Morphological and Ecological Segregation of Two Sympatric Lomatium triternatum (Apiaceae) Varieties in Montana

2013·4.9 MB·English

by  Peter Lesica

Madrono, Vol. 60, No. 3, pp. 211-216, 2013
MORPHOLOGICAL AND ECOLOGICAL SEGREGATION OF TWO SYMPATRIC
LOMATIUM TRITERNATUM (APIACEAE) VARIETIES IN MONTANA
Peter Lesica
MT
Herbarium, Division of Biological Sciences, University of Montana, Missoula, 59802
[email protected]
Pamela M. Kittelson
MN
Department of Biology, Gustavus Adolphus College, St. Peter, 56082
Abstract
The Lomatium triternatum complex is widespread in the Columbia River Basin. Lomatium
triternatum (Pursh) J. M. Coult. & Rose var. triternatum and L. triternatum (Pursh) J. M. Coult. &
Rosevar.anomalum(M. E.JonesexJ. M. Coult. &Rose)Mathiasaresympatricthroughoutmuchof
their range and are reported to have similar fruit but different leaves. Plants of the two varieties
repeatedly occur within 5-30 m ofeach other at an unusual site in northwest Montana, USA. We
collected data on leafand fruit morphological characters and habitat associations at this site to help
address whether the proper taxonomic rank for these two taxa should be revised. Terminal leaflet
shapeand specificleafareadifferedbetween thetwovarietieswithlittle orno overlap. Fruitmericarp
length and width also differed between the two varieties. These patterns are not consistent with
treatingL. triternatumvar. triternatum and L. triternatumvar. anomalum as sympatric, interbreeding,
conspecific taxa. The two varieties ofL. triternatum also occurred in different plant communities in
spiteofgrowingincloseproximity: var. anomalum occurswithtall, leafy forbs, whilevar. triternatum
is associated with xeric-adapted bunchgrasses and cushion-forming forbs. We conclude that the two
varieties are better recognized as separate species.
Key Words: Apiaceae, Columbia Basin, Lomatium triternatum Montana.
,
Lomatium is the largest North American genus Rose) Mathias. Here we explore the taxonomic
in the Apiaceae with its main center of diversity disposition of these two varieties. They are
in the Columbia Plateau and IntermountainWest reported to have similar fruit but distinctly
(Mathias 1938; Sun and Downie 2010). Analysis different leaves. Variety triternatum has bipinnate
of molecular data suggests that Lomatium and or biternate leaves with linear leaflets, while var.
other closely allied genera probably underwent anomalum has bipinnate to tripinnate leaves with
rapid diversification during the late Tertiary (Sun narrowly oblong leaflets. These two varieties are
et al. 2004). As a result, the genus is taxonom- reported to be “wholly intergradient” (Cronquist
ically difficult (Mathias 1938), with relatively few et al. 1997), but no further information is
taxonomically informative morphological traits, provided. Variety triternatum occurs from south-
numerous narrow endemic species (Darrach et al. ern Alberta and British Columbia to northern
2010), and a good deal of morphological California and northern Utah, while var. anom-
convergence (Sun and Downie 2010). alum is found sporadically from northwest
The Lomatium triternatum (Pursh) J. M.Coult. Montana to eastern Washington south to south-
& Rose complex is widespread in the Columbia west Oregon, across much of Idaho to northern
River Basin. The complex has been variously Utah and southwest Wyoming (Cronquist et al.
treated as seven species (Coulter and Rose 1900), 1997; Consortium ofNorthwest Herbaria 2011).
two species consisting ofseven varieties (Mathias Possession of similar morphological traits can
1938), and onespeciesconsistingofthreevarieties often indicate a close phylogenetic relationship
(Hitchcocketal. 1961; Cronquistet al. 1997). The between species. Indeed, similar morphology,
complex is loosely united by the following especially in the anatomy of reproductive struc-
characters: narrow involucel bracts, similar mer- tures, has long formed the basis of classical
icarp shape, and vaguely similar leaf shape. systematics (Stace 1980). However, similar mor-
Hitchcock and Cronquist (1973), in the most phology may actually be the result ofconvergent
recent treatment for the Pacific Northwest, use a evolution rather than an indication of close
tripartite classification with two subspecies: L. t. phylogenetic relationship (Niklas 1997; Judd et
subsp. triternatum and L. t. subsp. platycarpum al. 2008). Indeed, recent molecular studies show
(Torr.) B. Boivin (= L. simplex [Nutt, ex S. that convergent evolution has occurred in the
Watson] J. F. Macbr.). Subspecies triternatum is Apiaceae, and fruit morphology in particular is
composed of two varieties: var. triternatum and reported to be a poor indicator of phylogenetic
var. anomalum (M. E. Jones ex J. M. Coult. & relatedness in the family (Downie et al. 2002).
MADRONO
212 [Vol. 60
Under the biological species concept, two
species will rarely if ever interbreed to produce
fertile offspring, but interbreeding is possible and
likelyinazone ofoverlapbetweentwo subspecies
orvarieties ofthe same species (Mayr 1942; Stace
1980). Both L. var. triternatum and L. t. var.
anomalum occur together in an unusual geologic
setting in northwest Montana known as the
Giant Ripple Marks. Plants of the two varieties
m
repeatedly occur within 5-30 of each other.
The purpose of our study is to determine the
degree to which these two taxa may intergrade
ecologically and/or morphologically at this site in
order to inform an independent assessment of
their proper taxonomic rank. Fig. 1. Giant Ripple Marks geologicsiteinnorthwest
Montana. Ripplemarksareinthebottomofthevalley;
Study Site the ridges support Lomatium triternatum var. triterna-
tum and have a light color from the leaves and flowers
We conducted our study in the Giant Ripple ofEriogonum ovalifolium. The dark areas between the
Marks at the north end ofCamas Prairie Basin at ridges indicate more mesic vegetation and support L.
an elevation ofca. 885 m, 8 km south ofthe town triternatum var. anomalum.
ofHot Springs on the Flathead Indian Reserva-
tion in Sanders Co., Montana (47°30.9'N, and the forbs Lomatium triternatum var. anom-
114°35.0'W). The climate of this region is semi- alum, Artemisia ludoviciana Nutt., and Lupinus
arid with cold winters and hot summers. The sericeus Pursh.
most similar weather recording station to the
study site is at Lonepine, 15 km to the north. The Methods
mean January and July temperatures are —5.2°C
and 20.1°C, respectively. The average annual We located a 50-m transect across six haphaz-
precipitation is 286 mm (NCDC 2012). The ardly chosen ripple marks in early June 2009 so
majority of the precipitation occurs from April that approximately half of each transect was on
through June. the slope of the ridge, and the other halfwas in
The Giant Ripple Marks were formed when the swale at the bottom. We sampled the closest
Glacial Lake Missoula drained during Pleisto- Lomatium triternatum plant to the transect line at
cene glacial retreats, andwaterpoured south over 5-m intervals, providing 10 samples per transect
passes at the north end of the Camas Prairie and a total of60 target plants. We classified each
Valley. The water dumped its load of coarse plant into either var. triternatum (narrow leaflet)
sediments in two lines of enormous current or var. anomalum (broad leaflet) and measured
ripples (Alt and Hyndman 1986). These ripple the length and maximum width of the terminal
marksappeartodayastwo seriesoflow(5-20m), leaflet ofthe lowest leaffor each plant in the field
parallel ridges roughly perpendicular to the main and then immediately placed the leaflet in a plant
axis of the valley (Fig. 1). The vegetation of the press. Senescent ordamaged leaves were avoided.
Giant Ripple Marks is a mosaic of bunchgrass We dried collected leaves for 24 hr and weighed
prairie on the well-drained ridges with graminoid each one to the nearest mg.
and forb-dominated meadow vegetation and Vegetation associated with each target plant
occasional vernal pools on the deeper, fine was quantified by placing a 1-m2 plot frame with
textured soils in the swales between the ridges. the target plant in the center and estimating
The tops and upper and middle slopes of the canopy cover ofeach vascular plant species into
ripple mark ridges are dominated by the grasses one ofthe following classes: 0.1%, 1%, 5%, 10%,
Pseudoroegneria spicata (Pursh) A. Love, Poa 15%, etc. Native plot vegetation was ordinated
secunda J. Presl, and Elymus elymoides (Raf.) usingprincipal components analysis (PCA). Only
Swezeyand thecushion-formingforbsEriogonum species occurring in at least five plots were
ovalifolium Nutt, and Antennaria dimorpha included in the analysis. Non-native species, such
(Nutt.) Torr. & A. Gray. Bromus tectorum L. is as Agropyron cristatum (L.) Gaertn. Bromus
an annual, exotic grass that is common particu- tectorum, and Potentilla recta, were common in
larly where the perennial grasses have declined some plots but were excluded from the analysis.
due to grazing. Vegetation of the swales is Nomenclature follows Lesica (2012).
currently dominated by introduced species such We collected mature fruits (mericarps) from
as Poa pratensis L. and Potentilla recta L. The haphazardly chosen plants ofboth L. triternatum
most common native species in the swales include taxa on June 18, June 24, and July 10, 2008.
the grasses Elymus elymoides and Poa secunda Length and width ofrandomly chosen mericarps
2013] LESICA AND KITTELSON: LOMATIUM TRITERNATUMVARIETY SEGREGATION 213
Fig. 2. Width (a), length/width ratio (shape) (b), and specific leaf area (c) for terminal leaflets of Lomatium
triternatumvar. anomalum and var. triternatum. The box represents themiddle 50% ofthe scoreswiththe vertical
line the median; the upper and lower whiskers mark the range ofvalues excluding outliers; the dots represent the
5th and 95th percentiles.
=
and their enclosed seed were measured to the did not differ between the two varieties (P
mm
nearest 0.1 with a dissecting microscope. 0.11). However, leaflet width was significantly
Waimtohnign-mdoartpehsot(yPpe>mea0.s1u5r)e,mesnotswdeidcnoomtbdiinffeedr gtrreiatteerrnaftourmL(.Pt<. v0a.r0.01a;nFoimga.l2uam),caonmdplaernegdtht/owivdatr.h
measurements from different dates in our analy- ratio for the terminal leaflet of var. triternatum
ses to provide a sample ofn = 40 for each taxon. was greater than 10 but less than 10 for var.
Leafarea (cm2) was measured on the dried leaf anomalum (P < 0.001; Fig. 2b). Specific leafarea
collections with a LI-COR 3100 area meter (LI- (SLA) also differed between varieties. Mean SLA
COR, Lincoln, NE). Surface area of each leaflet was 148 cm2/g (SE = 9) and 76 cm2/g (±4) for
was measured three times and averaged. Specific var. triternatum and var. anomalum, respectively
leaf area (SLA) for each target plant was (t = 7.0, P < 0.001), and there was little overlap
calculated as the ratio ofthe average ofthe three between the two varieties (Fig. 2c).
surface area measurements to leaf dry mass Some fruit characters also differed between the
(cm2/g). We used two-sample t-tests to assess two varieties. Mean mericarp length was 7.0 mm
the difference in leaflet shape, specific leaf area, (SE = 0.1) and 8.6 mm (±0.2) for var.
and fruit size and shape. triternatum and var. anomalum respectively (t =
7.9, P < 0.001), and width wa,s 4.2 mm (±0.1)
Results and 4.8 mm (±0.1), respectively (t = 5.1, P <
0.001). However, fruit shape (length-width ratio)
=
Shape ofthe terminal leaflets differed between did not differ between the two varieties (P
the two varietieswith little overlap. Leafletlength 0.30). The ratio of the width of one side of the
MADRONO
214 [Vol. 60
and cushion-forming forbs. Different communi-
ties often host different congeners (Schluter
2000
).
By definition infraspecific taxa should be
interfertile to some extent where they co-occur
(Stace 1980), so wewould expect to find plants of
intermediate morphology when two varieties of
the same species are closely sympatric, as in our
study area. However, these two putative varieties
ofLomatium triternatum remain morphologically
distinct in spite of occurring in close proximity.
Mericarps ofvar. anomalum are larger than those
ofvar. triternatum with little overlap, and leaflets
of var. anomalum are wider than those of var.
triternatum with no overlap. The two taxa differ
in both fruit and leaf traits, which are unrelated
characters likely coded for by different genetic
pathways. Forexample, phenotypicdifferencesin
fruit and leaf traits have been shown to be
attributed to different molecular traits or devel-
opmental pathways (Lippman and Tanksley
anomalum triternatum 2001; Thul et al. 2009). Since traits among these
Fig. 3. First principal component analysis (PCA) two Lomatium taxa appear to be nearly discon-
scores for vegetation associated with Lomatium triter- tinuous it suggests that they are genetically
natum var. anomalum and var. triternatum. The box isolated in spite ofoccurring sympatrically.
representsthemiddle50% ofthescoreswiththevertical Specific leaf area (SLA) for var. triternatum
line themedian; theupperandlowerwhiskersmarkthe was nearly twice as great as for var. anomalum.
range of values excluding. The dots represent the 5th
and 95th percentiles. Higher PCA scores were associ- Both infraspecific and interspecific differences in
ated with more mesic vegetation. specific leafarea (SLA) have been linked to water
stress (Fitter and Hay 2002; Hoffmann et al.
2005). Plants occurring in drought-prone areas
mericap wing to the width of the enclosed seed tend to have lower SLA (thicker leaves) than
was 0.56 (±0.02) and 0.49 (±0.03) for vars.
2t.r2iT,thePern=attuw0mo.0a28vn)ad.riveatri.esanoofmaLloumma,trieusmpecttriivetleyrn(attu=m tH1h9oo9fs7f;emCianunnmnnoirneetghmaael.smiec2t0ea0ln5.v)1i.9r9o9Fn;umreWtnrhtiesgrhm(toReretiec,alh. p2etl0a0a1nl;t.
aoxcicsurorfedthienPdiCffAereanctcopulannttedcofmomrun1i7t%iesof. tThheePtfoCitraAslt ssSphLeacAdie—esel.eoga.f,vteessnun(eCxlhperaaezvsedssonphhaaevnneodtayKpliaocuwefrpmlaaSsntLniAcit1yt9h9a3in;n
variation in the associated vegetation data. Lichtenthaler et al. 2007). Thus, ifthe two forms
1 represented a gradient between xeric vegetation of Lomatium were ecotypes of the same species,
dominated by the grass Pseudoroegneria spicata wewould expect the swalevar. anomalum to have
and the cushion-forming forbs Antennaria dimor- a higher SLA than var. triternatum, which occurs
pha and Eriogonum ovalifolium, and amoremesic in more xeric habitats. Instead, we found the
assemblage dominated by taller, leafier forbs opposite. Given the difference in SLA, it is
Jii.cnecuMlsu.,dLinCogomualAttc.ihuil&mlemaRaocsmrieol,cleaafronplduiumLm.(TLot.r,rri.tLeu&rpniaAnt.uumsGrvsaaeyrr).- utohnfaltipklmaesoltryipctihthoyal.togItinhceatslheedficofarfsmeerseonafcreetshaiernseteesrtibwmropeleytdaiaxnagr,esatunhldet
anomalum. Twenty-seven plants of var. triterna- differences in SLA are more likely attributed to
tum (84%) had a PCA-1 score <0, while 22 of entirely different gene pools (White and Montes-
var. anomalum plants (79%) had a score >0 R 2005), with SLAbeinga geneticallydetermined
(Fig. 3). Other axes were not interpretable. morphological difference related to other traits
such as water use (Nobel 1980), leaftemperature
Discussion (Van Volkenberg and Davis 1977), or internal
anatomical differences (Nobel 1991). SLA may
The two Lomatium taxa discussed here were not evolve independently ofother morphological
associated with different habitats at our study traits; var. anomalum may belong to a lineage
site. Plants assigned to var. anomalum were found with relatively low SLA that has been conserved
only in swales dominated by tall, leafy forbs, along with other adaptations, such as large
whilevar. triternatum wascommon in morexeric, leaves, which correspond to its relatively mesic
ridgetop and slope communities dominated by habitat. Our combined morphological results
relatively sparse Pseudoroegneria spicata tussocks suggest that the two taxa are not exchanging
2013] LESICA AND KITTELSON: LOMATIUM TRITERNATUMVARIETY SEGREGATION 215
genes for SLA or other leaf and fruit traits and Hitchcock, C. L. and A. Cronquist. 1973. Flora of
may not even belong to the same lineage within the Pacific Northwest: an illustrated manual.
Lomatium. Indeed, recent molecular-based phy- University ofWashington Press, Seattle, WA.
logenetic research indicates that the two taxa , , M. OWENBY, AND J. W. THOMPSON.
belong to separate clades (D. Mansfield, College 1961. Vascular plants of the Pacific Northwest,
ofIdaho, personal communication). WPaAr.t 3. University of Washington Press, Seattle,
Lomatium triternatum var. anomalum and L. t. Hoffmann, W. A., A. C. Franco, M. Z. Moreira,
var. triternatum occupy different habitats in and M. Haridasan. 2005. Specific leaf area
northwest Montana and have different leaf and explains differences in leaf traits between conge-
fruit morphologies with little or no overlap even neric savanna and forest trees. Functional Ecology
when populations occur intermixed. In addition, 19:932-940.
differences in leafanatomy between the two taxa Judd, W. S., C. S. Campbell, E. A. Kellogg, P. F.
are likely genetically determined. Taken together, Stevens, and M. J. Donoghue. 2008. Plant
our results from this site of sympatry support systematics: a phylogenetic approach, 3rd ed.
treatment of Lomatium triternatum var. anom- Sinauer Associates, Sunderland, MA.
alum as a separate species: Lomatium anomalum Lesica, P. 2012. Manual of Montana vascular plants.
M. E. Jones ex J. M. Coult. & Rose. WBootratnhi,caTlX.Research Institute of Texas Press, Fort
Acknowledgments Lichtenthaler, H. M., A. Ac, M. V. Marek, J.
Kalina, and O. Urban. 2007. Differences in
WearegratefultoJohn Hoagforprovidingaccessto pigment composition, photosynthetic rates and
hisproperty. Laura Lynch and Donald Davis provided chlorophyll fluorescence images of sun and shade
assistance measuring and recording traits. P. K. was leaves of four tree species. Plant Physiology and
supported by a Research, Scholarship and Creativity Biochemistry 45:577-588.
Grant from Gustavus Adolphus College. Lippman, Z. Z. and S. D. Tanksley. 2001. Dissecting
the geneticpathwayto extreme fruit sizein tomato
Literature Cited usingacrossbetween the small-fruited wild species
Alt, D. and D. W. Hyndman. 1986. Roadside Lycopersicon pimpinellifolium and L. esculentum
var. Giant Heirloom. Genetics 158:413^122.
gMeTo.logy of Montana. Mountain Press, Missoula, Mathias, M. E. 1938. A revision of the genus
Lomatium. Annals of the Missouri Botanical
Chazdon, R. L. and S. Kaufmann. 1993. Plasticity Garden 25:225-297.
of leaf anatomy of two rain forest shrubs in Mayr, E. 1942. Systematics and the origin ofspecies.
relation to photosynthetic light acclimation. Func- Columbia University Press, New York, NY.
ConstioornatliuEcmoolfogNyo7r:t38h5w-e39s4t. Herbaria. 2011. Burke National Climatic Data Center (NCDC). 2012.
Museum ofNatural History and Culture. Univer- Website: http://www.ncdc.noaa.gov [accessed Sep-
sity of Washington, Seattle. Website http://www. tember 2012].
pnwherbaria.org/ [accessed 02 February 2013]. NlKLAS, K. J. 1997. Theevolutionarybiologyofplants.
Coulttheer,UmJb.elMl.ifaernade.J.CNo.ntRroisbeu.ti1o9n0s0.fMroonmogtrheapUh.So.f NobeUln,ivePr.sitSy. o1f9C80h.icaLgeoafPreasnsa,tCohmiycagaon,dILw.ater use
National Herbarium 7:9-256. efficiency. Pp. 43-55 in N. C. Turner and P. J.
CronHqoulimsgtr,enA.., 19N9.7.HI.nteHromlomungtraeinn,floaran:dvaPs.culKa.r hKirgahmteermp(eedrsa.t).urAedastprteasst.ioWinleoyf,plNaentwsYtoorkwa,tNerY.and
PplaarnttsA,ofSutbhcelaIsnsteRromsoiudnateai(nexWceesptt,FUabSaAle,s)V.olN.e3w, pla.nt1p9h9y1s.ioPlhoygysiAccoacdheemmiiccaPlresasn.dNeewnvYiorrokn,meNntYa.l
York Botanical Garden Press, Bronx, NY. Reich, P., M. B. Walters, and D. S. Ellsworth.
Cunningham, S. A., B. Summerhayes, and M. 1997. From tropics to tundra: global convergence
Wstersutctoubrye.an1d999c.heEmvioslturty,ioncaormypadriivnerggernaciensfaliln alenadf AincapldaenmtyfuonfctSicoineincnegs.,PUroScAee9d4i:n1g3s73o0f-t1h3e73N4a.tional
soil nutrient gradients. Ecological Monographs Schluter, D. 2000. Theecologyofadaptiveradiation.
69:569-588. Oxford University Press, Oxford, U.K.
Darrach, M., K. K. Thie, B. L. Wilson, R. E. STACE, C. A. 1980. Planttaxonomyandbiosystematics.
Brainerd, and N. Otting. 2010. Lomatium Edward Arnold, London, U.K.
tamanitchii (Apiaceae) a new species from Oregon Sun, F. and S. R. Downie. 2010. Phylogenetic
and Washington State, USA. Madrono 57:203- analyses of morphological and molecular data
208. reveal major clades within the perennial, endemic
Downie, S. R., R. L. Hartman, F. Sun, and D. S. western North American Apiaceae subfamily
Katz-Downie. 2002. Polyphyly of the spring- Apioideae. Journal oftheTorrey Botanical Society
ipcaarlsleevyisde(nCcyemsoupgtgeersutss)c:ommoplleecxulraerlaatinodnsmhoirpsphaomloongg- 137,:133-156,.and R. L. Hartman. 2004. An ITS-
the perennial endemic genera of western North based phylogenetic analysis of the perennial,
American Apiaceae. Canadian Journal of Botany endemicApiaceae SubfamilyApioideaeofWestern
80:1295-1324. North America. Systematic Botany 29:419-431.
Fitter, A. H. and R. K. M. Hay. 2002. Environmen- Thul, S. T., R. K. Lal, A. K. Shasany, M. P.
tal physiology of plants, 3rd ed. Academic Press, Darokar, andA. K. Gupta. 2009. Estimationof
London, U.K. phenotypic divergence in a collection of Capsicum
MADRONO
216 [Vol. 60
species for yield-related traits. Euphytica 168:189- Wright, I. J. and M. Westoby. 2003. Nutrient
196. concentration, resorption and lifespan: leaf traits
VanVolkenburgh, E. andW. J. Davies. 1977. Leaf ofAustralian sclerophyll species. Functional Ecol-
anatomy and water relations of plants grown in ogy 17:10-19.
controlled environments and in the field. Crop , P. B. Reich, andM. Westoby. 2001. Strategy
Science 17:353-358. shifts in leaf physiology, structure and nutrient
White,J.andC.Montes-R.2005.Variationinparameters content between species of high- and low-rainfall
related to leafthickness in common bean (Phaseolus and high- and low-nutrient habitats. Functional
vulgarisL.). FieldCropsResearch91:7-21. Ecology 15:423^134.