Sphinx moth pollinators for the endangered western prairie fringed orchid, Platanthera praeclara in Manitoba, Canada

2004·3.8 MB·

by  A Richard Westwood

JournaloftheLepidopterists'Society
58(1),2004,13-20
SPHINXMOTH POLLINATORS FOR THE ENDANGEREDWESTERN PRAIRIE FRINGED ORCHID,
PLATANTHERA FRAECLARA IN MANITORA, CANADA
A. RichardWestwood1 and Christie L. Borkowsky
DepartmentofBiology,UniversityofWinnipeg,Winnipeg, ManitobaR3B2E9,Canada
ABSTRACT. Thewesternprairiefringedorchid,Platantherapraeclara (Sheviak&Bowles),isanendangeredspeciesinNorthAmerica.
In Manitobaorchidsproducelowernumbers ofseedcapsulesthan more southern populations. Exploration ofthepollination biologyforP.
praeclaraiscriticaltopreservediisendangeredspecies.ThisstudyidentifieddiepollinatorsofP. praeclara usingconeandmalaisetrapsand
testedtheeffectivenessofmarkingpollinatorswithtracesofDay-Glow®orangemarkerpowder.AlthoughLepidopterawerenumerousinor-
chidplotsduringdailyobservationperiods,includingdayflyingSphingidae,nonewerepollinatorsforP.praeclara.Amongthe5856insectsfrom
49familiescapturedover45trappingdays,sixsphinxmoths,twospecimensofHylesgallii(Rottenburg)andfourspecimensofSphinxdrupifer-
arum J.E. Smith (Sphingidae), were foundwith two ormore orchid pollinia attached to theireyes and were confirmedas pollinators ofP.
praeclara. S.drupiferarumisuncommoninsoudiemManitobaandH. galliiappearstobealessefficientpollinatorthanS. drupiferarum.Pro-
boscislength,eyewidthandflightperiodmayinfluencetheefficiencyofbioticpollinationoftheorchid.
Additionalkeywords: Sphingidae,pollinia,pollination,P. praeclara,S. drupiferarum,H.gallii.
The western prairie fringed orchid, Platanthera seedcapsules can rangeup to49% (Bowles 1983, She-
praeclara (Sheviak & Bowles 1986), occurs in areas of viak& Bowles 1986, Cuthrell 1994). The audiors have
remnant tall grass prairie in southeastern Manitoba. hypothesizedthatreducedseedpodproduction maybe
Priortoitsdiscoveryinthemid 1980snearTolsoi, Man- linkedtolowpollinationsuccessin Manitoba. Thelevel
itoba, P. praeclara was not known to exist in Canada ofpollination success forP. praeclara and the identity
(Johnson 1985, 1991). The Manitoba population, atits ofthe orchidpollinators are unknownin Manitoba.
maximum has consisted ofapproximately 21,000 indi- Pollination is die process inwhich pollen grains are
vidual plants (although the population may fluctuate transferred to the stigma, which is followed by fertil-
widely from year to year) and it is the largest offour ization ofthe ovules and development ofseeds (Proc-
metapopulations (over 300 plants) in North America. tor et al. 1996). Many orchids require a biotic organ-
The remaining populations occur in North Dakota, ism (a pollination agent orpollen vector) to transport
South Dakota, Minnesota, Kansas, Nebraska, and die pollen to die stigma (van der Pijl 6k Dodson 1966,
Iowa (Sheviak & Bowles 1986, Bjugstad & Fortune Faegri & van der Pijl 1979). In P. praeclara die most
1989, Bray & Wilson 1992, Bjugstad-Porter 1993, striking visual characteristic of the flower is a large,
Pleasants &Moe 1993, Davis 1994, Sieg& King 1995, deeply fringed, tri-lobed lower lip and a long, slender
U.S. Fish &Wildlife Service 1996, Hofet al. 1999). spur containing nectar that suspends backward from
The orchid is protected under Manitoba's Endan- the flower (Figs. 1,2). The nectar spur maybe 36-55
gered Species Act and has been placed on the endan- mm long with a maximum diameter of2.7 ± 0.5 mm
gered species list in both Canada and United States (Sheviak & Bowles 1986). Orchids growto 38-85 cm
(Collicutt 1993, Davis 1995). In Manitoba,P. praeclara tall (Sheviak& Bowles 1986) with the determinantin-
oftenhas lowfruitsetandsubsequentseedproduction florescence containing 7 to 12 flowers (Sheviak &
(Borkowsky 1998). Although the western prairie Bowles 1986, Pleasants 1993). The pollinium (Fig. 1)
fringedorchidwillproduce avegetative form, there is is a specialized structure that consists ofpollen, a col-
littleevidencethatvegetativereproductionoccurs inP. umn andviscidium (Nilsson 1992). InP. praeclara die
praeclara (Bowles 1983, Sather 1991, Sieg & King minute grains of pollen are arranged into subunits
1995, Hofetal. 1999). Aplant mayalsogodormant, as called massulae (Pleasants & Moe 1993). These sub-
is typical in other species of orchids (Nilsson 1992). units form abi-lobed mass that is attached to the col-
Therefore,the recruitmentofnewplantsis dependant umn, which is secured to the viscidium (die entire
primarily upon successful pollination and subsequent structure is termed the pollinium). The pollinium is
seed production. In surveys of over 1000 plants sheathed,withtheexceptionoftheviscidium.which is
Borkowsky (1998) found that only 2.1% of orchid exposed and adaptedto cement itselfto the pollinator
stemsproducedone ormoreseedcapsules annuallyin (Bowles 1983). Each flowerhas onepollinium located
Manitoba between 1994 and 1998. In more southern on eitiier side ofdie stigmatic surface. This allows for
mm
orchid populations the percent ofstems that produce a6 to 7 separation between eachviscidium (She-
viak& Bowles 1986). Theopeningtodienectarspuris
' Corresponding author: R. Westwood, email: r.westwood® locatedimmediatelybelowthe stigmatic surface.
uwinnipeg.ca;Phone[phone];Fax:[phone]. The small opening to the nectar spur restricts the
14 Journalofthe Lepidopterists' Society
Fig. 1. Platanthera praeclara flower, anterior view, showing FlG.2. Platantherapraeclara flower,lateralview, showingnec-
pollinium(a),andtheopeningtonectarspur(b). tarspur(a).
position ofa potential nectar seeking insect, and in- Dodson 1966, Patt et al. 1989, Robertson & Wyatt
creases the likelihood that one or both viscidia will 1990, Bowles et al. 1992, Larson 1992).
come into contact with a likely pollinator. The The floral characteristics ofP. praeclara indicate a
pollinium is pulled from the sheathwhen apollinator moth pollination method (Faegri & van der Pijl 1979,
contacts the viscidium and then withdraws from the Sheviak & Bowles 19S6, Luyt & Johnson 2001), most
flower(Darwin 1904, Dressier 1981, 1993). likely sphingophily or phalaenophily. P. praeclara has
Orchids also produce a scent or fragrance which no developed landing platform on the flowers, there-
maybeweakorabsent in the daytime, increases inin- fore hovering pollinators are favored and butterflies
tensityatduskand mayremainstronguntilsunrisede- may be excluded (van der Pijl & Dodson 1966,
pendingon the age and condition ofthe flower (She- Dressier 1981). Toretrievenectar, theinsectmustcor-
viak & Bowles 1986). The relationship between the rectly align itselfwith the flower as it inserts its pro-
attractiveness ofthe fragrance to pollinators and the boscis,which increases the likelihoodthat the pollina-
intensityofthe scentis unknown inP. praeclara. torwill contact one or both ofthe viscidiawhich will
To be identified as a pollinator and to rule out in- adhere to the proboscis or eyes ofthe insect (Sheviak
discriminate visitors to the flower, the organism must & Bowles 1986). Consideringthe length ofthe nectar
make regular visits to the flowers during its lifetime spurofP. praeclara andpositionoftheviscidia,thelist
andeffectivelydepositthepollenonthestigma(Faegri of potential pollinators is further reduced to Lepi-
& van der Pijl 1979). The flower-pollinator relation- dopterawidiaverylongproboscis (e.g., mothsbelong-
ship may also be controlled by pollinator behavior, ing to the Sphingidae or hawkmoth family). Few, if
mouthpart morphology or taxonomy and by flower any, observations ofpollination offringed orchids by
morphology (Faegri &van der Pijl 1979,Wyatt 1983). hawkmothshavebeen madeinthe field(Bowles 1983,
Wyatt (1983) identifiednineformsofbioticpollination Sheviak & Bowles 1986, Pleasants & Moe 1993).
—
inorchids includingsphingophily(Sphingidae hawk- Based on proboscis length, Sheviak & Bowles (1986)
moths), phalaenophily (small moths), psycophily (but- suggestedthatthe followingspecies ofmoths couldbe
terflies), melittophily (bees), myophily (syrphid and potential pollinators of P. praeclara in the United
beeflies), sapromyophily(carrion anddungflies), can- States: Eumorphaachemon (Drury), Hyleslineata (F),
tharophily (beetles), ornithophily (birds) and chi- Sphinx drupiferarum J.E. Smith and Sphinx kalmiae
ropterophily(bats). J.E. Smith. Cuthrell (1994) collected two specimens,
The potential pollinators for P. praeclara in Mani- one each of Sphinx drupiferarum and Eumorpha
toba are not known. Although the literature suggests achemon with viscidia from P. praeclara attached to
that night-flying members ofthe Sphingidae may be thehead,fromalighttrapadjacenttoafieldoforchids
keypollinators ofP. praeclara, other members ofthis indie United States.
orchidgenus andseveralcloselyrelatedgeneramaybe To better understand the pollination biology of
pollinated by butterflies, other modi species, certain P. praeclara in Manitoba diurnally active Lepidoptera
Diptera, Coleoptera or Hymenoptera (van der Pijl & including day flying Sphingidae, nocturnally active
Volume58, Number 1 15
LepidopteraandotherinsectOrdersweresurveyedas approximately40 cm in lengdi and25 cm in diameter.
potential pollinators ofP. praeclara. Several types of An invertedwire mesh conewas placed at die baseof
passive trapsweretestedtocapture P. praeclara polli- the trap (much like a minnowtrap) and a hingedtrap
nators, and a method to mark individual pollinators doorplacedonthetopofthetrap.Threelegsattached
wasinvestigated. to the bottom oftrap allowed it to be placed over an
individual orchid. The legs were of sufficient length
Studysite and Methods that dieycouldbe pushed into the ground aroundthe
Five field experiments were established between orchid, allowing the mouth ofthe inverted cone tobe
1997and 1999todeterminedieidentityofP. praeclara suspended approximately5 to 10 cm above the termi-
pollinators. Three studv plots were located in the nalflowerofthe orchid. Traps were emptied dailvbe-
Manitoba Tall Grass Prairie Preserve, near Tolsoi, tween 0630 h and 0830 h and again between 1900 h
Manitoba (49°05'N, 96°49'W) and two additional and 2030 h. Captured insects were identified and ex-
moth light trappingsiteswere locatednear Lonesand, amined forpresence oforchid pollinia. One conetrap
Manitoba (approximately 21 km east ofthe Preserve) was placed in each ofdie three plots and rotated be-
and near Grunthal, Manitoba (approximately 26 km tween plants during the floweringperiod. Cone traps
northwest). Theclimateis aborealcontinental regime were placed over orchids that had die most flowers
with mean temperatures of 19.6°C and -18.8°C for open with intact pollinia available. Traps remained in
JulvandJanuary, respectively. Fifty-fivepercentofthe place over an individual orchid for 24—48 hours and
annualprecipitation (mean 579.1 mm) falls duringthe were placed in the plots between 1 Julyand 15 Julvin
periodofMaythroughAugust (CanadianClimate Pro- 1997, 1998and 1999.
gram 1993). Drainage in much ofthe prairie is poor Malaiseinsecttraps (Bioquip® EquipmentSpecial-
withsoilcomposedoflacustrineparentmaterial, sandy ities) were to used to sample for potential pollinators
loam to clay loam upper horizons and a thin organic over a groups oforchids (generally five or more indi-
surface layer. Stones, rocks and occasionallyboulders vidual plants). Each trap was approximately 2 m in
are also scattered across the prairie (Canada Soil In- height with a glass container at die top to collect
ventory 1989, Moore & Fortney1994). trapped insects. A piece ofVapona® insecticide strip
The Tall Grass Prairie Preserve has an abundance (0.2 cm2) wasplaced atthe endofdie collectingcylin-
ofgrasses, including Big bluestem (Andropogon ger- der to kill captured insects. One malaise insect trap
ardiiVitman) and Littlebluestem (Schizachyriumsco- was placedwithin each ofthe diree plots and rotated
parium (Michx.) Nash) (Gramineae) and is inter- among groups oforchids duringdie floweringperiod.
spersedwith bluffs ofvarious deciduous tree species, Trapswere emptied dailybetween 0630 h and 0830 h
includingwillow(Salixspp.) andpoplar(Popidus spp.) and again between 1900 h and 2030 h. Insects were
(Henne & Diehl2002). identified and examined for presence of orchid
Three 50 x 50 m plots were established in fields pollinia. Trapswere placed in die fieldbetween 1 July
were orchids were numerous. Each plot contained a and 15 July in 1997, 1998 and 1999. Traps were ro-
minimum of25 orchids. Plotswere separatedbyadis- tated within the plot to new groups oforchids every
tanceof500to 1000m. Toestablishthattherewereno three to five days.
dayflyinginsectspollinatingtheorchids, 15individual In 1999, five orchids in each plot under a malaise
orchidsweretaggedineach ofthe threeplots (totalof trap or one orchid under a cone trap were chosen to
45 plants) in 1997 and 1998. Each orchidwas visually test the effectiveness of Day-Glow® orange marker
observedfor 15 minutesonatleastthreeseparatedays dyepowderinidentifyingpotentialpollinators. Asmall
duringthefloweringperiodin 1997and 1998between amount of Day-Glow® orange marker dve powder
1100 h to 1500 h to determine if insects contacted was appliedwith aextrafine nylon brushtothe centre
plants and were responsible for removal of one or ofeach flowerofdie orchids chosen. Insects captured
more ofthepollinium from theflowers, eitherbyacci- inthetrapswereidentifiedandexaminedforpresence
dent or for the purpose ofobtaining nectar or pollen. ofpollinia removed from die orchids and also exam-
The number and identity ofpotential pollinators (Or- ined in a dark room under an ultraviolet light fordie
der, Family and Genus/species if known) landing or presence of die Day-Glow® powder. Insects widi
crawlingon die orchidswas recorded. markerparticles adheringto theirbodieswrereconsid-
Invertedcone insecttrapswere usedtosample po- eredtohave come incontactwddi die treated flowers.
tential pollinators from individual orchids. Cone traps To determine moth flight periods, two Wards all-
were constructed from light gauge steel tubing and weather insect traps® (Wards Natural Science) were
wireandcoveredin fine mesh screen. Conetrapswere placed approximately 1 km to die west ofthe orchid
16 Journalofthe Lepidopterists" Society
plotsand5kmtothesouthoftheplots. Twoadditional Table 1. Number and identity ofinsect Orders and Families
Ward'sall-weadierinsecttraps®were locatedatLone- collectedinconeandmalaisetraps 1997-1999.
sand and Grunthal, Manitoba to augment moth cap-
Order Family/Subfamily n
tures attheTall Grass Prairie Preserve. Trapswereop-
erated from 1 May to 31 August, 1997-1999. Traps Ephenieroptera Baetidae 1
Heptageniidae 1
hadaneightwattultra-violetfluorescentbulbas anat- Odonata Coenagrionidae 3
tractant. Thetrapswereusedtosurveynocturnal Lep- Orthoptera Acrididae 5
idoptera and determine the flight periods ofpotential Plecoptera Perlidae 1
Heniiptera Miridae 21
orchid pollinators. Flight periods were considered to Reduviidae 2
include the period between date offirst and last cap- Pentatomidae 2
Homoptera Cercopidae 5
ture in each year. Traps in the Tall Grass Prairie Pre-
Cicadellidae 13
servewereplacedinopenareas surroundedbymature Neuroptera Mantispidae 2
trees and were not visible from orchid plots. Lepi- Chrysopidae 9
doptera captured in traps were sorted, pinned and Coleoptera Carabidae 8
Scarabaeidae 5
identified to species. Elateridae 2
The flowering period for orchids in plots was Lampyridae 12
recorded during the study. Flowering period was de- Cleridae 5
Coccinellidae 8
finedastheperiodbetweentheappearanceofthefirst Tenebrionidae 1
flowerandlastflowerin aplot andpeakfloweringdate Chrysomelidae 5
was defined as the date when the most flowers were Curculionidae 3
Mecoptera Panorpidae 2
fullyopeninaplot.The meanoverlapin daysbetween Tiichoptera Limnephilidae 6
moth speciesflightperiodandorchidfloweringperiod Diptera Tipulidae 26
forallyearswascalculated. Mean overlapflightperiod Culicidae 55
Chironomidae 24
data was square root transformed to satisfy assump- Simuliidae 15
tions of normality and heterogeneity ofvariance for Tabanidae 4450
analysis of variance (ANOVA) (SPSS Inc. 1999). Syrphidae 45
Muscidae 546
Where ANOVA was significant a least-significant dif- Calliphoridae 228
ference (LSD) testwas used to determine the differ- Sarcophagidae 11
ences between means (a = 0.05) because ofits consis- Lepidoptera Pyralidae 14
Pterophoridae 2
tency (Saville 1990). To determine if potential Tortricidae 34
pollinators would fit pollinia distance separation re- Gelechiidae 40
quirements and nectar spur depth requirements ofP. Geometridae 18
Arctiidae 3
praeclara, measurementsweremadeonaminimumof Noctuidae 34
five individuals from all sphinx moth species collected Lasioeampidae 15
in the various trap types. Pinned moths were softened Spliingidae 6
Hesperiidae 10
in arelaxing chamberand the length ofthe proboscis, Nymphalidae 5
the distance between the outer margins ofthe com- Satyridae 2
poundeyes anddistancebetweentheinnermargins of Hymenoptera Ichneumonidae 23
compound eyes where measured in mm under a dis- VSepshpeicdiadeae 23
secting microscope. For data analysis die proboscis Megachilidae 2
length was square root transformed and distance be- Apidae 19
tweenoutereyeedgesanddistancebetweeninnereye
edges were log transformed to satisfy assumptions of
normality and heterogeneityofvariance. Morphologi- Identification of insects was based upon Hodges
cal measurements between species were subject to (1971), Rockburne and Lafontaine (1976), Morris
ANOVA and where ANOVA was significant differ- (1980), Hodges et al. (1983), Covell (1984), Borror et
ences between means were identified using a LSD al. (1989), Klassen et al. (1989), Layberryet al. (1998)
test. To examine the relationship between proboscis andHandfield (1999).
length and abilityofpollinators to retrieve nectar, the
Results
length ofnectar spurs and the depth ofnectarwithin
the spur (distance from distal end of spur to top of Observational survey. In 1997 and 1998 orchids
nectar line) was measured for orchids in each ofthe wereobservedbetween 1100h and 1500h forapprox-
direeplots. imately 70 hours. Although numerous insects (many
Volume 58, Number 1 17
Table2. SummaryofsixSphingidaecollectedfrom< ridmalaisetrapswitiiattachedpollinia1997-1999.
#of Pollinia Marker
Date Species Type pollinia location powder
11 July1997 Hijlesgallii cone .-> head n/a
11July1997 Sphinxdrupiferarum malaise 5 head n/a
15July199S Sphinxdrupiferarum malaise 11 head n/a
15July1998 Sphinxdrupiferarum malaise 3 head n/a
20July1998 Sphinxdrupiferarum cone 7 head n/a
6July1999 Hylesgallii cone 2 head present
belongingtopollinatingfamiliesorgenera) frequented Flight period. The flight periods for sphingid
the plots during die dailyobservation periods no indi- species found in the vicinity ofthe Tall Grass Prairie
viduals were observed to seek nectar or pollen from Preserve were based on the catches from the four
the orchids, to use the flowers as arestingplatform or black light traps during the period of 1997 to 1999
to sun themselves on the orchids. Individuals ofthe (Table 3). Flowering dates for P. praeclara are also
sphinxmodisHemaristhysbe (Fabricus) (16individuals) shown inTable 3. Generallydie moths collectedwere
and Hemaris diffinis (Boisduval) (28 individuals) were most abundant in the first severalweeks ofthe orchid
observedintheorchidplotsduringdieobservationperi- floweringperiod. TheflightperiodsofS. drupiferarum
ods but theywere not attracted to orchids, despite re- and H. gallii overlapped widi orchid flowering bv
peatedlv\isitingnearbyfloweringherbs fornectar. 34.6% and45.3% respectively.
Trapping experiments. Between 1997 and 1999 Morphological measurements. The proboscis
the cone and malaise traps caught 5856 individual in- length,widdi acrossdieeyes anddistancebetweenthe
sects from 49 families over45 trappingdays (Table 1). innereyemarginsweremeasuredforthe 15speciesof
The only insects found to have pollinia attached to Sphingidae collected in the vicinity ofthe Tall Grass
theirbodies belonged to the family Sphingidae (Table Prairie Preserve (Table 4). The mean orchid nectar
2). Sixsphingid modis, twospecimens ofthe Bedstraw spurlengthwas 45.27 mm (n = 1016, SE = 0.134) and
hawkmoth, Hijles gallii (Rottenburg), and four speci- the mean depth ofnectar widiin the spur was 12.44
mens ofdieWildcherrysphinx, S. drupiferarumwere mm (n = 1016, SE = 0.201).
collected with two or more pollinia attached to the
Discussion
eyes (Figs. 3, 4). H. gallii had 2 pollinia per moth,
while S. drupiferarum had3 to 11 polliniaper moth. The results confirm observations byFaegri andvan
Marking experiment. In 1999 one sphinx moth, der Pijl (1979), Sheviak and Bowles (1986), Cuthrell
H. gallii, was collected from acone trapwith traces of (1994) and Luyt andJohnson (2001) diat P. praeclara
Day-Glow® orange marker dye powder on both eyes ispollinatedbynocturnal Lepidopteraconfinedto the
and die proboscis. Both pollinia attached to the eyes family Sphingidae. It appears that there are no diur-
also had traces ofpowder, primarily on the massulae nally active insects diat seek nectar or pollen from P.
andviscidium.
Fig. 4. Anterior/lateral view, Hyles gallii (Rottenburg1. with
polliniafromPlatantherapraeclaraattachedtoeves. Notedielarge
FIG. 3. Anterior/lateralview. SphinxdrupiferarumJ. E. Smith viscidiumcementedtolower-centerofeve,widimassulaeprojected
dthpolliniafromPlatantherapraeclaraattachedtoeyes. forward.
18 Journalofthe Lepidopterists' Society
Table3. Flightperioddates,orchidfloweringperiods,numberofindividualsphinxmothscollectedinblacklighttrapsandpercentoverlap
betweenflightandfloweringdatesinthevicinityoftheTallGrassPrairiePreserveandfororchids 1997-1999.
Dati p%erOivoedrl&apfloofweflriignhgt
1997 1998 1999 n1 perioddays± SE
Peakorchidfloweringdate 2 fuly 4July 7July
Orchidfloweringdates 22June-16July 19June-20July 23June-19July
Sphingidae-flightperioddates
Ceratomiaundulosa Harris 12June-12July 7June-18July 9 |une-31July 14 92.0±4.9c-
Sphinxchersis(Hubner) 31 May-2July ]1July absent 6 14.1 ± 13.9a
SphinxkahniaeJ.E. Smith 1-26July 5-24June 10June-18July 8 58.0±23.2abc
SpliinxluscitiosaClemens absent 20May-30June 19June-11July t 34.6± 19.9abc
Sphinxdrupij"eranunJ.E.Smith 11July 1 June-25July 24June 5 34.6±32.7abc
SmerintluiscerisyiKirby 28May-27June 21 May-29July 16May-25July 58 73.7±26.3bc
Smerinthusjamaicensis(Drury) 25May-30June 1 June-lOJuly 29May-18July 123 65.3± 18.2bc
Poanesexcaecatus(J.E. Smith) 3June-20July 11June-15July 5 (une-20July 37 94.6±9.2c
Poanesmijops(J.E. Smith) 11-28June 9June^lJuly 1-27June 34 29.3± 16.9abc
Cressoniajuglandis(J.E.Smith) absent 8-30June absent 11.7 ± 11.6a
Pachijsphinxmodesta (Harris) 27Mav-28June 11June-5Aug 9June-9July 61 63.0±20.8bc
Hemaristhijsbe(Fabricus)3 31 May-25June 5-28June 4June-2July 16 25.3±6.9abc
Hemarisdiffinis(Boisduval)J 21 May-2July 10June-10July 1June-13July 28 62.0± 10.4bc
Darapsamyron (Cramer) absent 15 May-6July absent 5 18.0± 16.0a
HijlesgaUii(Rottenburg) 15June-31July 31 May-29June 6July 18 45.3±28.5abc
0.040
2.086
'Numberofmothscollected 1997, 1998& 1999.
2Meanswithineachcolumnfollowedbythesameletterarenotsignificantlydifferent(Fisher'sLSD,p>0.05).
3,4H. tliybeandH. diffiniscollectedbysweepnetindievicinityoftheTallGrassPrairiePreserve.
praeclara. The presence of pollinia on one H. gaUii (Drury) (collected by light trap) with P. praeclara
specimenandfourS. drupiferarum specimens andthe pollinia attached to the eyes. The present study adds
markerpowderandpolliniaon anotherspecimenof//. H. gallii to this list of pollinators while E. achemon
gallii confirm these species as pollinating agents ofP. does notoccurin Manitoba.
praeclara in Manitoba. Despite the larger size ofthe The mean proboscis lengths for S. drupiferarum
mm
malaise trap covering more orchids, there was little andH. gallii capturedinthisstudywere38.40 and
difference in the capture rates ofmoths with pollinia 33.50 mm, respectively. Subtracting the mean nectar
between trap types. The Day-Glow® orange marker depthfromtotalnectarspurlengthprovides adistance
dyepowdersuccessfullymarked onepollinator. of32.83 mm, thus it appears thataproboscis lengthof
Polliniawere attachedto the centerofeach eye on 30-35 mm is required to reach the nectar in P.
both moth species. S. drupiferarum had more pollinia praeclara. Based on proboscis length. Sphinx chersis
per specimen. The distance between the outer edges (Hubner) and S. kahniae mayalsobepotentialpollina-
ofdieeyeswasgreaterin S. drupiferarum thanH. gal- tors in Manitoba (Table 4). These species hadasignif-
lii (Table 4), but it is unknown ifthis small difference icantly longer proboscis than all other species col-
in eye separation (0.43 mm) would affect die attach- lected (Table 4), with no other species having a mean
ment ofpollinia. Perhaps S. drupiferarum, is more ag- proboscis length greater than 23 mm. The separation
gressive at retrieving nectar and presses harder into ofinner and outer eye margins between species was
the centre ofthe orchid thus capturing more pollinia. less distinctwithsomespecieshavingashortproboscis
Alternately H. gallii may visit the orchids less fre- yet still having similar eye positioning and separation
quently resulting in fewer opportunities to remove (Table4).
pollinia, orbe discouraged from visiting orchids once Orchid flowering periods and moth flight periods
severalpolliniahave beenattachedtotheeyes. H. gal- are restricted in terms of overlap in Manitoba. The
liiis morenumerous in thevicinityofthe orchids than overlap in more southern areas of the range of P.
S. drupiferarum, thus it seems likelythatdieyare less praeclara maybe greater and therefore higher levels
attractedto the orchids. ofpollination may occur resulting in more seed cap-
Cuthrell (1994) described two specimens of S. sules per plant. S. drupiferarum is uncommon in
drupiferarum (one caughtbylight trap andone amu- southern Manitoba and the most numerous sphingid
seum specimen) and one specimen of E. achemon speciesindieTallGrassPrairiePreservedonotpossess
Volume 58, Number 1 19
Table4. Lengthofproboscis,thedistancebetweentheoutermarginsofthecompoundeyesanddistancebetweendieinnermarginsofcom-
poundevesforSphinxmodiscollectedinthevicinityoftheTallGrassPrairiePreserve, Manitoba.
Meandistance Meandistance
Meanproboscis betweenouter betweeninner
lengtii eyemargins eyemargins
(mm± SE) (mm ± SE) (mm± SE)
Cerato/niaundulosaHarris 9.14 ± 0.76c2 5.05±0.05h 2.04±0.02d
Sphinxchersis (.Hubner) 40.32±0.4Sh 5.89±0.03k 2.40±O.llef
SphinxkahniaeJ.E.Smith 33.64± 1.31g 5.31 ± O.lTbi 2.05±0.18d
SphinxluscitiosaClemens 22.56 ±0.17f 4.46± 0.05ef 2.34±O.Olbc
SphinxdrupiferarumJ.E. Smith 38.40 ±0.97h 5.59± 0.06i 2.34±0.02e
SmerinthuscerisyiKirby 2.21 ±0.06a 4.76±0.13g 1.88±0.03c
Smerinthusjamaicensis(Drurv) 1.73±0.75a 3.85± 0.03c 1.70±0.02b
Poanesexcaecatus(J.E. Smith) 2.89±0.03b 4.63±0.05fg 2.09±0.03d
Poanesmyops(J.E. Smith) 1.87±0.04a 4.07±0.03d 1.83±0.06bc
Cressoniajuglandis(J.E. Smith) 2.10 ±0.05a 3.31 ±0.08a 1.37+0.53a
Pachysphinxmodesta (Harris) 1.95 ±0.17a 5.73±0.07j 2.53±0.02f
Hemaristlujsbe(Fabricus) 12.75±0.52d 3.88± 0.02c 1.88±0.08c
Hemarisdiffinis(Boisduval) 9.48±0.20c 3.58±0.06b 1.89 ±0.03c
Darapsamijron (Cramer) 17.68±0.18e 4.28±0.09e 2.12±0.04d
Hylesgallii(Rottenburg) 33.50±0.96g 4.88±0.03gh 2.44±0.02ef
p = 0.001 p =0.001 p=0.001
Fh--37-1 F1M.9S=3J5J"2 F2,9s=40.1
1Numberofmothsmeasured.
-Meanswithineachcolumnfollowedbythesameletterarenotsignificantlydifferent(Fisher's LSD,p>0.05).
aproboscis ofsufficient length to take nectar from P. Baker (1961) stateddiat sphingophilous flowers of-
praeclara. H. gallii populations fluctuate greatlvfrom ten have a low frequency ofpollinator visitation and
yeartoyearinsouthern Manitoba, oftenwithveryfew compensate by producing numerous seeds. Further
individuals appearing in some years. Therefore, low researchis requiredtodetermineifothersphinxmoth
pollinatorpopulations maybe restrictingseedproduc- species are pollinatingorchids and ifthe current level
tion insomeyears fororchids in Manitoba. ofseedpodproductionis abnormallylowornormalfor
The Tall Grass Prairie Preserve is surrounded bv P. praeclara in Manitoba.
agricultural lands, a mixture of intensivelv farmed
Acknowledgments
grains and oilseeds and livestock production. The lar-
valhost plants for S. drupiferarum canveryregionally This project was funded by the following agencies: Manitoba
andincludeMains spp., Prunus spp. andlilac, Syringa CMoannsietrovabtaioSnu,sttahienaUbnlieveDresivteyloofpWmiennntipFeugn,dMaannidtodbieaWOrocrhliddWSiolcdileitfve.
vulgaris L. (Hodges etal. 1983). Theseplants arepre- Fund.TheauthorswouldliketoacknowledgeJasonGreenall,Peggy
sentinthevicinitybuttheirdistributionis patchy. Itis Westhorpe, Lome Heska, LizReimerandTracyLoeweninprodd-
unknownwhetherinsecticideusage (Suzan etal. 1994) iBnogoansesifsotranrceeviaenwdinsguptphoertmatnoudsicirsispttu.dPyhaontdogDrra.pCh.sWbavnRg.\a\n"desRtawcohoed.l
on adjacent farmlands may restrict the population of
larvalforms ofS. drupiferanim andH. gallii orifweed Literature Cited
control may restrict access to host plants. S. drupifer- Baker,G.E. 1961. Theadaptationoffloweringplantstonocturnal
arum maybeamore efficientpollinatorthan H. gallii, andcrepuscularpollinators. The Quarterly ReviewofBiology
bthuettohrechniudmfbieelrdsowfaSs. dlrouwpiinfetrhaisrusmtuaddyu.ltTshefrreequmeanytibneg Bjuc3so8rt:ca6hd4i,d-7A1.(.PJl.a&tanWt.heFraoprrtauenclea.ra)1:989m,onTihtoeriwnegstearnndprraeisreiaercfhr,inpgpe.d
considerable competition between nectar sources 197-199. In Bragg,T B.&J. Stubbendieck(eds.),Proceedings
given the amount ofintensively managed lands adja- oftheElevendiNorthAmericanPrairieConference. University
cent to the Tall Grass Prairie Preserve, which would BjucosftaNde-bProarsktae,r,LiRn.co1l9n9.32.92Thpep.westernprairiefringedorclud(Pla-
furtherreducethetime spentbymothspollinatingor- tantherapraeclara):itsresponsetoburningandassociatedfungi.
chids. Stronglightsources from farmssurroundingthe M.Sc.(RangeManagement)Thesis.RangeManagementDepart-
orchid fields mayalso attract moths, preventing them BORKmOeWnStK,YU,nCi.veLr.sit1y99o8f.WyMoamniintgo,baLaTarlalmiGer,asWsyPormaiirnige,Pir—e\sleir+ve1—16npapt.-
from visiting orchids. S. drupiferarum occupies the uralresourcesinventoryIII. CriticalWildlife HabitatProgram.
wnhoritchhemrnamyosaltsoecxotnetnrsiibounteoftoistsporraadnigcepionpulMaatniiotnoboac,- Borr3o0r,pp.D.J.,Triplehorn,C. A. & N. F.Johnson. 19S9. Anin-
troduction to die study ofinsects. 6di ed. Saunders College
currence andlowerpollination rates ofP. praeclara. Publishing,Toronto.875pp.
20 Journalofthe Lepidopterists' Society
Bowles,M. L. 1983. ThetallgrassprairieorchidsPlatantheraleu- Klassen,P.,A. R.Westwood,W B. Preston &W. B. McKillop.
cophaea (Nutt.) Lindl. and Cypripedium candidum Muhl. e.v 1989. ButterfliesofManitoba. Manitoba MuseumofMan and
Willd.: some aspects oftheirstatus, biology, and ecology, and Nature. 1989Winnipeg,290pp.
implications towards management. Natural Areas Journal Larson,R.J. 1992. PollinationofPlatantheradilatatavar.dilatata
3:14-37. inOregonbythenoctuidmoth Discestraoregonica. Madrono
Bowles,M.L.,R.Flakne&R.Dombeck. 1992. Statusandpopu- 39:236-242. W
tlahteiroanlfeluucctoupahtaieoans(oNfutdti.e)eLaisntdel.r)nipnraIilrliienoifsr.inOgreedgoarncohi1d2:(2Pl6a^t4a0n.- Laybteerrrflyi,esRo.fAC.a,nPa.da.HNaRlClR&esJ.eaDr.chLaPfroesnst.aiUnniev.er1s9i9t8y.ofTThoerobnutto-
Bray,T. E. &B. L.Wilson. 1992. StatusofPlatantherapraeclara Press,Toronto.280pp.
Sheviak&Bowles(westernprairiefringedorchid)inthePlatte Luyt, R. & S. D. Johnson. 2001. Hawkmodi pollination ofthe
RiverValleyin Nebraska from Hamilton to Garden counties. African epiphytic orchid Mystachlium venosum, with special
TransactionsoftheNebraskaAcademyofScience 19:57—-62. reference to flower and pollen longevity. Plant Syst. Evol.
CanadaSoilInventory. 1989. SoillandscapesofCanada Man- 228:49-62.
itoba. LandResourceResearchCenter,ResearchBranch,Agri- Moore, R. & G. Fortney (eds.). 1994. Management plan for
J.
cultureCanada,Ottawa,Ontario.Agric. Can. Publ.5242/b. 22 the tall-grass prairie preserve in Manitoba. Critical Wildlife
pp. 1:1 million scale maps compiledbyCanada-Manitobasoil HabitatProgram.36pp. + appendices.
survey. Morris, R. F 1980. Butterflies and mothsofNewfoundlandand
Canadian Climate Program. 1993. Canadian climate normals: Labrador:theMacrolepidoptera.Research Branch,Agriculture
1961-1991. MinisterofSupplyandServices.Rpt.No.56-61/2- Canada.Publication 1691.407pp.
1993.266pp. Nilsson,L.A. 1992. Orchidpollinationbiology.TrendsinEcology
Collwihciuttet,friDn.geRd.or1c9h9i3d.PlSattataunstherreapoprrtaeocnlatrhaeinweCsatneardna.prCaoirmi-e Patt,andEMv.o,luMt.ioWn7:M25e5r-c2h5a9.nt, D. R. E. Williams & B. D.
J. J.
mittee on the Status of Endangered Wildlife in Canada Meeuse. 1989. PollinationbiologyofPlatantherastricta(Or-
(COSEWIC).39pp. chidaceae) in Olympic National Park, Washington. Amer. J.
Covell, C. V., Jr. 1984. A field guide to the moths of eastern Bot.76(8):1097-1106.
NorthAmerica. HoughtonMifflinCo.,Boston.496pp. Pleasants,J. M. 1993. Pollinationecologyofthewesternprairie
Cuthrell, D. L. 1994. Insectsassociatedwiththeprairiefringed fringedorchid.Platantherapraeclara. North DakotaParksand
orchids, Platanthera praeclara Sheviak & Bowles and P. leu- TourismDepartment.24pp. +figures.
cophaea (Nuttall) Lindley. MSc. Thesis, North Dakota State Pleasants.J. M. & S. Moe. 1993. Floraldisplaysizeandpollina-
University,DepartmentofEntomology.76pp. tionofthewesternprairiefringedorchid,Platantherapraeclara
Darwin, C. 1904. Thevarious contrivances bywhich orchids are (Orchidaceae). Lindleyana8:32—3S.
fertilizedbyinsects.2nded.JohnMurry,London.300—pp. Proctor,M„P.Yeo&A.Lack. 1996. Thenaturalhistoryofpolli-
Davis, S. K. 1994. ManitobaTall Grass Prairie Preserve natural nation.TimberPress,Portland,Oregon.479pp.
resourcesinventoryI.CriticalWildlifeHabitatProgram.48pp. Robertson,J. L. & R.Wyatt. 1990. Reproductivebiologyofthe
. 1995. Nationalrecoveryplanforthewesternprairiefringed yellow-fringed orchid, Platantera ciliaris. Amer. J. Bot.
orchid,Platantherapraeclara. CanadianNatureFederation,En- 77(3):388-39S.
dangeredPlantandInvertebratesofCanadaProject.23pp. Rockburne,E.W.&J.D.Lafontaine. 1976. Thecutwormmodis
Dressler,R.L. 1981. Theorchids:naturalhistoryandclassification. of Ontario and Quebec. Research Branch, Canada Dept. of
HarvardUniversityPress,Cambridge, Massachusetts.332pp. Agriculture.Publication 1593. 164pp.
. 1993. Phylogeny and classification ofthe orchid family. Sather, N. 1991. Recoveryplanforwesternprairiefringedorchid
DioscoridesPress,Portland,Oregon.314pp. (Platanthera praeclara Sheviak & Bowles). Unpublished Re-
Faegri, K. &L.van der Pijl. 1979. Theprinciplesofpollination port.St. Paul, Minnesota.
ecology.WPergamonPress,NewYork.244pp. Saville,D.J. 1990. Multiplecomparisonprocedures:thepractical
Hodges,R. 1971. ThemothsofAmericanorthofMexico.Fas- solution.TheAmericanStatistician44(2):174—180.
cicle21.Sphingoidea.E.WClasseyLtd.&R.B.D.Publications Sheviak,C. &M.L.Bowles. 1986. Theprairiefringedorchids:
J.
Inc.,London. 158pp. + colorplates. apollinator-isolatedspeciespair. Rhodora88:267-290.
Hodges, R.W,T. Dominick,D. R. Davis, D. C. Ferguson, G. Sieg, C. H. & R. M. King. 1995. Influenceofenvironmentalfac-
Franclemont,E. G. Munroe& A. Powell. 1983. ChJ.eck torsandpreliminarydemographicanalysesofathreatenedor-
J.
list ofthe Lepidoptera ofAmerica north ofMexico. National chid,Platantherapraeclara.Am. Midi.Nat. 134:307-323.
Museum of Natural Histoiy. MRC-127, Washington, D.C., SPSSINC. 1999. SPSS,version10.0.0forWindows.Chicago:SPSS
U.S.A.284pp. Inc.
Handfield,L. 1999. LesguidedesPapillonsduQuebec.Broquet Suzan,H.,G.P.Nabhan&D.T.Patten. 1994. Nurseplantandflo-
Inc.657pp. ralbiologyofarareNight-BloomingCereus,Peniocereusstriatus
Hennmeo,diD,.HCe.mi&leuJ.caK.neDviaedheln.sis20(0S2a.tuNronitiedsaeo)ninthseouNtehveardnaMbaunci-k U.S.(BFriasnhdeg&ee)WiF.ldBluixfbeaumS.erCvoincsee.rv1a9t9i6o.nBPiloaltoagnyth8e:r4a61p^r7a0e.clara
toba. Lepid. Soc.56:112-114. (western prairie fringed orchid) recoveryplan. U.S. Fish and
J.
HOF,J.,C.H.Sieg&M.Bevers. 1999. Spatialandtemporalopti- WildlifeService,Ft. Snelling, Minnesota,vi+ 101pp.
mizationinhabitatplacementforathreatenedplant:diecaseof Van derpijl, L. andC. H. Dodson. 1966. Orchidflowers: dieir
the western prairie fringed orchid. Ecological Modeling pollination and evolution. University of Miami Press, Coral
115:61-75. Gables,Florida.214pp.
JOHNSON, K. L. 1985. Rare plant alert: another new orchid: die Wyatt, R. 1983. Pollinator-plantinteractionsanddieevolutionof
prairie white fringed orchid (Platanthera leucophaea). Mani- breedingsystems,51-95pp.In Real, L. (ed.),Pollinationbiol-
tobaNaturalistsBulletin8:12-13. ogy.AcademicPress,Inc.,Orlando,Florida.338pp.
. 1991. Marking News, Botanical Survey of Manitoba.
Newsletter#6. Manitoba Museum ofMan and Nature, Win- Receivedforpublication 30January2003; revisedandaccepted29
nipeg, Manitoba.6pp. July2003.