper, 1965); and (d) while simultaneously leaving behind in the old range both coevolved predator/parasites and competitors (Elton, 1958). The possible permutations of (a)--(d) provide the basis for intriguing questions as to the role that historical chance and evolutionary product have played in determining new range expansion. Furthermore, such increasingly commonplace events are an opportunity to study biota in flux, the changes that occur and the nature of the new equilibrium among species (Harper, 1977). The first three circumstances, at least, occurred during the 19th century in the semi-arid intermountain West of North America with the entry of Bromus 0304-3746/81/0000--0000/$02.50 © 1981 Elsevier Scientific Publishing Company

146

tectorum L. (downy brome, downychess, cheat and most commonly, cheatgrass). The life history of this winter annual was preadapted to the varied softs and macroclimate (specifically, wet, cold winters and dry, hot summers) of a region under modification. Seldom in the recent transformation of the Earth's vegetation by mobile Occidental man and his plants has the vegetation of such a large area been transformed so swiftly and (apparently) permanently. On > 410,000 km2 (Fig. 1) of the intermountain West, steppe vegetation dominates and extends from southern British Columbia to the southern third of Nevada and Utah between the Rocky Mountains and the Cascade/Sierra Nevada Ranges. Today the alien annual grass, Bromus tectorum L., is the most ubiquitous, and in many rangelands, the dominant plant species. The land area encompassed by what Daubenmire (1978) has called the Agropyron spicatum ~,~"

BRITISH

i _(,~

COLU

11~L@ I

,

(

108" I

:~TA

I

sAsKAT

t

\\ l . .k. . . .

0

\.

100

---f'~"

MONTANA

Walla

OREGON

IDAHO I

b

• Klarr~lh Falls

~

~..........

L.

. . . . . . . . .

ORNIA

~-

. . . . . .

I ,

•

wYOMING

]1w

_

R°

I I

~A"

i

Fig. 1. Spread of Bromus tectorum from 1889 to its modern distribution (post 1930) in the intermountain West. Locations in stippling are based on annotated herbarium specimens (see Appendix) and contemporary accounts (see text). Fig. l a shows railroad system by 1898 (compiled from the U. S. Geologic Survey, 1898 and British Columbia Dep. Agric., 1897). Maximum distribution of B. tectorurn (Fig. l f ) corresponds approximately to the limits of the Agropyron spicatum Province [sensu Daubenmire ( 1 9 7 8 ) ] . Undifferentiated grasslands occur both south and east of the intermountain steppe (shading in circles).

147

Fig. lb.

Province (Fig. If) conforms remarkably well to the new range of B. t e c t o r u m in the treeless sections of western North America. This article attempts to reconstruct with contemporaneous records (ca. 1890--1930}, primarily from the Columbia Plain, the events b y which B. t e c t o r u m spread through this region. This ecological chronicle provides the basis for unraveling the identity of those species likely to have been competitively displaced. Recollections or reminis* cences written after the events chronicled here have deliberately not been used. Furthermore, due to the many c o m m o n names for members of this genus (often with overlapping or inclusive c o m m o n name usage), included in this account are only citations in which I can verify the B r o m u s taxon being referred to. ORIGINAL VEGETATION AND EARLY SETTLEMENT (1850--1904)

Prior to 1850 the Columbia Plain supported a perennial caespitose grassdominated steppe, in which the gradient formed b y effective precipitation centering in autumn--winter was the single most important environmental feature. At its eastern relatively mesic extent the steppe environment supports commun-

148

ities of Agropyron spicatum (Pursh) Scribn. and Smith, Festuca idahoensis Elmer, and an assemblage of broad-leafed dicot forbs. On the steppe's western margin and south into Nevada the habitats are progressively drier and are characterized b y Artemisia tridentata Nutt., Chrysothamnus spp. and Purshia tridentata (Pursh) DC (Daubenmire, 1969). The remoteness of the intermountain West from the eastern United States protected it from all but minor development before 1850. But beginning in the 1850's and continuing sporadically for'the next 20 years, gold and silver strikes were reported north and east of the Columbia Plain. These events initiated habitat modification as the Columbia Plain became an open range for cattle which were driven annually to the mines, via the Okanogan River Valley into British Columbia and east into the mining districts of northern Idaho (Meinig, 1968). As cattle trails were established local overgrazing was inevitable. Such heavy grazing was to have profound consequences in a grassland with an almost complete absence of sod-forming grasses. The soil surface is covered instead with a fragile layer of cryptogams b e t w e e n the clumped grass. Consequently, rapid introduction of cattle and later sheep and feral horses on the

Fig. lc.

149

Fig. l d .

open range of the Columbia Plain resulted in more soil surface alteration than comparable settlement on the Great Plains to the east. Also unlike grasslands in mid-continent, bison had been sparse in the Agropyron spicatum province for the previous 2500 years (Gustafson, 1972). Members of the so-called "Stevens r e p o r t " (results of a reconnaissance in 1853--55 for the transcontinental railroad route through the region) reported that they did not see antelope or buffalo nor did t h e y find these animals to be c o m m o n to the indigenous people (Cooper, 1860}. Large congregations of ungulates with their trampling activity and repeated movement over the same terrain had been u n k n o w n Nevertheless, b y the time of the General Land Survey in 1881 surveyors were c o m m o n l y including in their section notes, descriptions such as "hundreds of parallel and diverging trails too numerous for topographic record". The cryptogam layer was being broken to a much greater extent than ever before by livestock hooves causing a proportional increase in sites of disturbance. Homesteading rapidly increased with the railroad b o o m beginning in the early 1880's. Land which had been open range before 1880 was parcelled into 160 acre sections, fenced and planted to spring wheat. As with the response to grazing activity earlier, the vegetation responded to plowing in a manner

150 different from that in the Great Plains. The generally friable soil covered with caespitose grasses (e.g.A. spicatum and F. idahoensis) offered comparatively little resistence to plowing. Plant volunteers would have been minimal. By 1890 the "improved" farm acreage in the Columbia Plain was in excess of two million acres (Meinig, 1968). Alien bromes (e.g. Bromus brizaeformis Fisch and Mey, B. secalinus L., and B. ciliatus L.) (Sandberg and Leiberg, n.d.) began to appear with agricultural settlement in the Columbia Basin and Great Salt Lake Basin. I can only speculate as to their mode of entry into these areas, for there would have been ample opportunity in an era prior to any inspection of goods for pests. For example, seed grain was commonly adulterated with weed seeds by unscrupulous grain dealers (Washington Dep. Agric., 1914). Beginning in 1878 and extending well into this century, William Suksdorf travelled extensively in south central Washington and along the lower Columbia River. Because he was a professional plant collector preparing lots of dried specimens for sale, he collected all vascular plants without a p p a r e n t taxonomic restriction. By 1882 he had collected Bromus mollis L. at White Salmon, Washington [W. N. Suksdorf (s.n.) WS] and by 1884 he had found

Fig. le.

151

Fig. lf.

B. brizaeformis at Spangle, Washington [W. N. Suksdorf (s.n.) WS]. His "Flora Washingtonensis" of 1892 lists 4 alien bromes; B. ciliatus, B. mollis, B. racemosus L. and B. secalinus within south central Washington. A herbarium specimen of B. secalinus from North Yakima in 1892 includes the notation " t o o c o m m o n in wheat fields" [L. S. Henderson (s.n.) WTU]. Similarly, B. mollis [F. Leonard (s.n.) Hot Spring Lake, RSA] and B. secalinus [M. E. Jones, 1009, Salt Lake City, RSA] were collected in Utah in 1884 and 1879, respectively. With the advent of agriculture, alien weeds were appearing for the first time. B. mollis was apparently most prevalent within this cluster of newly arrived alien bromes in the Washington steppe. Shear (1901) writing of the Columbia Plain reported that " . . . . . bunch grasses have been practically exterminated over large areas and their places occupied more or less b y w e e d y annual plants, especially the soft chess [Bromus hordeaceus (molIis)] . . ." -- a view supported b y Sampson (1917) based on his observations in 1907 in the nearby Wallowa Mountains, and b y Henderson (1903) for the steppe in Idaho. The earliest flora of the Palouse region lists B. mollis as "very c o m m o n " in a 35 km radius of Pullman, Washington (Piper and Beattie, 1901). Elmer (1899) de-

152

scribed the grass in the vicinity of Pullman as " . . . only t o o c o m m o n l y found in house lots and waste places". Cotton (1904b) listed soft chess as "becoming quite abundant in some of the worn-out parts of the [Washington] range, while in others it is just appearing". More specifically, he observed that B. mollis was among the prevailing grasses on the hills bordering the Columbia River in southern Klickitat County, an area of extensive early grazing activity. B. brizaeforrnis was also locally abundant (Brodie, 1898; Piper and Beattie, 1901). The deteriorated range conditions which had fostered the initial spread and establishment of these alien species had apparently b e c o m e critical b y 1896, when the Northern Pacific Railroad ended an open range policy on its holdings in Washington in an effort to stop the evident range abuse (Cotton, 1904b). Nevertheless, there was considerable pessimism by federal range investigators and laymen alike (Brodie, 1898). Cotton (1904a) concluded that "at the present time nearly all the perennial grasses have been destroyed. There are, however, enough to furnish a seed source". "An Illustrated History of Whitman C o u n t y " (1901) claimed that " . . the native grass is gone, and experiment has not yet fully demonstrated the adaptability of any other grass to this soil and climate". These gloomy assessments are all the more remarkable when it is recalled that only 40 years had elapsed since the first sustained herds of livestock had been kept anywhere on the Columbia Plain (Cotton, 1904b; Meinig, 1968). Circumstances were similar in Oregon (The Resources o f the State of Oregon, 1892) and in British Columbia. Prospective settlers concerned about disappearance of the famed 'Bunch grass' in interior British Columbia were being reassured that "another class of vegetation, no less nutritious has succeeded it" (Anderson, 1882b). B. t e c t o r u m had arrived b y 1889 in the interior Pacific Northwest. Macoun (1890) recorded the grass as "introduced in meadows and cultivated fields at Spence's Bridge [British Columbia] ". [J. Macoun (s.n.) CAN]. It is unlikely that B. t e c t o r u m had arrived in the grasslands of southern British Columbia much earlier as Macoun (1877) had collected extensively in the same area in 1875, but had not found B. tectorurn (although he did collect the aliens B. ciliatus and B. r a c e m o s u s in interior British Columbia). The grass was collected in Washington north of Ritzville along the Great Northern Railxoad r o u t e in 1893 [J. H. Sandberg and J. B. Leiberg, 191, US, G. F. Russell, pers. comm.] and at Provo, Utah in 1894 [M. E. Jones, 5503, R S A ] . All three locations were noncoastal wheat growing districts at that time (Anderson, 1882a; Hollister, 1882; Meinig, 1968), strongly suggesting that B. t e c t o r u m first entered the region as a grain contaminant. Alternatively, there is no evidence that B. t e c t o r u m established first along the Pacific Coast and then proceeded inland along lines of commerce. For example, much commerce moved via the Columbia River Valley from the time of initial European contact. While B. t e c t o r u m was being collected repeatedly on ballast dumps at Portland, Oregon, beginning in 1902 [no collector given, CAS], Suksdorf, living up-river at Bingen, Washington, did not collect it east

153

of Hood River, Oregon until 1912 [W. N. Suksdorf, 7526, WS]. It is still possible however, that the grass was carried from the coast into interior Washington and British Columbia in a single event. Several other possible entry routes are also unlikely. A northward expansion of range from California or Nevada is improbable on two grounds: (i) all the collections of B. tectorum I have seen from central Oregon southward into Nevada are post 1900; and (ii) neither Tracy (1888) nor Griffiths (1902; 1903), traversing the Great Basin, reported B. tectorum, although both collectors were particularly concerned with assessing the status and diversity of grasses. Tracy (1888) reported B. mollis, B. secalinus, and B. racemosus found along his route from Reno, Nevada to Salt Lake City via northern Nevada. Griffiths (1902) reported B. secalinus, B. mollis and B. rubens L. occurring locally along Bartlett Creek in northern Nevada. Instead B. tectorum spread from east to west across the Great Basin from one or more entry points in the valley of the Great Salt Lake Basin (Pammel, 1910; 1915). Entry (other than in contaminated goods) from the east across the northern Rocky Mountains prior to 1900 also seems unlikely. Pammel (1897) found B. tectorum locally in the foothills of eastern Colorado, but the grass was not among his list of introduced and new plants in the Bitterroot River Valley near Hamilton, Montana (Pammel, 1905). By 1902, Cusick [W. C. Cusick, 2844, WS] found B. tectorum in cultivated fields in the Klamath Valley of southern Oregon, another early wheatgrowing district (The Resources of Southern Oregon, 1890). There is no evidence that the plant had spread that far south from the Washington-Oregon border by 1900, although it may have been carried there in grain. As it had been recognized by the early 1890's that the rangelands in the Pacific Northwest had been badly overgrazed, an effort was mounted to find more resistant species or varieties (Lamson-Scribner, 1899). B. tectorum was deliberately introduced at least once during this turn of the century search for new grasses for the overgrazed ("exhausted") range. Seed sent from the U. S. Department of Agriculture in Washington was sown on April 15, 1897 at the college experimental farm at Pullman, Washington. A thick sward had formed by June, but none of the plants reportedly flowered (Brodie, 1898). Since B. tectorum is a winter annual with mainly autumn germination, it is quite likely that not all the seed germinated in the spring trial, but rather emerged unnoticed the following autumn. Consequently, Pullman may have been an additional focus from which cheatgrass spread. Hedrick (1965) cites a deliberate introduction about 1915 by a peddler extolling the virtues of a new "100-day" grass which may have caused the wholesale introduction of the plant in Wallowa County, Oregon. Collections of B. tectorum became increasingly frequent by 1898 in eastern Washington at Spokane [C. V. Piper (s.n.) US], Pullman [D. A. Brodie (s.n.) WS], and Pasco [A. D. E. Elmer, 1047, US] and in northeastern Oregon in 1897 (Cusick, 1908). In part this was certainly a function of there being more plant collectors at the College and Experiment Station at Pullman, but the locations for other bromes collected during this time strongly suggest that

154

b y 1900 the flora of the Columbia Plain was being collected comprehensively. The range of the alien b y ca. 1900 (Fig. l b , c) appears to have been still quite small and confined to local populations. While B. tectorurn was apparently slowly increasing its range between 1889 to ca. 1900, there is little indication in the reports of the time that the observers recognized the plant's potential for spread. Piper and Beattie (1901) referred to B, tectorum as "sparingly introduced in waste places" in eastern Washington. Cotton (1904a) reported that in Washington, "the plant has become very abundant along the railroad right-of-way between Toppenish and Wapato" b u t that it was seldom seen elsewhere. Griffiths' (1903) only reference to B. tectorum in his regional survey of range conditions in 1902 was for the Okanogan River Valley of Washington where " the bulk of forage is [various native perennial grasses] . . . with introduced Bromus tectorum and chess (Bromus secalinus) in the lower draws and benches, where the native grasses have been injured b y the trampling of s t o c k . . . " . The trails made during the 19th century mining b o o m supported B. tectorum by 1900. Whether cheatgrass had b y that time migrated south from early British Columbia locales or north from the Columbia Plain is unresolvable. Only Elmer (1899) observed that is was "making rapidly its h o m e in dry arid lands" west of Pullman. Two main types of agro-ecosystems (rangeland and cereal grain fields) were developing in the intermountain West b y 1900. B. tectorum was well suited to both newly created environments. Cheatgrass and other annual bromes, were weeds of wheat fields long before they arrived in the West (Brady, 1852; Wood, 1863). In addition, the wide ecologic amplitude of the grass allowed entry in rangeland. This dual ability to reside in both agro-ecosystems allowed the alien, unlike many other weeds, to persist even as land was converted from one use to another. 1905--1914

The years 1905--1914 (Fig. l d ) bracket the decade in which B. tectorum first became b o t h widespread and locally abundant. B. tectorum was prominent in northeastern Oregon b y 1910 (Cusick, 1908; 1909) in the Great Salt Lake Basin b y 1912 (Pammel, 1915; Kearney et al., 1914) and in Idaho (Osborne, 1911), both in the Snake River Plain as well as adjacent to the Columbia Plain. Man's activities were responsible not just for the weed's introduction, b u t also it seems for determining the rate and m o d e of its range expansion. With virtual completion of the railroad system in the northwest b y 1900 (Fig. la), seed of B. tectorum could have been spread rapidly over the region mixed in grain, in the discarded cattle bedding straw along railroad sidings (Henderson, 1898), in the dung of the transported cattle in which it remains viable (Maser, 1974), or as the discarded straw packed with dry goods (Craig, 1892). The morphology of the seed and spikelet aid in such transport. Ten of the eleven alien bromes found t o d a y in the Pacific Northwest (Hitchcock et al., 1969) have caryopses with an awned lemma attached, but B. tectorum

155

is especially prone to dispersal b y animals as the rachilla disarticulates above the glumes, with several caryopses plus the rachilla being a c o m m o n dispersal unit. The resulting multi-awned disseminule is readily attached in fur, especially wool. Spread and maintenance of B. tectorum in agricultural fields was aided b y allowing the chaff from stationary threshing machines to collect into piles in the fields. Livestock then grazed in these harvested fields in autumn (Meinig, 1968). This practice had the two step effect of always leaving a residual seed supply in the fields, b u t more importantly, creating through the trampling activity in fallow fields, safe sites for germination in autumn. Additional impetus to the spread of B. tectorum came with the change from spring to winter wheat cultivation which had reached its apex b y ca. 1915. Before the advent of widespread farm mechanization, there had been reluctance to sow winter wheat, as any delay in autumn harvest jeopardized the time for sowing the next crop (Meinig, 1968). However, as early as 1884, 80% of the crop in some districts was winter wheat (Greene, 1888) and the practice expanded as new grain varieties were developed. Consequently, fields were being plowed and prepared as small grain seed beds in the phenologically appropriate time for maximum cheatgrass germination. Much of the early range° land was being plowed for the first time. Widespread adoption of winter wheat agriculture and its accompanying practices, ensured that much of the intermountain steppe region not in rangeland would be quickly subjected to cheatgrass invasion. Once a field was contaminated, the grass proved to be a persistent occupant, while on rangeland, cheatgrass largely replaced the once dominant perennial caespitose grass (Daubenmire, 1970). Decimation of cheatgrass b y fire or smut damage is short term, as recruitment continues from the residual seed reserve in the soil (Klemmedson and Smith, 1964). This persistence of annual bromes as a group in the seed bank of grain fields led unknowing farmers to think their wheat seed had degenerated into a weed, and in the parlance of the day, the wheat had " c h e a t e d " the farmer (Piper, 1920). 1915--1930

B. tectorum was becoming b o t h a widespread and dominant weed b y 1915. First mention of B. tectorum persisting in large numbers is found in the reports of county agents in Washington. Not surprisingly, alfalfa fields in the Okanogan Valley, identified as an early locale for the alien (Griffiths, 1903), were becoming infested with cheatgrass along with B. secalinus and B. sterilis (Hughes, 1915). B. tectorum was also becoming prominent in alfalfa fields in southern Benton county. In a series of annual reports beginning in 1913, the c o u n t y agent chronicles the growth of the problem (Wash. State Univ. Coop. Ext. Serv., County Agent Rep., 1914). The agent states that by 1914 "thousands of acres are being almost taken by this 'cheat'. Everyone is alive to the seriousness of the p r o b l e m . . . " . Since the detailed reports of weed seed con-

156 taminants b y the State Seed Analyst for 1914--1916 (Wash. Dep. Agric., 1916) show no cheatgrass seed in commerical lots of alfalfa, the spread was probably accelerated unwittingly by the unregulated exchange locally of contaminated alfalfa seed. South of Benton County in adjacent Oregon, Peck (1916) decribed B. tectorum b y 1915 as "the most plentiful grass, distributed in enormous abundance over almost the entire area s t u d i e d . . . " [Umatilla County]. The burgeoning problem was detected first in alfalfa because as a hay crop it was cut near the ground in midsummer, thus incorporating the already mature cheatgrass, while the relatively tall wheat varieties of that era (i.e. Little Club, Fife and Bluestem; Meinig, 1968) were cut above the height of most mature B. tectorum individuals. However, B. tectorum was found even in harvested wheat. Alvord (1918) found cheatgrass among 40 weeds he collected in seed lots from grain elevators in Washington in 1916--1918. Results of the incomplete returns of a state-wide survey of the 10 worst weeds as viewed b y county agents in 1916 show that in the steppe counties of Spokane and Benton, B. tectorum was already a serious pest, while in Douglas and Grant counties it was not as yet detected (Wash. County Agents' letters to I. D. Cardiff, 1916). Shantz (1916) provides photographic documentation of the spread of cheatgrass to Franklin c o u n t y in south central Washington. In several of the photographs (P-4 and U-4) near railroad lines at Connell and Pasco, respectively, a dense sward of B. tectorum can be clearly seen in the foreground, an identification supported b y his notes for these photographs. Elsewhere, B. tectorum was apparently not yet as conspicuous. Among bromes Nelson (1914) found only B. mollis and B. brizaeformis c o m m o n in "waste places" in southern Chelan county on the western border of the Columbia Basin. Neither Weaver (1917) nor Drum (1916) list it among ruderal or w e e d y species for the relatively moist meadow steppe communities (Festuca idahoensis/Symphoricarpos albus zone, sensu Daubenmire, 1970). I conclude that the plant while present was still in low numbers, since B. tectorum was regularly being collected in these eastern Washington counties (chiefly Whitman and adjacent Latah County, Idaho). The contemporary published floras of eastern Washington and southern British Columbia provide further evidence of the plant's spread. "The Flora of Southeastern Washington and Adjacent Idaho" b y Piper and Beattie (1914), an updated and geographically expanded version of their 1901 volume, continues to list B. tectorum as "sparingly found in waste places", while the abundance descriptions of many other species were revised. But, Henry (1915) lists B. tectorum as "very abundant in the dry regions along the Fraser and Thompson (Rivers); Penticton". In the 25 years which had elapsed since it was first collected at Spence's Bridge, B. tectorum had apparently invaded much of the steppe in British Columbia, while still only locally abundant in north and south central Washington. The plant was also becoming prominent at the southern limits of its new range. Shantz (1925) described the sagebrush lands of Utah and northern

157

Nevada as including B. tectorum. The time span 1919--1930 otherwise provides little documentation of the invasion. 1930--PRESENT

The lack of information for the time span 1919--1930 is all the more unfortunate, as b y 1930, B. tectorum seems to have occupied its current range, like a group of coalescing "leopard spots" which rapidly covered the Agropyron spicatum Province (Fig. le). (Forsling and Dayton, 1931; Daubenmire, 1942; Fleming et al., 1942; Humphrey, 1945; Tisdale, 1947.) The vegetation had changed to such an extent that references appear b y 1936 to "cheatgrass (B. tectorum) lands" in central Washington (Wash. State Univ. Agric. Exp. Stn., 1937). Forsling and Dayton (1931) state that throughout Oregon, Washington, Idaho, Nevada and Utah, A. spicatum ". • • • has practically disappeared from much of the sagebrush t y p e where such use [heavy grazing] has prevailed, being replaced by d o w n y chess (Bromus tectorum)". By 1928 large tracts of abandoned farm land or overgrazed rangeland in the Snake River Plain of Idaho supported " . . . immense uniform areas of d o w n y bromegrass on burned-over sagebrush lands" (Piemeisel, 1932). Pickford (1932) considered B. tectorum as the only important annual grass in the disturbed areas once dominated b y A. tridentata in northern Utah. While it does not seem possible to d o c u m e n t the plant's spread in detail, there was a definite acceleration of range expansion b e t w e e n i 9 1 5 - - 1 9 3 0 . A plot of estimated area occupied versus time (ca. 1900--1930) yields a sigmoidal curve (Fig. 2). These same area estimates also show the straight line relationship of a " c o m p o u n d interest disease" in the plot of In (x/(1-x)) where x is the decimal equivalent of eventual range occupied at time t (Van Der Plank, 1963). Like an epidemic, B. tectorum was limited in its initial expansion b y a lack of inoculum (only small populations) and numbers of foci (points of entry). The grass was limited in ca. 1915 b y neither inoculum nor tissue to be infected (available territory). Finally, b y 1930 the lack of suitable unoccupied territory restricted further range expansion. Leopold (1941) best characterized 1.0(]

"6 "~ ~ 0.75 .~" ¢,~ -"1

~'~ o~ i~ I..LI 0.25 ff

1900

1920

1910

1930

Year Fig. 2. Progress o f B. t e c t o r u m as a p l o t of area o c c u p i e d versus t i m e b a s e d on accumulative percentage o f eventual range occupied by 1 9 0 0 , 1 9 0 4 , 1 9 1 4 , 1 9 2 0 a n d 1 9 3 0 .

158 the impression gained from the range expansion during this period that ". one simply woke up one fine spring to find the range dominated b y a new w e e d " -- cheatgrass. DISPLACEMENT OF THE NATIVE COLONIZERS

B. tectorum ultimately occupied t w o general groups of niches in western North America: occupation of the first is well documented, as a persistent resident of sites once occupied b y perennial caespitose grasses (Fig. l f ) (Forsling and Dayton, 1931; Daubenmire, 1942; Fleming et al., 1942; Humphrey, 1945; Tisdale, 1947) and of a second group, colonizing grasses. Given the plant's role in Eurasia as a member of ruderal communities (Braun-Blanquet et al., 1952; Bor, 1970) and sand dune succession (Mirov, 1951), display of the colonizer's role in North America is to be expected. Furthermore, as a cleistogamous winter annual it displays more than half of the 12 features Baker (1974) considers ideal weed characteristics. Yet its success as a colonizing species in western North America has been so phenomenal as to almost exterminate the evidence - - t h e native weedy grasses. There are few observations before this invasion on which we may rely, b u t these records are nevertheless useful. Leiberg (1897) observed that the region's only native brome, Bromus carinatus H. and A., was abundant wherever settlements had been established in the low elevation forest adjacent to steppe in northern Idaho. Griffiths (1903) and more recently Reid (1942) noted large quantities of B. carinatus in poorly cultivated fields or other disturbed sites in the low elevation sites in the Blue Mountains of southeastern Washington and adjacent Oregon. B. carinatus t o d a y is found in many steppe communities with a variable life form. Under the t e m p o r a r y conditions found on disturbed sites it is cleistogamous (Harlan, 1945). Stebbins (1957, 1974) has emphasized the possible origin of inbreeding in species that o c c u p y temporary, colonizing habitats. To Stebbins (1957) facultative self-pollinators in Eurasia (similar to the North American B. carinatus) were intermediate stages b e t w e e n self-incompatible species and those which are regularly self-pollinated, including B. tectorum. Was a similar phylogeny proceeding in western North American bromes (the rate limited only b y the comparative lack of tempora~-y site-producing ungulates), only to be supplanted b y the product of similar selection in Eurasia in the form of B. tectorum? Festuca may have behaved similarly. Two cleistogamous winter annual fescues, F. microstachys Nutt. and F. octoflora Walt., n o w restricted to minor roles in a variety of steppe communities may have been important colonizers before alien entry. Cotton (1904a), who surveyed the range conditions over much of the arid steppe in central Washington in 1902, found F. octoflora and F. microstachys as increasers with over-grazing. Cotton's report includes a photograph of an Artemisia tridentata stand badly over-grazed b y sheep. T o d a y a stand with such disturbance would be dominated b y B. tectorum, b u t in

159 1902 Festuca octoflora (as noted in the photograph's caption) was abundant with such disturbance. Sandberg and Leiberg's (n.d.) 1893 botanical survey across the northern half of the Columbia Basin agrees with Cotton's observation on the role of F. microstachys with disturbance. The genetic variability displayed b y the F. microstachys complex (Kannenberg and Allard, 1967) seems consistent with the plant's role of colonizer before competitive displacement. CONCLUSIONS

A recurring question in the history of what Elton (1958) has called "the great historical convulsions" of the Earth's biota through invasion, addresses the reasons for the range expansion b y transoceanic immigrants. The evidence collected here supports the argument for functional displacement of the colonizers, F. octoflora, F. microstachys, and B. carinatus b y alien bromes beginning in the late 19th century. The caespitose form of A. spicatum, a colonizer of bare sites in the meadow steppe of Washington, was probably similarly affected (Daubenmire, 1960). This transformation was set into motion by: (a) initially, a marked increase in the level of microsite disturbance by livestock; (b) the later, but continuing conformation of agricultural practice to the phenology of the recently introduced alien grasses; which (c) were preadapted to the major climatic features of the region. Whether this is a case of the native species not adapting rapidly to the changing habitat, or their niche overlap being t o o extensive with the aliens with which they competed poorly, has not yet been determined. However, as the participants are now known an experimental solution is possible. Role diminution or range reduction in itself is not an unusual phenomenon in weed history. Salisbury (1961) cites many examples of weeds now rare in the British flora which were once common. But the number o f species affected and the swiftness with which this process was accomplished in the Agropyron spicatum Province is remarkable. In the 80 years after this process began, the role of the two native annual fescues, for example, has been so irreparably altered that without the historical records, I d o u b t that an accurate list of the native w e e d y grasses could have been reconstructed. Wholesale alteration via alien establishment is not typical of grasslands in North America, although the transformation of the Interior Valley of California parallels the events chronicled here (Burcham, 1956). In the Great Plains where agriculture has certainly been intensive, and there has been long term o p p o r t u n i t y for alien invasion, the sere on abandoned or grazed land may include aliens, but not to the exclusion of native colonizers (Costello, 1944; Mentzer, 1950; Weaver, 1958). Furthermore, except for the local prominence of Poa pratensis L. (McInteer, 1952; Dix, 1959), natives soon dominate the mature c o m m u n i t y after succession. I can only speculate at this time as to the possible role that b o t h long association with herding ungulates and the prominence of the rhizomatous life form among the Great Plains grasses has played

160 in their resistence to alien invasion. In this regard, intriguing questions remain as to the historical and evolutionary events which eventually led to the expression of the "geometrical powers of increase" for B. t e c t o r u m in the intermountain West. ACKNOWLEDGEMENTS I thank the herbaria of California Academy of Science; Central Experimental Farm, Ottawa; Gray Herbarium, Harvard University; Missouri Botanical Garden; National Museum of Canada; New York Botanical Garden; Oregon State University; Provincial Museum of British Columbia; Rancho Santa Ana Botanical Garden; Smithsonian Institution; University of British Columbia; University of California, Berkeley; University of Idaho, Botany; University of Idaho, College of Forestry and Range Sciences; University of Minnesota; University of Nevada; University of Utah; University of Washington; University of Wyoming; Utah State University; Washington State University; and the Museum of the University of Oregon, which provided specimens collected during the 40 year span examined here. I thank the following who aided at various stages in the study; T. Abraham, T. Brayshaw, V. Brink, J. Christy, R. Daubenmire, R. Fisher, E. Franz, A. J. Gilmartin, F. C. Hall, G. Harris, J. D. Helms, D. W. Hedrick, E. Kukla, C. McQuade, C. J. Mason, J. Mastrogiuseppe, R. Murray, R. Old, B. Roche, G. F. Russell, J. Singleton, F. Sneva, H. St.John, J. N. Thompson, J. Wacquant, and D. Wagner. Support was provided by National Science Foundation Grant DEB76-80952. APPENDIX

Annotated specimens of B r o m u s t e c t o r u m L. cited in the text and/or used in compiling Fig. 1 are: J. Macoun (s.n.), Spence's Bridge, B. C. 1889 (CAN); J. H. Sandberg and J. B. Leiberg, 191, N of Ritzville, Washington 1893 (letter Sandberg to Beal 1893, Archives, Smithsonian Inst.) (US); M. E. Jones, 5503, Provo, Utah 1894 (RSA); D. A. Brodie, (s.n.), Pullman, Washington 1898 (WS); A. D. E. Elmer 1047, Pasco, Washington 1898 (US); C. V. Piper, (s.n.) Spokane, Washington 1898 (US); D. Griffiths and J. S. Cotton, 7, Prosser, Washington 1902 (WS); No collector given, Portland, Oregon 1902 (CAS); W. C. Cusick, 2844, Klamath Valley, Oregon 1902 (WS); L. N. Goodding, 1106, Nephi, Utah 1902 (RSA); J. S. Cotton, 1066, S of Wallula, Washington 1903 (WS); B. Hunter, (s.n.) LaGrande, Oregon 1904 (WS); W. N. Suksdorf, 644-ORE, The Dalles, Oregon 1904 (WS); M. Arnold (s.n.), Skelton, Nevada 1905 (RENO); P. A. Rydberg, 6023, N of Salt Lake City, Utah 1905 (RM); M. E. Jones, (s.n.) Theodore, Utah 1908 (RSA); C. P. Smith, 1696, Logan, Utah 1909 (UTC); Mrs. J. S. Thompson (s.n.), Tuscarora, Nevada 1909 (RENO); J. M. H. (s.n.), Paterson, Washington 1910, (WS); J. F. Macbride, 252, Boise, Idaho 1910 (UC); G. B. Mains, 12, Emmett, Idaho 1910 (RM); R. K. Beattie, (s.n.) mouth of Walla Walla River, Walla Walla County, Washington 1911 (WS); R. K. Beattie,

161

(s.n.) Attalia, Washington 1911 (WS); D. Bennitt, 47, Twin Falls, Idaho 1912 (RM); A. A. Heller, 10488, Reno, Nevada 1912 (RENO); W. N. Suksdorf, 7526, Rattlesnake Creek, Klickitat County, Washington 1912 (WS); E. Wilson, 1061, Summerland, B. C. 1912 (UBC); E. Wilson (s.n.) Penticton, B. C. 1912 (UBC); C. R. Ball, 1766, Nephi, Utah 1912 (MIN); H. T. Darlington, (s.n.) Tucannon River, Columbia County, Washington 1913 (WS); Mrs. R. D. Cooper (s.n.) near Bend, Oregon 1914 (OSC); B. S. Martineau, 46, Humboldt County, Nevada 1915 (RENO); D. Packard (s.n.) Provo, Utah 1915 (UTC); N. J. Billings (s.n.) near Wallowa, Oregon 1915 (OSC); H. J. Rust, 665, Spencer, Idaho 1916 (CAS); E. M. Warren (s.n.) Vernon, B. C. 1917. (UBC); W. E. Lawrence, 592, 13~ miles E of Heppner, Oregon 1917 (CAS); O. Mink, 13, Approx. 1 mi. N of Jarbidge, Nevada 1917 (OSC); R. S. Ferris and R. Duthie (s.n.) Washoe County, Nevada 1919 (CAS); W. E. Lawrence, 2246, 7 miles S of Bend, Oregon 1919 (CAS); R. S. Ferris and R. Duthie, 375, Summer Lake, Lake County, Oregon 1919 (RM); R. S. Ferris and R. Duthie, 635, Mitchell, Oregon 1919 (CAS); W. E. Lawrence, 2336, Hermiston, Oregon 1919 (CAS); K. Whited, 209, Redmond, Oregon 1920 (WS); I. C. Otis, 930, Leavenworth, B. C. 1920 (UC); J. A. Harris, C2062, S of Grantsville, Utah 1920 (MIN); Mrs. M. E. Soth, P-83, Pocatello, Idaho 1921 (RM); S. V. Wheaton (s.n.) Curlew, Washington 1921

(ws). REFERENCES Alvord, E. D., 1918. A Study of Washington Weeds. MSc Thesis, Wash. State Univ., Pullman, 55 pp. Anderson, A. C., 1882a. The British Columbia Directory, 1882--1883. R.T. Williams, Victoria, 402 pp. Anderson, A. C., 1882b. A Brief Account of the Province of British Columbia, its Climate and Resources; An Appendix to the British Columbia Directory, 1882--1883. R. T. Williams, Victoria, 33 pp. An Illustrated History of Whitman County, 1901. Walt Lever, 469 pp. Baker, H. G., 1974. The evolution of weeds. Annu. Rev. Ecol. Syst., 5:1--24. Bor, N. L., 1970. Gramineae. Flora Iranica No. 70. Brady, J., 1852. Thirty-second Congress, First session, House Exec. Doc. 102, 650. Braun-Blanquet, J., Roussine, N. and Negre, R., 1952. Les Groupments Vegetaux de la France Mediterraneenne. (Prodrome des Groupements Vegetaux de la France). Centre National de la Recherche Scientifique, C.N.R.S., Paris. British Columbia Department of Agriculture, 1897. Fifth Report, 1895--96. Sketch Map. Brodie, D. A., 1898. Grasses of Washington. Bacc. Honors Thesis, Wash. State Univ. In: Manuscripts, Archives, Special Collections, Wash. State Univ. Libraries, Pullman, 203 pp. Burcham, J., 1956. Historical backgrounds of range land use in California. J. Range Manage., 9: 81--86. Cooper, J.G., 1860. No. 1 Report on the Botany of the Route. Part II. Botanical Report. In: Reports of Exploration and Surveys to Ascertain the Most Practicable and Economical Route for a Railroad from the Mississippi River to the Pacific Ocean, Made Under the Direction of the Secretary of War, 1853--55. Vol. 12. Washington, DC, pp. 13--39. Costello, D. F., 1944. Natural revegetation of abandoned plowed land in the mixed prairie association of northeastern Colorado. Ecology, 25: 312--326.

162 Cotton, J. S., 1904a. A report on the range conditions of central Washington. Wash. State Agric. Coll. Ext. Stn. Bull. 60, 45 pp. Cotton, J. S., 1904b. Forage conditions of eastern and central Washington. MSc Thesis, Wash. State Univ., Pullman, 74 pp. Craig, M., 1892. Some Oregon weeds and how to destroy them. Oregon Agric. Exp. Stn. Bull., 19, 21 pp. Cusick, W. C., 1908. A catalogue of grasses in northeastern Oregon. In: Manuscripts, Archives, Special Collections. Wash. State Univ. Libraries, Pullman, 49 pp. Cusick, W. C., 1909. An annotated list of the flowering plants and ferns of northeastern Oregon: Wallowa, Union, Baker and Umatilla Counties. In: Manuscripts, Archives, Special Collections. Wash. State Univ., Pullman. Darwin, C., 1856. On the Origin of Species. (A facsimile of the first edition, Harvard University Press, Cambridge, 1964), pp. 60--67. Daubenmire, R., 1942. An ecological study of the vegetation of southeastern Washington and adjacent Idaho. Ecol. Monogr., 12: 53--79. Daubenmire, R., 1960. An experimental study of variation in the Agropyron spicatum-A. inerrne complex. Bot. Gaz. Chicago, 122: 104--108. Daubenmire, R., 1969. Ecologic plant geography of the Pacific Northwest. Madrono, 20: 111--128. Daubenmire, R., 1970. Steppe vegetation of Washington. Wash. Agric. Exp. Stn. Tech. Bull., 62, 131 pp. Daubenmire, R., 1978. Plant Geography. Academic Press, New York, NY, 338 pp. Dix, R. L., 1959. The influence of grazing on the thin-soil prairies of Wisconsin. Ecology, 40: 36--49. Drum, B. B., 1916. Some common weeds of the Palouse country, their seeds and seedlings. Bacc. Honors Thesis, Wash. State Univ. In: Manuscripts, Archives, Special Collections, Wash. State Univ. Libraries, Pullman, 104 pp. Elmer, A. D. E., 1899. Flora of the vicinity of Pullman of all Pteridophyta, Gymnospermae and Monocotyledons. Bacc. Honors Thesis, Wash. State Univ. In: Manuscripts, Archives, Special Collections. Wash. State Univ., Pullman, 70 pp. Elton, C. S., 1958. The Ecology of Invasions by Animals and Plants. Methuen, London, 181 pp. Fleming, C. E., Shipley, A. and Miler, M. R., 1942. Bronco grass (Bromus tectorum) on Nevada ranges. Nevada Agric. Exp. Stn. Bull., 159, 21 pp. Forsling, C. L. and Dayton, W. A., 1931. Artificial reseeding on Western mountain rangelands. U.S. Dep. Agric. Circ. 178, 48 pp. General Land Office, 1881. Land Survey Field Notes. Microfilm, Reel 69. U.S. Dep. Inter. Bur. Land Manage., Spokane, Washington. Grant, V., 1977. Organismic Evolution. Freeman, San Francisco, 418 pp. Greene, F., 1888. Report on the interior wheat lands of Oregon and Washington territory. Fiftieth Congress, First session Sen. Exec. Doc. 229. Griffiths, D., 1902. Forage conditions on the northern border of the Great Basin. U.S. Dep. Agric. Plant Ind. Bur. Bull., 15, 60 pp. Griffiths, D., 1903. Forage conditions and problems in eastern Washington, eastern Oregon, northwestern California, and northwestern Nevada. U. S. Dep. Agric. Plant Ind. Bur. Bull., 38, 52 pp. Gustafson, C. E., 1972. Faunal remains from the Marines rockshelter and related archaeological sites in the Columbia Basin. PhD Thesis, Wash. State Univ., Pullman, 183 pp. Harlan, J. R., 1945. Cleistogamy and chasmogamy in Bromus carinatus. Hook. and Arn. Am. J. Bot., 32: 66--72. Harper, J. L., 1965. Establishment, aggression, and cohabitation in weedy species. In: H.G. Baker and G. L. Stebbins (Editors), The Genetics of Colonizing Species. Academic Press, New York, NY, pp. 245--265. Harper, J. L., 1977. Population Biology of Plants. Academic Press, New York, NY, 892 pp.

163 Hedrick, D. W., 1965. History of cheatgrass -- present geographical range and importance of cheatgrass in management of rangelands. In: Cheatgrass Symposium. U.S. Dep. Inter., Bur. Land Manage., Portland, Oregon, pp. 13--16. Henderson, L. F., 1898. Twelve of Idaho's worst weeds. Univ. Idaho Agric. Exp. Stn. Bull., 14, pp. 89--136. Henderson, L. F., 1903. Grasses and forage plants in Idaho. Univ. Idaho Agric. Exp. Stn. Bull., 38, pp. 191--256. Henry, J. K., 1915. Flora of Southern British Columbia and Vancouver Island. Gage, Toronto, 363 pp. Hitchcock, C. L., Cronquist, A., Ownbey, M. and Thompson, J.W., 1969. Vascular Plants of the Pacific Northwest. Part. I. Vascular Cryptogams, Gymnosperms, and Monocotyledons. University of Washington Press, Seattle, WA, 914 pp. Hollister, O. J., 1882. The Resources and Attractions of Utah. A. Zeehandelaar, Salt Lake City, 93 pp. Hughes, J. A., 1915. Cheats and brome grasses. The Oroville Weekly Gazette, Oct. 22, 11 (18): 4. Hulbert, L. C., 1955. Ecological studies of Bromus tectorum and other annual bromegrasses. Ecol. Monogr., 25 : 181--213. Humphrey, R. R., 1945. Common range forage types of the inland Pacific Northwest. Northwest Sci., 19: 3--11. Kannenberg, L. W. and Allard, R. W., 1967. Population studies in predominately selfpollinated species. VIII. Genetic variability in the Festuca microstachys complex. Evolution, 21: 227--240. Kearney, T. H., Briggs, L. J., Shantz, H. L., McLane, J. W. and Piemeisel, R. L., 1914. Indicator significance of vegetation in Tooele Valley, Utah. J. Agric. Res., 1: 365--417. Klemmedson, J. O. and Smith, J. G., 1964. Cheat-grass (Bromus tectorum L.) Bot. Rev., 30: 226--262. I~mson-Scribner, F., 1899. Grass and forage plant investigations on the Pacific Coast. U.S. Dep. Agric. Div. Agrostology Circ. 22, 7 pp. Leiberg, J. B., 1897. General report on botanical survey of the Coeur d'Alene Mountains in Idaho during the summer of 1895. Contrib. U.S. Natl. Herb., 5: 1--85. Leopold, A., 1941. Cheat takes over. The Land, 1: 310--313. McInteer, B. B., 1952. Original vegetation in the Blue Grass Region of Kentucky. Castanea, 17: 153--157. Macoun, J., 1877. Report of Professor Mascoun. In: Report of Progress for 1875--1876. Geol. Surv. Can., pp. 110--232. Macoun, J., 1890. Catalogue of Canadian Plants. Part. V. Acrogens. W. F. Brown, Montreal, p. 397. Maser, C., 1974. The sage vole, Lagurus curtatus (cope, 1868) in the Crooked River National Grassland, Jefferson County, Oregon. A contribution to its life history and ecology. Saugetierk. Mitt., 22: 193--222. Meinig, D. W., 1968. The Great Columbia Plain. University of Washington Press, Seattle, WA, 576 pp. Mentzer, L. W., 1950. Studies on plant succession in true prairie. Ecol. Monogr., 21: 255-267. Mirov, N. T., 1951. Geography of Russia. John Wiley, New York, NY, 362 pp. Nelson, J. C., unpublished. Notes on the flora of Chelan County, Washington. Wash. State Univ. Libraries, Pullman 14 pp. Osborne, O. M., 1911. Weed pests of Idaho and methods of eradication. Idaho State Agric. Exp. Stn. Bull., 71, 36 pp. PammeI. L. H., 1897. Notes on the grasses and forage plants of Iowa, Nebraska and Colorado. U. S. Dep. Agric. Div. Agrostology Bull., 9, 47 pp. Pammel, L. H., 1905. Notes on the flora, especially the forest flora of the Bitter Root Mountains. Iowa Acad. Sci. Proc., 13: 87--100.

164 Pammel, L. H., 1910. The problem of weeds in the west. Iowa Acad. Sci. Proc., 17: 34--46. Pammel, L. H., 1915. Weeds of the Farm and Garden. Orange Judd, New York, NY. Peck, M. E., 1916. A section of Upper Sonoran Flora in Northern Oregon. Iowa Acad. Sci. Proc., 23: 317--324. Pickford, G. D., 1932. The influence of continued heavy grazing and promiscuous burning on spring-fall ranges in Utah. Ecology, 13: 159--171. Piemeisel, R. L., 1932. Weedy abandoned lands and the weed hosts of the beet leaf hopper. U. S. Dep. Agric. Circ., 229. Piper, C. V., 1920. Forage Plants and Their Culture. MacMillan, New York, NY, 618 pp. Piper, C. V. and Beattie, R. K., 1901. The Flora of the Palouse Region. Allen, Pullman, Washington, 208 pp. Piper, C. V. and Beattie, R. K., 1914. The Flora of Southeastern Washington and Adjacent Idaho. New Era Printing, Lancaster, Pennsylvania, 296 pp. Reid, E. H., 1942. Important plants on national forest ranges of eastern Oregon and eastern Washington. U. S. Dep. Agric. Pacific Northwest Forest Range Exp. Stn. Range Res. Pap. No. 1, 64 pp. The Resources of Southern Oregon, 1890. The Southern Oregon State Board of Agriculture. Salem, Oregon, 152 pp. The Resources of the State of Oregon, 1892 (rev.). The State Board of Agriculture. Salem, Oregon, 230 pp. Salisbury, E., 1961. Weeds and Aliens. Collins, London, 384 pp. Sampson, A. W., 1917. Important range plants: their life history and forage value. U. S. Dep. Agric. Bull., 545. Sandberg, J.H. and Leiberg, J.B. (n.d.). Report upon a botanical survey of portions of eastern Washington. In: Records of The Bureau of Plant Industry, Soils, and Agricultural Engineering. Div. Bot., Rec. Group 54, U.S. Natl. Arch. Wash., DC, 122 pp. Shantz, H. L., 1916. Field Note Book, No. 82. Washington, August--September, 1916. In: Special Collections. Univ. Arizona Library, Tucson, AZ, pp. 200--231. Shantz, H. L. (in I. Tidestrom), 1925. Flora of Utah and Nevada. Contrib. U. S. Natl. Mus. 25, 665 pp. Shear, C. L., 1901. Field work of the Division of Agrostology: a review and summary of the work done since the orgainization of the division, July 1, 1895. U. S. Dep. Agric. Div. Agrostology Bull., 25, 67 pp. Stebbins, G. L., 1957. Self-fertilization and population variability in the higher plants. Am. Nat., 91: 337--354. Stebbins, G. L., 1974. Flowering Plants. Evolution Above the Species Level. Belknap Press, Cambridge, 399 pp. Suksdorf, W. N., 1892. Flora Washingtonensis. A Catalogue o f the Phaenogamia and Pteridophyta of the State of Washington. In: Manuscripts, Archives, Special Collections. Wash. State Univ. Libraries, Pullman, 15 pp. Tisdale, E. W., 1947. The grasslands of the southern interior of British Columbia. Ecology, 28" 346--382. Tracy, S. M., 1888. Report of an investigation of the grasses of the arid districts of Texas, New Mexico, Arizona, Nevada, and Utah in 1887. U. S. Dep. Agric. Bot. Div. Bull., 6, 61 pp. U. S. Geologic Survey, 1898. Nineteenth Annual Report 1897--98. Part V. Forest Resources. Atlas, plate II. Van Der Plank, J., 1963. Plant Diseases: Epidemics and Control. Academic Press, New York, 349 pp. Washington County Agents' letters to I. D. Cardiff, 1916. In: Manuscripts, Archives, Special Collections. Wash. State Univ. Libraries, Pullman. Wash. Dep. of Agric., 1914. First Annual Report 1913--1914. Seed Lab. Rep., pp. 101--111. Wash. Dep. of Agric., 1916. Second Annual Report 1914--1916. Seed Lab. Rep., pp. 164-178.

165 Wash. State Univ. Coop. Ext. Serv. County Agent Rep., 1914. Annual report of the Benton County Agent. In: Manuscripts, Archives, Special Collections. Wash. State Univ. Libraries, Pullman. Weaver, J. E., 1917. A study of the vegetation of southeastern Washington and adjacent Idaho. Univ. Nebraska Stud., 17: 1--133. Weaver, J. E., 1958. Native grassland of southwestern Iowa. Ecology, 39: 733--750. Wood, A., 1863. Class-Book of Botany. Barnes and Burr, New York, p. 791.