____...IIII:~·i::s;c:.: '

THE BIOSYSTEMATICS :OF PROPULLANS GRAY AND

SYMPATRIC POPULATIONS OF ERYTHRONIUt~ ALBIDU~1 NUTTALL ()

AThesis Submitted to the .. Faculty of

University of Wisconsin" LaCrosse La Crosse, Wisconsin 54601

Jo Ann Banks

lriPartial Fulfillment bfthe Requirements for the Degree

• of

Master of Science in Biology

August 19]8

© Mfr­ -1<,/; 8tL'n

UNIVERSITY OF WISCONSIN .. LA CROSSE La Crosse, Wisconsin 54601 COLLEGE OF ARTS, LETTERS, AND SCIENCES

Candidate: Jo Ann Banks

We recommend acceptance of this thesi.sto. theColl ~g~ofArts.'. L.etters, and Sciences in partial fulfillment of this candidates' requirements for the degree Master of Science in Biology. The candidate has completed her oral defense of the thesis.

'l3i.?Y Date ~ ,-31-7!- ::::::::r; '. ~ y~ ~':::? - .v"=" .JI.. zr~--'- ~.- Datj,b '61(g/&a: ~J'}~p:~ ../t)at-e

This thesis is app"oved for the College of Arts, Letters,

79-02634 ii

ABSTRACT

A biosystematic.comparison of Erythroniumpropullans.ilndsympatric populations of f. albiQum was conducted to clarify the relationships of these •.. twospecte.s.. .iE.propu11 ans . iSiendemicto,southeaste.rn.Minnespta., while I. albidum is common and widespread throughout the eastern , including southeastern ~1i nnesota. Themorphol ogy,geography, .cy­ tology, ecology, and reproductive biology of each species were investigated. Populations of E. albidum and E. propullans within Rice and Goodhue Counties, Minnesota, were utilfzedinthesttidY.. A morphological analysis of six. charactersjndicateclthatt. albidum and I. propullans are morphologically distinct species, and that th,ree size classes of exist: f. propull ans, depictedasa.dwarfed f. ~­ bidum, an intermediate size class, and L al:!~jdum. The intermediate size class suggested that La1bidum andE:proplrl-ra:ns--hyhri-d-Fze-te--formferti le offspring. The rari ty.of the hybridfurther.suggested that sOme repro- ductive isolation between the two species exists. The pollen-ovule ratio, protandry, and the dimorphi c of I. propull ansand I .. al bi dum suggest that both species are facultatively outcrossing. Breeding experiments demonStrated that I. albidum is facultatively outcrossing and able to pro~ duceseed when. crosseclwith I. propul1etDS_. Selt-ferti1i zationi~E.a 1­ bidum required a physical mechanism to transfer pollen f~oril-tlieantherto the. . E. propull ans, however, produced seed only when crossed with E. albidum. Pollen .was found to be effectively transferred either between C10rle?of I. albidum ()Y' f. propullans by AndrenacarliniCkl1., a solitary oligolectic bee whose behavior waswells~i~~

ACKNOWLEDGEMENTS

I am expressly grateful for patience and perceptive. advice during the course of this study to its completion. I am also especially grateful to Dr. Jerry Davis, my major advisor, and to Dr. Thomas Claflin, Dr. S. H. Sohmer, Dr. Thomas Weeks, and Dr. Weinzierl for their advice and .guidancewhiJe serving on my thesis conmittee. Mr. Mark Heitlinger and the Nature' Conservancy of Minnesota is gratefully acknowledged for their interest in and funding of this study. Each of the following individuals have made invaluable contributions to this study for which I am most grateful: Dr. Daniel Crawford, for his advice. and preliminary investigations in the chemotaxonomy of Erythronium; Dr. H~gh Iltis, for introducing me to the problem; Dr. Thomas Morley, for his advice and suggestions concerning ; Mr. Orwin

Rustad, for assisting me in locating populations ".+ c .....O/-h .... " ... and S. W. Batra, Dr. Lloyd Knutson,A. Menk~, and G. of the I sect Identification and Beneficial Insect JntrOcIuction Institute, Belts­ ville Agricultural Research Center, for their time and effort ih the identification of insects. Bruce Biltgen, Steve Swanson, and Diane Kell are given a special thanks for all of the help they have given me with this thesis during ~he past three years. iv

TABLE OF CONTENTS

PAGE LIST OF TABLES .. ·...... LIST OF FI GURES viii INTRODUCTION...... 1 LITERATURE REVIEW ... 2 .. •• ..... 41: ••• 2 Cytology . ; ••. ·5 Life Hi story .. 5 STUDY SITE DESCRIPTION ...... 10 Regional Description . 10 toea1ity Descri pi ons .. 17 MATERIALS AND METHODS ...... 25 Morpho logy ... 25 Cytology . '...... 25 Demographics and Phenology . 26 ·Pollen and Ovule Production. 26 Pollen-Ovule Ratios ..... 27 Vectors of Pollen Transfer. 27 .~r~§cijrJ9 ~ tllciJ §!s 27 RESULTS ...•• .. 33 MOrphology •...... ••• 33 Cyto logy.. .•...•••• Demog'raphics and Phenology ... Po 11 en and Ovu 1e Production. •. 44 Pollen-Ovule Ratios ....•. Observations of Floral Biology ...• Observations on Pollination Biology. Breeding Studies . 55 Seed Di spersa1 . . • ...... DISCUSSION . 74 ""MQrRll()~l(lgy,.,~,==c'.=,"=~L.. •.•••••• •• ••• '.. .•..• ••• 74 Demographics and Phenology . ·...... 74 Reproductive Capacity...... 78 Vectors of Pollen Transfer •. 78 Breeding Studies ...... ' ...... 80 Evolutionary Relationships...... 82 Species Persistence. ••••••• ·...... 83 CONCLUSIONS •••••••.••• 85 ...... 06 ...... '" . .. .. ·XION3ddV La . ·031IJ 3~nlV~31Il "vi

LIST·"OF TABLES PAGE TABLE 1. Characteristics comparing .... ~.~.-_ ... (from Ireland, 1975, and D;;-h"",+.-,,,,.. 1Qhh\- ~- - .• •• 6, 2. COlllpariSonamongsites of 5 morphological characters an- alyzed in E. al bidum ....•.... - .... 34 3. Comparison among sites of 5 morphological characters an- alyzed in E.. propullans .....•, . .. .. 35 4. Comparison of morphological characters between E.pro- pull ans, the intermediate sizeelass, andLalbidum. .... 36 5. Number of flowering and non-flowering individualsofE. propull ans within: a .25 m2 area,wtth.th.e.g.pproximat~=- il.reaeach respective clone occupies. ...•...... 43 6. Pollen and ovule production and pollen-ovule ratios of ,I. propull ans and I. a1bidum ...... •. 45 7. Percent stainable, unstainable and micropollen in E. pro­ pl.lllans andLal bidum ... ~-~---;-;--~--; ••.··•..i=...·.. J"~~~_,A.:~"~".j:'rbb between the sites utilized in 1977 the occurrance of apomixis in I. pro- ...... 56 9. Comparison of data between the sitesutilizedin1977 and 1978 measuring the occurrance of self-fertilization in I. propullans and I. albidum... . .•.•..• ... •...... 58 10. Summary of res from treatment to det~rmine apomixis and self-fertilization within f. propullans and I· albidum in 1977 and 1978 .•.•....•...••....•.• .. 59 11. Comparison of data between the sites util ized in 1977 and 1978 measuring the success of intraclonal crosses in

m ~... E. a1bi-dum· ,·e ..• ..>:.•.... r.···.. =oc=...... 60 12. Summary of results from treatments to determine success of intraclonal crosses in I. propul1ans~~~~. albidum in 1977 and 1978 ...... •..• •• . - - . .. 61 vii

13. Comparison of data between sites measuring thesucc.ess of interclonal crosses in I. propullans and I. albidum in 1977 and 1978. •.••••••• .• •.•••.•••• 63 14. Summary of results from treatments to determine success of interclonal crosses in I. propullans and I. albidum in 1977 and 1978. .. ••••..••.•.•.•••.•.• . 64

15. Comparison of data between thesitesuti.l.i.iedjli-~J9_IZ~ and 1978 measuring the success of jnterpopulational crosses in E. propullans and I. albidum .•.... " ..• •. 65 16. Summary of results from treatments to determine the success··ofinterpopulational crosses in I. propullans and E. albidum in 1977 and 1978 .•...... 66 17. Comparison of data between sites measuring thesuc.cess of interspecific hybridization between I. propullans and E. albidum in 1977 and 1978 .....•.••. •. .•• .• 68 18. Summary of results from treatments to determine success of interspecific hybridization between I .. propullansand E. a1bi dum i n1977 and 1978...-.--.-.---.-.----.---.--.-----.-----.-.. •. 69 19. Comparison of data between sites measuring the natural ~evel of seed production in I. propullans and I. albidum ln 1977 and 1978. ....•. • ...•.•. .. •.•• .. 70 20. SUl1llTlary of results determining the natural level of fruit and seed set of I. pro(JUlTa:i1s-and-r.--allJldUffiI--plus the annual seed crop of each species...... ••.•• .• 71 21. Comparison between I. propullans, the intermediate size class and E~albidum for 6 morphological characters ... •. 75 22. Comparison in the morphology, ecology, and chromosome number of I.propullans, I. albidum, and I. mesochoreum . .. 76 23. Summary of results obtained from breeding experiments in I. propullans 'and I ..albidum. ...•..••••.•.• .• 81 v;;;

LIST OF FIGURES

FIGURE Range of distribution of E. albidum! E. mesochoreum! 1. andE. propullans .... -: ...•... -:.m •.... ~ ... . . 4 2. Distribution of I. propullans in Rice and Goodhue Counties! Minnesota...... •.•. .. 12 3. Distribution of E. propullans near the terminus of the Des Moines lobe at the Bemis morraine during the CarjPhase of Wisconsin glaciation (adaptedfromWr;ght! 1956) .....•...... ·················14 4. Major presettlement floristic provinces in Minnesota and location of I. propull ans pOPulmatforimsU{~d~p~ea--~- from Marschner! 1930) ...... •... c ••·•• •• •• 16

ll 5. Extent of "Big Woods type vegetation and distribution •of I. propullans populations in Rice and Goodhue Counties! Minnesota (adapted from ~1arschner! 1930) . ... 19 ('~O 6.. Location of study sites in Rice and Goodhue . ... 21 • III a • • .. . ..

7. Bagged E. propullans flowers after treated for one of the various crosses •.....•.••.•• .. ... 31

8. An I. propullans! intermediate sized!.and I· albidum flowering individual ... .. _ ...... •.•• . .. . 38

9. "fneflowersof E. albi·dum (top)! theinterlTlediafesize class! and I. prbpullans ....•.•...••.••.... 38 10. The lateral stem offshoots of plants of the intermediate size class . 11.lhefruitofEcocalbidum (left) !the t[lterll1~

14. i.. propullans (left) and I.albldum flowers showing the lnner and outer whorl of petalsandstamens...... •. ... 48 15. dimorphism in E. albjdum,theshbrt~Y'\'Jh()\"l dehiscing prior to the longer inner whorl.•••.••••~.~ ••.. 48 16. Stamen dimorphism in I. Qropullans...... •...... 48 49 17. E. propullans showing 4-merous condition..... 49 18. E. P.!:Q.Pullans flower showing 5-merous condition. 49 19. E. propullans flower showing 6-merous condition...... 51 20. Andrena carlini Ckll .. . . . 21. Andrena carlini entering an I. albidum blossom ...... 51

Andrena carlini collecting nectar frornI· t.. ~t'-··_··­ 22. 53 blossoms ......

23. Andrena carlini collecting nectar from 53 blossoms . ., .0 •--••••••••• . .. . 54 24. Andrena carl ini .wi th poll en sacs on .rear.J egs. . 25. Bombylius major collecting nectar from I. ~bidum...... 54 26. Seeds of I. propullans (left) and I· albidum(right) presentQn~..§.ach ng the outgrowth of the seed ...... 73 coats 0 •••••••••••••••• .... .------

INTRODUCTION

and critical habitat, to aid in the recovery of. stressed populations, to implement stewardship'plans that will help insureit$persistence. 2

LITERATURE REVIEW

Taxonomy The genus Erythroni um extends throughout the north.erntemperate regions of the world. The greatest species diversity of Erythroniumoccurs in western North America (Applegate, 1935). In eastern North America, the species are separated by flower color into a yellow group of five taxa (Parks and Hardin, 1963) and a white group of three taxa. The white group is composed of I. propu 11 ansGray,E;ralhidum_Nutt~.~and __ E.mesochoreum Knerr. I. albidum has the largest geographic range (Fig. 1) extending from '"Ontario and Minnesota southward through Kentucky, , , and , and eastward to Pennsylvania. mesochoreum is known from , , eastern , eastern , Oklahoma and Texas. I· pro­ pullans is known only from southeastern Minnesota. Asa Gray, who first described I. propullans in 1871, cited three distinguishing features of I. propullans to warrant its specific ranking: one-half the lengthoftheE.albidum flower, , the origin of a lateral offshoot or runner just below the base of the rather than from the' base of the as in I. albidum, and the color of the flower which he described as ,bright pink or rose. Rosen­ t1ahl 1919 described the variation in perianth parts, a fourth distin­ guishing feature of I. propullans. The f1ower$b~y~either 4, 5, or 6 stamens and petals. Variations in the number of perianth parts has not Fig. 1. Range of distribution of I. albidum, I. mesochoreum and I· pro­ pullans. 4

~ albidum. ~. mesochoreun _ ~. l2!2P-ullans 5

been repo~ted in other species of the genus,. StamendimorphJs.m is common in Erythroniurn(Applegate, 1935; Graff, 1916S Meads, 1893; Pickett, 1917;

Utech and Kawano, 1975a; 1975b}, including I. albidum,I.rileso¢horeum,

(Robertson, 1966) and E. propullans (Rosendahl, 1919). E. mesochoreum is morpho1ogi callysimilar'l;oiE. albfdlIll'Cal'1dwas re- - ..- J . garded as a variety of ~ •. albidum (Knerr, 1891a; 1891b;Rickett, 1937; Robertson,1966). The morphological, ecological, arid chromosomal differences

between the two species as described by Ireland (195ZLand Robertson (l966)

are listed in Table 1.

Cytology

The base chromosome number of thegemls .Eryt;hron·;'l:1m-+X==Jl or 12)

was estab1ished by Utech and Kawano (1976). The yellow ,ETYthrorli urn , spede,? of the eastern United States are represented by dip1oid( 2N==24)

and tetraploid (2N==48) species and Hardin, 1963). Poly- within the white flowering taxa: 2N==22 in I. reland~ 1957; Robertson, 1966} and 2N==44inE.,?lbidum (Cave,

1939; Ireland, ~957; Robertson, 1966;Sllirth, 1955). The chromosome number of I. propullans has not been reported.

Life Hi story Several aspects of the life history of I. albidum, with pmnhas .y~g~ta:tj},1gY'gprQgltcti~Hl,have beE!n investigated by Blodgett (1894; 1900; . 1910} and Schaffner (1901). Being a vernal species, the shoots first

emerge in Harch or early April. By the time the forest canopy has formed,

~I'\n.."v~m... +,,'u the first of June, the leaves of Er.xthroniumare well into

life cycle is therefore begun and completed 6

Table 1. Characteristics comparingE~ albidumandE. mesochoreum .. (from Ireland, 1957, and Robertson, 1966) - -

E;mesochoreum ..-E. albidum

flower length 2.7-3.3 cm completely reflexed usually straight, pt::rianth shape sometimesreflexed

width (flowering 1.0~2.5 cm 1.U-Z.O cm p)) leaf traits usually flat and usually conduplicate • mottled and nonmottled

1 or 2 rl.J~r1-ners"_"--·"~-·--=cccc"'=r-'dropper propagation of ,I-leaved forms .. prairies, pastures, habitat preference shaded ravines and moist woods dry .•• woods

haploid chromosome number 7 within a two month period. Vegetative Reproduction

During the flowt;!rillg phenophaseOf E. albidlllll(late Apl"'il or early May) one to three runners up to 20 cm long form from the base of the non­ flowering individuals. Occasionally, one runner forms at the base of the flowering plants. One new bulb is formed attht;!apex of each runner. If runners are not formed, then new bul bsare formeciaciJac§ntto the.. parent bulb in the f1 oweri ngand non-flowering plants.. If.. bulbsarenot formed, the persistence of that individual is terminated. Vegetative persistence of I. mesochoreum by bulb multiplication is similar to that of I. albidum.

l The short runners of I. mesochoreum are called "elrOppers ! (Ireland, 1957)

Eo mesochoreum produces "droppers II each year, wlli 1e I. albi dum produces

"droppers ll only from. first year bul bs.• In flowering plants of I. propullans, the runner originates not from J the base of the bulb of the parent , but from the sheathed below the leaf bases at a distance abOve~tne--51nb-tGraY;-~1871;Blodgett V;ilnl"lpV;pllshes out from the leaf sheath , elongates, and produces One-third of the vascular bundles below the point of stolon origin continue up the flowering stalk (Blodgett, 1909). The remaining two-thirds of the bundlesentt;!rintotherlmner. According to Blodgett (1909), this diversion of vascular tissue from the fTower stalk results in the dwarfed size of the I. propullans flower. In the non-flowering I. propullans individuals, the formation of "runners and

"'~it""'.j;'';;';';;''i.L;';'LL'''''~~.cJC L..;.. ... - - . f Ioweringt:·alhidumim:ticviduals·; The production of up to three new bulbs each year by ea.chparellt plant results in populations of E. albidum and I. propullans with extensive, dense clones. B

E. albidum bulbs .present at the.end()f the fruiting season remain. - . . . dormant until the la~ter part of October when leaves and flower buds differentiate within the subterranean bulb (Blodgett,189Ll) •. The bulb ceases further development and overwinters until the following spring. Sexual Reproduction I. albidum plants occur in two forllls: thesingle-,;;reaVed,non-flewering form, the two-leaved, flowering form. Because the bulbs of the flowering plantswereJ.I$ually larger in size and deeper in the soil than the bulbs of the non-flowering plants, Blodgett(1900)$uggested that flowering was dependent on the bulb attaining the IInecessary vigor, depth, and size". He estimated that a minimum of five years was required for a seed to become a mature plant capable of flowering. Studi es to determi ne the 1eve1 ofapomrxfs-~]~neae-greeofcompa t­ i'bility or the success of experimental hybridization have yet to be per­ formed in any'Erythronium species. A relationship between pollen-ovule ratios and breeding systems has beeninves-t-i·gacted--by..... Cru.denJ19ZZh this -techliiqUehasnotbeenattempted for any· sped·es .of .•• Erythroni um. Poll inatorshavebeen surveyed only for E.- mesochoreumand --E. ja- poni cum (a Japanese· species) • Utech. and Kawano (197§a).•descri bed the pollination behavior of Xylocopa appendicuJatilcircumvolans (a bumble bee). This insect was well suited to the flower of E.japoniculll and effectively pollen between clones. An ethological study of Andrena ery- -... _.... Ckll. by ~1ichener and Rittenmeyer (1956) established that

··'···sectga'theredpollencof f. mesoeiTfrreum,ho'",'e'1er,mu'-f.!i~Lenc:e\'Jas made to the efficiency of Andrena erythron;i as a vector of pollen transfer. The ecology and pollination behavior of solitary bees; including Xylocopa and 9

~nd~, has been reviewed by Li nsl ey (1958) and Faegri and van del" Pi j 1

(1971) . The fruits of E. albidum develop dUY'lngthe latter part of May - .---- corresponding with the closing of the forest canopy. At this time, the leaves begin to senesce. By the end of July, the runners, parent bulbs, and fruits are gone, 1eavi ng only ne.w bul bs and seeds to conti nue the 1i cycle the following spring. The annualseedcropofE.albidumaverages only two· seedsperfloweri ng p1ant (Strui k,1957).Se.rnander.(1906) noted that an outgrowth (caruncle) present 00 the seed of E.'denscanis .is utilized as a food source by which harvest and disperse the seeds. This phen- omenon was reviewed by Gates (1943).

Prior investigations elucidating the biologyofErythronium provides minimal information concerning ~. Eropullans, perhaps the most intriguing

es within the genus. ~1any questions remain to be answered. What

the origin l and subsequent evolution·o-f-E·=-propuTTa.-ns?~VJhYisits

presentdi on so'limited? What is its 'relatiOnship to ~. albidum

and E. lllesochoY'el.lll1?Why is §.. propuHanS dwarfed? What can be done to insure the persistance of its populations? To anSWer these questions, , a field study was designed that would elucidatethe

cytology, ecology, and reproductive biology of ~. propullans and sympatric

populations of ~. albidulll as a comparative standard. 10

STUDY SITE DESCRIPTI

Regional Description All documented populations of f. propullans are Counties of southeastern Minnesota (Fig. 2). the two county area s dissected by the Cannon River, the Straight River, the Little Cannon River, and the headwaters of the North Zymbro River. Popula.tions of f· propullans occur in small ravine systems connected to each of these rivers. The bedrock of each valley is predolTlinatelySt. Peter Sandstone; overlying

alluvial soils are loamy and well drained (Carlson ~.~., 1975). Glacial till. deposits of pre-.Wisconsin glaciation persist in Rice 'and Goodhue Counties. Two Wisconsin morrainal ridges, one east and one of Faribault, Rice County, are evident. According to Wright1s (1956) interpretation lof Wisconsin glaciafio!1-in-Mfnne!mta;--theeastern Bemis orraine represenfsfhe extent of movement of the Des Moines ice lobe during the CarY substage of glaciation; the western Altamont Morraine was

.f'n...morl during the retreat of the Des Moines lobe during the later Mankato substage. populations of f. propullans exist adjacent to, but outside the

~~------""-"-""""""""" area of Wisconsin glaciation (Fig. 3). ll The floristic province in which f. propullans occurs is the "Big Woods of Minnesota (Daljbenmire., 1936), a mesic hardwood forest dominated by -sugar" "mapie, bassw-ooa~ and elrif:-' Prior- to- human=se--ttlemen~,."'BigWoodsll type vegetation in Minnesota extended from Rice and Le Sueur Counties in .+hn~~+ to Mahnomen County in the northwest g. 4). The present distribution of I. propullans and the pre-settlement extent of IIBig Woods" ...,ttl o II)

-0 s:: n:s

s:: ..... ,

.....Ol I.J.. ::OFV====:;==~==------n lb a... *o n

I!""

c Q ::l (It -0 o -0' c Q o-' :::J '"

ZI "

Fig. 3. Distributioh of I. Er0pullans near the terminus of the Des Moi nesJobe().~tDeBelllis morrai ne formed duri ng the Cary Phase of Wi scon- sin glaciation (adapted from Wright; 1955).

I I i ,I g I I . ·...... i- l .. r" .. -_ :- -- - ..;-

·I ...• • • ' . ·• I • .I • : ...... _.-:-:...... __ ~ ...... __ ..J . i...... i ~ ~ .. :. . ,...... ;. : : J...... ; ..... ,;. ... t ... __ ,

vI ..

Fig. 4. Major presettiement floristic provinces in Minnesota and location i' of I., propuHans

I I I I --'..T...... · -r" --.- ···r- ...... ·.... -:-_ ... _-- ...... ~..

I I I • t I I • t • I I , • I ;-······l~}--··;· r····· t-··_- , . ... _-; : : :...... _....·.. t. 0 __

pOOMPJOH - JClHUO) PClX!W ISClJ0::J pOOMSsog - ClldoW

...... puol>tJOd puo YOUUOAOS. -~ ~.- Cl!J!OJdD 17

in Rice and Goodhue Counti.es. is. shown "tn Fig. 5; prairie angQak $C.\YClnnah compris.e most of the remaining vegetation. The "Big Woodslltype vegetation that exists. today is evident only in Nerstrand State Park, Rice County, and along ravines and riVer bottoms. The latter is typically too steeply sloped todevelop for agricultural or coll1ll1eY'cialpUY'poses.~=::=:===:

tty Descriptions The des.criptions of the five sites utilized tn this. study are indi­

viduallydiscussed~ Site I Site I (Jig. 6) was located in Goodhue County? approximately two miles west of Kenyon (T109N, R18W, Sec. 8) in anopen,·djsturbed~woods.This

#. site lies directly adjacent to the North and has a s ope of less than 50 from the midstream bed. The soil is loamy, and leaf litter, although not present everywhere, often formed a layer of2 cm or more. This site was disturbed by three factors; grazing by farm animals, nearby railroad construction, and the loss of canopy cover and leaf litter by the death of elm trees. The latter might be the most significant dis­ turbance as the canopy cover and leaf litter strongly influence the en­ vironment at ground level. The vegetation of Site I is becoming dominated by grasses, Urtica dioica, and Laportea canadensis. Only four small clones of I. propullans, each occupying no greater than 20 square meters, were observed at this site in 1977 and 1978. ·····~======cc==~ccc ...... • · •..... The ground-cover species associated with I. Rropullans and I. albidum at Site I were Anemonel1a thallictroides, Asarumcanadense, Cla.ytonia virginica., acutiloba, Trillium nivale, and Viola spp. The pre­ dominate species at ground level was . It formed several II type Ivegetation and! distribution of I. propullans populations I'in Rice and (adapted from Marschner, 1930) •

... ,

_ Big Woods vegetation

* do~umentented pulations E.t:!' Vt-!VIIU"~

~, II Cannon Fall s ~~ ~ # aUI~· .. .Ne'r nd Zumbrota.*

Rice Go~dhue County ,

...... \.0 Fig. 6. Location of study s~tes in Rice and Goodhue Counties, Minnesota.

I I

Im ;

_...~...._._.__ ._--_._-_....IIIIIIIIIIIIIIIIII _ .J ··_··:·--f···~·············:-··········/:• .. l"l ...... ; ..... ,...I:: f~~ : . . .: ,; . ~ / . a:- f• • .-ft ...I .• i..~ .I .: .. 1 • ..}

_...... -... - ... _- _... ----­

-IZ 12,

dense clones thqt almost made a continuouS. rn~t from the river bank to the margin of the woods, approximately 20 meters.. Woody species at this site included Acernegundo?Acersaccharllrn, Crataegus sp., Ostrya v;rginica,Prunus serot;na,Prunusvirginica, Quercus alba, Quercus macrocarpa,QuercuspalustrJs,IiJJa ailler;<:ana,Ulmus amer­ icana (standing dead), and Xanthoxyl urn arner; canum.

; --23

Site III Site UL (fjg. 6) was; located in Faribault~RiceCount¥~ in a woods. 0.5 miles north of the Wilson Center School .5 miles north of Site II , on the north facing slope of a ravine draining into the Cannon River (TlION, R20W, Sec. 20). The slope from the midstr~alTll!~g\\'().§!§.§~~~~~=IbE:!§Qtl was loamy and. well drained with an abundant covering of leaf litter. This particular woods was the least disturbed ofalLsites. bidum clones were scattered on the north facing slope .. and toward the per- --- . iphery of the woods (an alfalfa field enclosed much of the wooded area). On the north facing slope, I. propullans was more abundant than _. albidum. The associated ground cover species included Anemonella thallictroides,

Asarum canadense, Claytonia virginica, DentarJa~lacin-i-a~ta~,-D~i-centl"a canadensis, cucullaria, Hepatica acutiloba, canaden- , sis, and Viola spp. Woody species within this site were Acer including many seedlings, Fraxinus americana, Fraxinus pennsylvanica var. subintegerrima, Prunus virginiana, Tilia americana, Ulmus americana, and Ulmus rubra. Site IV Site IV (Fig. 6) was located in Faribault, Rice County, in a rich woods behind St. MarY's School (type 10cality>ofLpropuHans), 1 mile south of Site II (TllON, R20W, Sec. 31). The slope was 240 from the mid- stream bed and soil was loamy, well drained and covered with leaf litter. The stream below the ravine drained into the ight _River. Although I.

.., .'" ... " ~. __...". albidum grew abundantly on north and wouth facing slopesLonly one small clone (approximately one square meter in area).ofE.propullans was found on the north facing slope. li.

The s,ame woody and herbaceous species as those found at Site II were observed here. The accumulation of human refuse was conspicuous at this site. Site V Site V (Fig. 6) was located 4 miles north of the Fari lt ty 1imits in a di sturbed woods adjacent to the Cannon 0 River (TllON, R20W, Sec. 20). The slope was 20 from the midstream bed. This site was unique in that a small clone.. composed of approximately 25 individuals was observed to be intermediate in size between £. propullans

and E. albidum. One large clone of £. propullans, approximately 3.5x8.5 meters, was present with several I. a1bi dum plants ; nterspers~dth~oughout the

" clone. Although I. propullans grew abundantly on the slope, theY were ab­ ,sent on the adjacent flood plain area grazed by cattle. A few I. albidum Were found on the flood p1ain, however. The~S"ame-woodY~alld-'herbaceous

spedesas thosefoundatSite II and IV were observed at 5i te V. 25

METHODS AND MATERIALS

Morpn()]ogicalcharacteristics of E.albidumiandE.iiPY'O[)l1ll ariswere analyzed by measUring live flowering plants and by inspectingherba.rium specimens to numerically define and compare the 'tWQspecjes.Theiherharium specimens which were examined are listed in Appendix I. The following five characters were measured: (1) flower length, (2) length, (3) length and width ofthe larger leaf of thef1()\'I~!"i,!g_~p]_~!:I:t~J4tleafJ~n~~b... ped­ uncle length ratio and (5) distance from the peduncleapexiitothepoint of runner origin o~·the stem (in £. propullans). Approximately 200 £. pro­ pUllansand £. albidum plants from Sites I, II, III, and V were measured. A Student's t-test was performed for each character to distinguish signif­ icant differences between the taxa. Non-flowering plants were not measured because of the difficulty encountered in ass gningthese ... toeither species. Single-leaved forms of £. propullans and £. albidum are similar in shape and are difficult to distinguish.

Cytology Chromosome preparations were obtained according to the method out- 1ined by 1965. Acti ve ly growi ng runner tips of £. I?ropull ans were gathered in the field and treated. for four hoUl~s in a 1:1 mixture of 0.003 M aqueous 8-hydroxyquinoline and saturated aqueous solution of alpha-bromo­ naphthalene to obtain straight and well spread chromos.omes. Runner tips 26

were stained in a 9:1 mixture of 1% aceto-orecin and IN HCl heated oyer an alcohol burner, and then squashed. Preparations were microscopically examined at 1000x using a Leitz Dialux photomicroscope. Photographs were taken using Kodak Panatomic X film. Chromos'ome counts from mi crosp()Y'~rnothercells were not attempted . ---"--'---'--'-'--"'--.-"'-'-"'-""""'''''-'--''.. "".~,._.~ ..._. ---" ----- Mi crosporogenes; soccurs within the subterranean bulb of I .. propu 11ans in the fall; and only. in those bulbs thataredestinedt()becoll1eflowering individuals. The uncertainty in the timing andchoic;eofbulos would re­ sult intoo many bulbs being destroyed to justify their use in this method.

Demographics and Phenology

Clones of E. propullans at eachstudy~.sjtEt.w~.rg3ssessedfor three factors: (11 area ofclone, (2) nWnber of non"'floweringplants, and (3) number of flowering plants per 0.25 m2. Phenological data on flowering and insect activity \'Jere recorded, particularly relationship between ambient temperature and the daily opening and closing of the flowers, the presence of insects, and the annual duration of the flowering interval.

Pollen and Ovule Production Eighteen flowers of I. albidum and 26 flowers of I. propullans with undehisced anthers were gathered among the sites where breeding exper­ iments were conducted to determine pollen and ovule production in both species. A total pollen count was obtained by rupturing one ~n·~hc.~ each flower in warmed glycerine jelly to release the pollen, and by staining them with methylene green (Radford, Dickison, Massey and Bell, 1974). The number of ovules per flower was also recorded. 27

A staining procedure was used to determil1ElpQI1Ell1yjgl:>iJjty (actual germination of poll en grains was not performed). One anther from each of 37 I. propu11 ans and 21 I. a1bi dum flowers between 1977 arid 1978 were collected from among all sites where breeding experiments> were conducted.

The pollen grains of each anther werE!stCljllegwjth=1%~antL1DJ:Ll:>lYe~jJI lactopheno1 (Swanson and Sohmer, 1975), with a minirllum of 500 grains

scoredperplant~ Pollen was recorded as stainablep61Ten,unstaJnabl e pollen, or micropo11en. ..

Pollen-Ovule Ratios The pollen-ovule ratios of I. albidum and E. propullans were deter­

.mined by mu1 tip1ying the total numberofpo]Jengratns_pE1'~antherbythe number of anthers per flower, and bya'iv'idingthi-s valaeby the number of ,. ovules per flower.

Vectors of Pollen Transfer The search for poll inators was conducted throughout the fl oweri ng phenophase of Erythronium in 1977 and 1978. All of the types of insects that visited Eo propu11ans and I. albidum between the hours of 9:00 am and 4:00 pm were collected. After preliminary sorfil1g,speGimenswere sent to experts for specific determinations of verifications. Extensive observations were recorded on pollinator activity and' flower morphol in relation to pollination.

Breeding Studies Pollination experiments were conducted to determine the propensity of I. propu1lans and I. albidum for apomixis, self-compatibility, and 28

hybridization. Sites I, IT, and IV in 197TandSitesI;II, and III in 1978 were utilized for this purpose (Site IV was not used in 1978 because of the scarcity of I. propullans at this site).Tl1eHitionaieanddesign for each treatment is described in each of the following subsections. Treatment I: Propensity for Apomictic Reproduction In both species, agamospermy (production of seed withoutipollen) was to be detected by removing all stamens (undehisced)priortoanthesis. Any possibility for chance cross pollination was assl.lredbyenclosing the emasculated flowers within glassine bags. Seeds prbducedUnder these conditions would establish the level of apomictic reproduction. In 1977 and 1978, 39 I. 'p!opullans and 40 I. albidum flower buds representing all sites utilized were treated this way. 'Treatment II Propensity for Self-fertilization p. Selected flower buds of I. propullans and I. albidumwereenclosed in glass1ne bags prior to anthesis. The pollen available to that flower, therefore, belonged only to that plant. In 1977, no further treatment was given. In 1978, the bags were removed duringanthesis, and the anthers wererilbbedagail'1stthe stigmatic surface of thepiSti1 either before or after the stigmatic surface of the pistil had Tullyexpal1ded. The bags were then replaced. The production of seeds bytheseplantsfn1977 would demonstrate the level of self-fertilization evident in plants not visited by pollinators. The production of seeds by these plants in 1978 would demonstrate the level of self-compatibility in the species. In 1977 and 1978,4Jt.prOpullal1s and 48 I. a1biduiTI;f·epfesen'tfng ciTT sites Utilized , were treated this way. Treatment III: Crosses Within Clones Prior to anthesis, all stamens were removed from flowers of I. pro- 29

,

Tt'eatmelitV:· I1iterpopUlati ona1 eros ses FloWerswereemasC:Ulated and enclosed prjortoanthesis.Whenthe stigma was fully expanded; pollen of the same species from plants that occurred at another site was introduced onto the>stigma. Theglassine bags were then replaced. Seed production in this experimental>group when contrasted with Treatment IV would indicate the level of seed production that a maximally outcrossing system could produce. In 1977 and 1978, 31 I. propullans and 43 I. albidum flowers were treated this way. Treatmen'ecvrT Crbsses Hel:weell E. propull ans-antl -E~ a fbTClum==== Flowers of both species were emasculated andericl()sed prior to anthesis. Pol1enfromI.propullans was introduced onto the expanded I. albidum stigma; pollen of I. albidum was introduced on~o the I. propullans 30

stigma. Seed production within this group would suggest the extent to which gene flow might be possible between the two species. In 1977 and 1978, a total of 49 S. propullans and 30 E. ,albidum were il1c:llidedil1 this experimental group. Treatment VII: Natural Level of Seed Production A total of 52 E. propullans and 43 I. albidum plants, representing 1 sites utilized in the breeding studies in 1977 and 1978, were selected ,and marked. This control group established normal levels.of... s.eed.pro,.. duction in each species. The annual and I. a1bi dufuwa.Sdeterlllined by dividing thetota1 number of. seeds ob­ tai ned from ea.chspecies b.y the numberOfflOweringplal1fsselected wi thi n this control\group. Reciprocal¢l"'()sses between the pollendonq,tingclO!1c{an(It.he pollen " recipient clonew.ereperformed for Treatments J II,. JV/,V,andVI. Sevel~a1 individuals from different areas of a site were used as w>llendonors to obtain a. Trueiridicafioriof the popula.ffolf asa-wnole.--'Ha\1(r]5(jTTination ,was··carrfed'Ql.Ifon1Y\>ihel1fhe stigma' had'flillYexpanded; approXimately three to fiv(Ftlpysfollowinganthesis of theflowerbud.Whehfully ex­ panded, the stigmatic surface was well exposed and Was covered with many minute hairs which facilitated lodging of pollen: To facilitate retrieval of the fruit, the glassine bags were stapled to tongue depressors which were secured in the ground (Fig. 7). Fruits were harvested five weeks jlfter hand pollination and were allowed to sit for five days until the

ca.ps lJlesarieaa.l1dsprit~=rel easi ng'the seeds-fO't'COllhtlng:-At=thetime· of hal~vest, the peduncle and fruit were prostrate on the ground (un 1ess supported by a tongue depressor) with the peduncle unattached from the remainder of the plant. Although the fruit and most of the peduncle were If: 32

green at the time of harvest, the fruit had to be r~trjgyed at that time

or lost to decay.

, 33

RESULTS

Morphology The range of variation in the five characters measuredf()r Ealbidum is large within ea.ch site (Ta.ble 2). Va.riability in E.al\bidummorethan likely represents the ecotypic differences between individuals. For ex­ ample, at Site I where the length of the flower ranged from 1.5 to 3.2 cm, sma 1.1 er flowers were observed growi n9 in grassy, matted areas. The range of I. propullans measurements was considerable at Sites II and III (Table 3). Nany of these f.>' pr()pullans plants had measurements , approaching that ()btained from the intermediate size class (Table 4) and larger than• the population means or ranges of I. propullansI at other sites

(Sites I and IV). These plants may represent the hybridization of an intermediate or hybrid class and I. propullans. At all sites, significant size differences betweenL propullans and

E. albidum were evident (Table 4). The ave~age length of the I·albidum flower, 2.78 cm, was more than twice the averagelengthof.the E.pro­ pullans flov-ter. The range in flower length did not overlap between both species. Size differerences were also apparent in the mean length of the peduncle, 6.48 cm in I. propullans and 12.25 cm in I. albidum, and in the mean length and vJidth of the leaf, 9.06 cm x 2.02 cmin--E~propufYans as compared to 12.25 em x 2.3cm in £. 1 dum. As the mean leaf length to peduncle length ratio was 1.5:1 in £. propullans, the height of the flower is consistently shorter than the height of the leaf. In £. albidum, the "

Table 2. Compa.rison among sit~s of iSm6rpholbgieal characters analyzed in t. albidum•.

.flower Jength .pedul1el e 1ength leaf length 1eaf: pedunel e leaf width

Site I I mean " 2.76 em 12.95 em 13.25 em 1.01:1 2.33 em sample slize 20 20 20 20 20 standard deviation .420 1.100 1.160 .130 .340 range 1.5-3.2 em 11.3-15.2 em 10.5-17.8 em .82-1. 25:1 1.3-2.8 em

Site II mean 3.1 em 11.1 em 11.8 em 1.07:1 2.12 em samp1e s'r ze 11 11 11 11 11 standar~ deviation 1.030 1.180 1.150 .060 .380 range 2.6.-3.2 em ~.5-13.6 em 10-14.2 em .97-1.18 em 1.8-2.4 em

S te III I meary • i 2.48 em 112.53~em 12 em .96:1 2.11 em sample slrze 29 I 29 29 29 29 standar~ deviation .260 I 1.080 1.080 .070 .230 range 1.6-2.9 em 110.2-15 em 9.8-14.2 em .82-1.1:1 1.7-2.5 em

1 Site V mean 3.13 em 111.34 em 11. Sem .01:1 2.79 em sample sl~ze 15 , 15 15 15 15 standard! deviation .320 1.40 1.120 .090 .470 range 2.7-3.6 em 8.8-13.4 em 9.4-13.3 em 1.3:1 2-3.5 em w ~ ,

Tiable3.Comparison among sitesqf' 5 l'llorpfi01ogicalc:haracters analyzed in •.s. propul1ans.

flower length pedunqle length leaf length 1eaf: pedunCl e leaf width distancel

Site I mean i 1.1 em 5".38 em 9.83 em 1.86:1 1.6.6 em sample size ,I 16 16 16 16 16 standard devia~ion .083 .770 .670 .090 .070 range 1-1.2 em 4-6.5 em 7-11.1 em 1. 37-2.52:1 1.2-2.3 em

Site II mean 1.19 em 6.32 em 10.19 em 1.46:1 2.06 em 12.21 em sample size 40 40 40 40 40 40 standard deviat~on .210 1.530 1.120 .200 .270 1.680 range .9-1.5 em 5-10.8 em 8-13 em .8-1.89:1 1.4-2.8 em 9.6·H6.5 em

i Site I II :I mean 1.4 em 6.97 em 8.73 1, em 1.25:1 2.29 em 11.34 em sample'size ; 30 30 I 30: 30 30 30 " standatd devia~.·.lion .100 .830 \ .850 .090 .200 " 2.920 range i !,' 1.3-1.5 em 5.5-8.8 9m 7-10.9 em 1.06-1. 42:1 1.9-2.5 em 8.1-14.4 em I ! \1 I .. Site V i.11 ! I i mean . .96crn 5.08 em 8.36!em 1.66:1 1.~8 em 8~94 em sample isize II 20 20 I 20i 20 20 20 standard deviat~on .070 .680 ! .790 .180 .160 .690 range .8-1.1 cm 3.9-6.2 cm 6.7-9.5 em 1. 32- 2. 11: 1 1.4-:2 em 5- 1.4 em

w 1distanee from orligin stem offshoot to the apex of the peduncle U1 ,

Table 4. Comparison of morphological characters between I. propUllans t the intermediate size class and i. albidum.

flower length ped~ncl el leri'9th leaf length, leaf:peduncle leaf ~idth distance1

I. Eropul1ans mean 1.19 cm 6.48 cm 9.06 cm 1.5: 1 2.02cm 11.24 !cm sample size 106 106 106 106 106 . 90 standard deviat10n .214 1.462 1.626 .• 281 .. 336 2.117

! range .8-1.5 cm 3.9-10.6 cm 4-13 cm 1.06-2.52 1.2-2.8 cm 6-14:4 em

Intermediate size class

mean 1.9 cm 9.78 cm 11. 74 cm 1.2:1 11 cm 13.:31 cm sample size 12 12 12 12 12 12 standard deviation .141 . .81p 1.1682 .152 343 1.458 •• I range 1.6-2.1 cm 8,6-11[3 ~m 9.3-~3.9 em :1 1.7-2~9 cm 10.8-14.8

E. albidtim I mean 2.78 cm ~2.251 cm 12.22 em 1 2.3~ cm sample size 85 8S! J5 8b I i standard deviation .358 1.485 U423 .422 range 1.5-3.6 em 8.8-15.2 cm 4-17.8 cm 3:1 1.7-~.em \~ 1distance from the origin of the~tem offshoot to the a!pex of the peduncle 37

two are approximately of equal heights. Based on these measurements, the two taxa are quite distinct as to the relative size oftheir above ground parts, and may be readily distinguished without noting the presence or absence of 'a lateral offshoot. Three size classes of plants were evident from __th~e:~_numerical analysis of each of the

smallest intermediate size class encountered at individuals) was found to be significantly different and I· .9l- bidum size 'classes in each of the four . 8 and 9). The interme'diate siz!2! class has a mean flower length of 1.9 cm, a mean pe-

" ".;' .." , .", duncle length of 9.78 cm, a mean leaf length - peduncle length ratio of 1. 2: 1, aliCia mean distance of 13.31 cm from origin on the stem tO,the apex of the peduncle. ate size class with a stem offshoot (Fig ng that of E.-a:Tbfdl.fm--fSClfffi cult to or I. propu 11 ans) . The fl O\"ieri ng to form fruit and produce normal seeds each of the three~sfze classes was white. _ pro- pullans flow~rs as, described by Gray (1871) in this ) study.

Cytology All chromosome counts taken from mitotically dividing cells within the runner tip of I. propullans were in the low 40's, suggesting that 38

,.

cfigs.§;:~~:~=['Lg.~,=?:~n E.:·ErQPylJ ans,jDtE?rm; d; ate. sized ,_a_n(:L .. E. a1- bidum flowering individual. Fig. 9, the flower of I.a1bidum TtoPT the intermediate size class, and E. propu11ans. 39

,

11 Figs. 10-11. Fig. 10, the lateral stem offshoots of plants of the in~ terl11eala'fe.:~slze=crass~='=Fig:·· 11, the' triA itof······E:···aT51d-iAmTlern~·········the i ntermedfa.tesize .cTass ,arid I. propu 11 ans (right). 40

f . .Er.Q£ullans, like f. albidum, is a tetraplQicl $Pec:ie$. An exact chromo­ some number was not obtained in any of the preparations due to artifacts and the clumping of the chromosomes. The best chromosome preparation is illustrated in Fig. 12.

Demographics and Phenology ConHguousclones of I. a1bi dum formextens i ve (Fi g. 13) .••.... thlSphenollleno!1 was not

The largest clone of f. propull ans nnC:AY'\ftlfi located at Site 5 (Table 5). In general. E to I. albidum) tends to occu!ras relative'ly small, of clones. The densjty of plantsihI. prbpuTTal1sclOl'lesral1gedfr01ll38·1b 212 , plants per 0.25 m2 (Table 5). The percent Of flowering plants as compared to non-flowering plants ranged from 1.1 to 22;9%(2 to 32 flowering plants per 0.25 m2). Th'is density reflects the manner and mode of vegetati~e reproduction. / Each fertile rloweringplant new bulbs

each year, and the sterile, non-flowering plants nY'"r1I11"''' new bulbs each year. Therefore, high population densities of young sterile plants sould be expected. The normal blooming interval of Erythronium in southern Minnesota is between the last two weeks of April and the first two weeks of May. In 1977, the first Erythronium blossoms opened the 13th of April, and in 1978, the first flowers began blooming on the 20th of April. In 1976, Minnesota recorded a severe summer and wi nter drought which c1i rn iYli shed the number of I. propullans individuals in many populations the following 1977 spring season. In the spring of 1978, .I. propullans was more abundant and new •SUl? Undo.Ad •3 .:1-0 sawOSOwO.AlIJ •~I . 6~.:t 42

,

Fig. 13 .. A typical population of E. albidum showinq its dominance as a ground cover species in early spring. Table 5. Number of floweri andnon-f16weririg individuals of~. ~ropullans a area. with the appl'"ox- i:nate area ea'ch respect; ve one cccupi~~ .. .-

Site: ...T I II II II III IV V nUi:Jt)er of ncnl-flowering individuals 35 82 125 70 125 168 104 180 numbtr Of f11,erin g. individuals 3 2 15 16 13 28 6 32

%f10wering individuals 8~6~'· 2.4% p"..7. ! 22.9% 10.4%' 16.8% 5.8% 17.8% I area of clon' (square meters 11.88 17.2 1.3~ 6.48 290.5' 20.35 393.5'

~ w 44 populations were located at this time. The daily opening and closing of flowers corresponded with diurnal temperature fluctuations during favorable weather conditions. In 1978, th~ .[. propullans and E. albidum petals did not assume their fully re­ flexed positions until the afternoon hours (between 12:30 pm and 1:30 pm .<~."-."'-'~-'-"------or later if on a north facing slope) when the ambient temperature was at least 120 C. The optimum interval of insect activity (including pollinators) also began after 12 :30 to 1:30 pm. During inclement weather conditions (rain or low ent temperatures), the owers rema.ined closed and pol- linator activity ceased.

Pollen and Ovule Production...... _--_ . PoHen

Pollen-Ovule Ratios The pollen-ovule ratio of both species is similar although pollen pro­ ductivity is much higher in f. albidum. A mean ratio of 767:1 in £. pro- 45

Jable6. Pollen~ndovuleproductionand len-ovule E. albidum.

pollen/anther po 11 en/flower ovules/flower p:o

1. .P.r.Q.2U 11 ans mean 953 4,194 5.32 767:1 sample size 26 26 26 26 standard deviation 370 .2,310 1.17 261.3 " range 467-1,872 1,862-11,232 2-8 374-1,310:1

mean 5.596 33,575 35.9 965:1 sample size 18 18 standard deviation 1,339 8.037 7.7 176.5 range 2,205-7,836 J.3,230-47,OI6 24 .... 43 427 .. 1,200:1 46

Table 7. Percent stainable, unstainable and mi len al-

mean . sample size standard deviation range 47

pullans and 966:1 in I. albidum was obtained. The range of pollen-ovule ratios between individuals in both species was broad (Table 6). Based upon a pollen-ovule ratio index CCruden, 1976; 1977), the breeding system of E. propullans and E. albidum is predicted as facul- tative xenogamy, i.e. ,each spedes is se1f-compati-ble~~_and~-adapted for out- , ,_.', , _, , _" , "..: _,_."." .. :: :. .. ,,:. : '_ :: .. ':.. "_.•.._•• _.._., __.".~ ., .. _._ .. _._. __. ~. __._'"."__•.~_ _.,, __•••,,. ·_,,·_"_,,_< __~,~,~.'=~=.H_"""==.= _ _,_".", .." _,~ ," _".". , crossing 796:1

error of

Observations of Floral Biology The f1 ower of E. a1bi dum cons ists ()_L'tWl:LW!1QrJ~_QtJ)eta1s, three , inner and, three outer, wi th one (Fig. 141~' ", At full anthesis of the flower, the ~ -'--', pendent,' bell-shaped flower (Figs. 15 and 16). The flowers reflex, as

, previouslY mentioned, at approximately increases, and remain reflexed until in conjun.ction with loss of sunl single day· when climatic conditions C'; f0l" pollination by bees or other di

/' imately six hours. E. tpropu 11 ans flowers i kewi se cons ist of two whor1s of peta1s. Each whorl, however, is composed of either two or three petals, so that the flowers are either 4, 5, or 6-merous (Figs. 17-19). The stamens, which are attached opposite to the petals, are equal in number to the number of ~etals. In any I. propullans clone, any combination of 4, 5, or 6-merous flowers' exists. The timing of the opening and closing of the I. propullans 48

Flgs:'=T~=r5~:"=Frg:Tlf';T=C=propull ans ( 1.eftrahd···E~······a:l-bicrl1mn···floweFs ····sh6wi ng the inn~rancl9~ter ~horl of petals and stamens. Fi9.15, stamen di­ morphism in E. albidum, the shorter whorl dehiscing prior to the longer innerwhorl.- Fig. ·16, stamen dimorphisminLpropullans. 49

Figs. 17-19. Fig. 17, I. propul1ans flower showing 4-merous condition. Fig. 18,I~ propullans flower showing 5"'l11erbUS cbndition. Fig. 19, I. propu 11 ans flower. showi ng 6-merous cond i t;QI1._ . 50

flower is identical with that of E. albidum. Stamen. dimorphism occurs in both E.propullans and E. albidum; the

stamens attached to the outer whorl of petals are shorter than those Figs. 15 and 16). As the flower opens for the

fi pollen is

anther. one day after dehisence of the outer \'Jhorl. Although pollen may be ob-

served on the anthers for three days following initial dehisence, a sudden

spring shower was observed to wash all pollen from an anther in a matter of minutes. The stigmas of both I.. albidum and E. propullans spread three to five days ,after the shorter whorl of anthers--dehi-£ee.-I-R-~-. a-l bi dum,

the stigma spreads into three distinct lobes, the

stigmatic :surface merely expands to expose (the

lobi ng of the s ti gmas~'J1Aent in E. !!.~~Lj:~~~~~~~~_ in I.. . pullans). E~ch species is therefore

fore the stigmatic surface spreads

Fl ora1 nectaries were observed

I to taste~ Altho~gh nectar was not detectablemi~_~~~B~~_~llans,the nectar~ gathering behav-ior--of-manybees on I.. propullans flowers suggested that

nectar was present.

Observati on Pollinator Biol Iheemergence of many miner bees, bumble bees and other insects was

sy~chronous with the opening of Erythronium flmvers. The majority of visits to Erythronium blossoms were made by the miner bee, Andrena carlini Ckll. (Fig. 20), although one bumble bee (Bombus bimaculatus Cresson) was observed '. 51

Figs. 20-21. Fig. 20, Andrena carlini Ckll. Fig. 21, Andrena carlini enteringanL·albidum blossom. 52

to visit an I. albidum flower. Andrenacarlini were observed to collect nectar and pollen and mate on both I~albidumand E. prOpJllJans flowers. The nectar gathering behavior - bidL!m and I. propullans flowers the rear, flexed ding' onto the an1:he'rswi ith the bee

gan to systematically collect nectar the entire pi~s,ti 1. Whi le gathering nectar;ft~plfsh-el:i-th~nth-ers-away from its body and in contact with the stigma. After emerging from the ...... ' '.' .' ', .•..... ," " ' , , ,." .., , " " , ,., ,.', ,...• _...•...... , ...... • E- a1bi dum flower, aeonsi derab1e amount of pollenwasbbse the ventral surface of the bee. bE:!gs emerging from E. p\~opull ans flowers. Several A. __ ...... visit E. albidumflower~to gather_pollen only. In this case, the bee oriented itself away from the base of the flower and clung onto the anthers and

filaments~.' with its front two pair of legs,_---_.,------~-"'-,,."",-,--",------freeing its two rear legs to . .' , " .. _._.,.,.' , ,., ,.. ,.. -.., " .."." ", - ',.' .. .' .. ' . .' .' .' .' .. _.. _-_ _., _, ,',•.. _.. .' collect pollen. Four--pol+enpackets (two on each of the rear legs) were observed on the pollen-gathering bees (Fig. 24).

Up to tenA~>carlini were observed collecting nectar \'lithin a single

ClOneQIcJ:.jl]bidumorL propull ansL~Cicb\liSi~t()~Clc;~fJ()\'lerJastedfrom five to t\'lenty-five seconds in I . .PIopullansand up to 30 seconds in f.. albidum. ShO\'ling consistency to either species, a single bee visited up to fiv~ flowers \'lithin a clone before moving on to another clone.

Where clones of I. ~llans and I. albidum gre\'l adjacent to each 'swossoLQ SUgLLndoJd '3 WOJJ Jg+~au 6u~+~aLLO~ ~U~LJg~ euaJpUV '£l-ll 's6~j

£5 54

Figs:=2tj.:::;C2~5C:=~Flg:C='2Zj:~frncrrena Cel rlTriiw-i tn- polTen- sacs on rear legs. Fig. 25, Bombyliusmajor collecting nectar from I. albidum. 55

Breeding Studies

The==~esutts==f~QIlL=eA=(;l1ccpol] j nationtrgCl.1:[Ignt,~~il~~ out] i n~JLjn the Methods and Materials section, is discussed separately. Treatment I: Propensity for Apomictic Reproduction Neither the I. propullans nor I. albidum flowers that were emasculated in 1977 or 1978 produced fruit of seed (Tables 8 and 10). Apomictic repro- Table $. Comparison lof data between the sites utilized 1977 and 19.78 Imeasuripg~he Occurance of apomixis "in I. Ipropullansand I. albidum.

· Site I Site II ; Site III Site IV Total

I~-~ ---~ ---,--~~~------,------i -I. propullans - 1977 number bagged/number retrieved 15/15 7/7 22/22 %wHh fruit 0% • 0% 0% I· ro: ullans - 1978 • num! er bagged/number l"'etri eved 6/6 6/6 5/5 17/17 I %w!Hh fruit 0% 0% 0% 0% I.. alb! dum - , I num: er bagged/nlJmb~rretri~ved I 14/14 1/1 15/15 % Wil th frui t ':0% 0% 0% I. alb~dum - 1978 I numper bagged/number retrieved 110/10 10/8 5/5 25/23 %with fruit 0% 0% 0% 0%

(J1 ~ 57

duction, if it occurs at all, must occur at an incidence so rare that it could not be detected with the sample sizes employed in this study. These results suggest that pollen is required to facilitate seed set.

.Treatment 11: Propensity for Self-fertilization

No fruit or subsequent seed deve1 opment()ESUY'l"~d in~h~=§.propu l~ans f18wers that ..wereise1f..pol1inated in 1977 or 1978 (Tables 9· and 10). In E';a1bidum, no fruit or seed was produced by plants that were trea.ted for self-fertilization in 1977. In 1978, however, 50% of those flowers treated for self-fertilization produced fruit, wtth an average of 2.33 seeds per . fruit (T~b1e 9). Pollen was not manually rubbed on the stigma in 1977 as it was in 1978. The two year data suggests that without physical assistance, pollen cannot be transferred to the stigJTIa,.ELl'('~th~l"~~E!~,--d_set in ~. ll­ bidum was highest when plants were pollinated shortly afterthedehisence of the anther and prior to the spreading of the stigmas, as they were at Site IlL At Sites I and II, where seed production was low.er, the flowers

seed when self-pollinated. If ~.proQullansisse1f.. compatible, it is at an incidence so rare tha,t it could not be demonstrated with the sample sizes employed. ,Treatment III: Crosses Within Clones In 1977 and 1978, no fruit or seed in ~. propullans .plants crossed with pollen from flowers of the same clone (Tables 11 and 12). ~. a1bidum flowers crossed with a flower of the same clone in 1977 and 1978 resulted in 40% of all plants producing fruit and a mean of 4.2 Tabll¢ 9. Comparisor of data be~ween !the sites. utilized in 1977 and 1978 measuhng the occurrance of self~ferti1ization lin i. propu'nansand i. alb~dum. .

Site I Site II Site III Site IV Total , i. propul1ans - 19717 I n~mber bagged/number retrieved 7/7 15/15 22/22 I i %1 with fruit 0% 0% 0% f. propullans - 1978

nllmhpY' bagged/number retrieved 5/5 9/9 6/6 0% 0%

E. _1'_' __ '" - ;

-E. 11'-'--'",

1-4

()l 00 ·lab1e lQ~ . Su~mary qf re~uH~ frqm treatmentstodetermin~ aP9111i.x1 ~Iand sel f­ ~<:rtilizationl\'Iithin I~ prop~l1ans and I~ albidum in 1977 land 19178.

apomi xi s self-fertilization

E~ propullans number bagged/number retrieved 39/39 43/43 %with fruit 0%

albidum number

()"I \.0 Table n.,' Cornparison'ofdatabetween: the ~ites utilized in 1977 and 1978 measuring,rthe success of intra- clonal crosses in I. propullans' and I. albidum. .

" Site I Site II Site ItI Si te IV Total r. propul~ans - 1977 ii ]! numb~r bagged/number retrieved 3/3 10/10 13/13 ;1 %wil~h fruit 0% 0% 0% 'I r. propul~ans - 1978 number bagged/number retrieved 5/4 6/6 5/5 16/15 %with fruit 0% 0% .0% 0%

, E'••lbid1- 1977 - numb'r bagged/numberr~trieved 16/16 16/16 %wi~h fruit 31% 31% mean number of seeds/fnuit 3.6 3.6 rang~ of seeds/fruit 1-7 1-7 '. EI. albidu~ - 1978 II - nu...nb~r bagged/numberretri eved 5/2 5/5 !f/9 %with fruit 0% 60% ~6%

mean I number of seeds/fruit 4.7 8 6 , range of seeds/fruit 2-8 4-12 2~12 o0-. e 12. SummarY of results from.treatments to determine success crosses in E. propuillans and I. albidum in 1977 a,nd

E. propull ans I. _. - ,__n,. nu. ber bagged/number retrieved 29/29 % . ith fruit 0% me: n number of seeds/fruit raj ge of seeds/fruit ! I

.....m 62

seeds per fruit (Tables 11 and 12). These results indicate that intra­ clonalhybridizaUon in I.propullans is.notlikelytoproduce seeds. In

E. albidum, the similar levels of seed production ofself~fertilized (Treatment II) andintraclonal hybridization reflects the genetic uniformity of all I. albidum plants within a TreatmentlV:iCrosses •••. Between Clones Within the Same Study ,Site In19ZZand19Z8inoseeds flower's that were crossec!with pollen site (Tables 13 and 14). During the bidum similarly treated produced fruit with a mean fruit (Tables 13 and 14). In E. albidum, these results in the results from Treatments II and III indicate tha'tt.bg,J)~e-r~~I1,t~ge of seed set in­ creases as the di stance of the poll en source increases ... · Interclonal hy­ bridization. in I. propullans is not likely to produce seeds~ Treatment V: Interpopulational Crosses No fruit or seeds were formed in the E. £!:2.Eu 11 ans crossed with ~ member from different sites in 1977 or 1978 (Tables 15 and 16). In £. ,albidum, 63% of the plants so treated in 1977 and 1978 formed fruit wi,th an average of 9 seeds per fruit (Tables 15 and 16) .•. The level of fruit and seed set in E. albidum for this treatment is similar. to that noted in Treatment IV. The range in the number of seeds produced from site to site was large, from a mean of 5.5 'seeds at Site II in 1978, to a mean of 18 ~eeds at Site I in 1978 (Table 15). These results indi~ate that in E. albidum, the highest levels of seed set are a result of outcrossing, and that a vector is required to transfer len from site to site. If £. pro- puilans is capable of setting seed from outcrossing, it is at levels so low that it was not detected. Table '13. Comparison of·data. between sites measuring the sucee.ss of inter-elona', crosses in .s. propu'lans and .s. albidum in 1977 and 1978. - .

Site I 'Site II Site I II Site IV Total·

.s. propul1ans - 1977 num~er ·bagged/number retrieved 8/8 3/3 11/11 %with fruit 0% 0% 0% I .s. propullans - 1978 number bagged/number retrieved 5/5 5/5 10/10 %with fruit 0% 0% 0% E~ albid~m - 1977 - num~er bagged/number retrieved 11/111 11/11 .' %with fruit 64% I 64%

me~ .. ~umb~~i~~ S~:~f/trult 8.1 I 8.1 ra~ eof seeds/fruit 1-13 I 1-13 ! I E. albid m 1978 I - nu~lber b?gged/numberretrfeved 5/5 10/10 ., 5/5 [ %w!ith fruit 0% 60% I 30% ! mean number of seeds/fruit 17.7 i 17.7 range of seeds/fruit 14-22: 14-22

O'l W Summary of re$ul tsl from treatments to determine' success bf i in E. propullans and I. albidumin 1977

E. propullans I. albidum

• eyed 21/21 21/21 0% 47% 12.5 3-23

0\ "J::::o T~ble,l~~, 'Comparison o~databetween th~ sites utilized in' 1977 and 1978 measuring' the su'ccess of inter- populatiipnal crosses in I. propul1ans and I. albidum.

Site I Site' II Site III Site IV Total

I. propullans - 1977 numper bagged/number retrieved 12/12 12/12 %with fruit 0% 0% ' i I. propul11 ans - 1978 number bagged/number retrieved 6/5 8/5 5/5 19/15 %with fruit 0% 0%, 0% 0% E. albidt - 1977 - num er bagged/number retrieved 8/8 16/16 %with fruit 62% 75% I I ,. me~ nnumber of se¢d$/fruit 6.4 f I I raf I 2..15' ge of seeds/fruit I 1":"24 I I I I I. albi urn - 1978 I nU~ber bagged/numb~rretrieved 5/5 27/22 ?/7I '! , %With frui t 14% 60~ 100% 54.5% I ! me~n number of seeds/fruit 18 5.5j 17.2 11.4 ! .. _ range of seeds/fruit ... 18 3-1[1 14-20 3-20 O'J c.n " table 16. Summary of .resu fr?m treatments to determine the success of ~nterpopulational crosses E. propullans and £. albidum in ·1977 and 1978.

E. propullans E. albidum

31/27 43y38 umber O~ !>withfruit..bag•... ge..d/...••..•....••... n.. u•...••m>..•....b. er,'...... •'·et..• r.ieved 63% ~.ean numbeY'ofsieedsYfru.i t 9 angeofseeds/firuit 1-24

O'l O'l 67

Treatment VI: Crosses Between E~ . prbpullansandE~.albidum In 1977, O%ofOtheI.propullanspIahOts crossedwittiE..a1bi dum produced seeds. In 1978, however, 20% of ~. propullansproducedfruitwith

a mean of 1.16 seeds per fruit (Tables 17 and 18). The discrepancy in resul ts i s probably due to the drought condi tionsof-1911--whi-c:h-made this

an atypi ca1 season and may have had an i nfl uence on the abicl i ty of ~. pro- pullans n E. "'" _. .. ,

1978 ~rossed with E. propul1ans pol seeds per fruit (Tables 17 and 18).

from interspecific hybridization between E. albidumand E. propullans dicate that gene flow between the

that seeds were set in crossed onto the stigma raises some questions

t i nctness, of these speci es .

81119ngthe£.. propullans plants selected

fruit with mean of 1.3 seeds per fruit in 1977 and 19 and 20). Of the E.albidum control plants sel

produced fruit averaging 5.8 seeds per fruit This .suggests· that in nature, less than one-fourth()falIE.propullansand abdut two-thirds of all E. albidum flowering plants produce seeds and fruit. The annual seed crop among all study sites averaged 0.3 seeds per flowering

pIant[l~1"ygarjnf. proRllll ans and four~e_e~s R~r_fJow~~in~L .. E.l...~.n.t. per year in E. albidum. Based upon the evidencee$-"tgbl f§hed in Treatments I-VI, seed production in .s. propullans in nature is probably a result of pollen transfer from~. albidum rather than from £.. propullans flowers. The

natural level of seed set in £.. albidum (5.8 seeds per fruit) is close to ,; :: ":;' i: < " :' , ',' ',' Table 17~ Comparison of data betweeh sit~s me~suri'ng the success of /interspecific hybHdization between ~. propuUlans and I. albidum in 1977: and ~978.

Site I Site II Site III Site IV Total

, I. propuilans - 1977 number bagged/number retrieved 15/15 15/15 %witn fruit 0% 0% I. propul1ans - 1978 number bagged/number retrieved 7/4 13/12 14/14 34/30 %with fruit 0% 8% 36% 20% mean ~umber of seeds/fruit 2 1 1.16 rangelof seeds/fruit 2 1 1-2

I. albid. m- numbe 8/S 8/8 %wit 75.~% 75.5% mean' . 6.1 6.1 rangefI 1-24 1-24 i I· 14/~ 22/14 66.61% 65.5% 5.5 10.25 3-11 3,..10 I

i 1

~ 00 II

! : . . " ',' .. ',' >.' ',::. J lTab1 e 18. S~mmary of resul ts from treatments to det~rmin~su~c~ssofiint¢r- ispecific hybridization between [. propullans and I. albiduminJ.977 and 1978.

[. .e!opull ans 1

I 49/4~ I.r".mher bagged./ n".,.... mber retr;eved 20%1 I~withfruiti .• '•• ·i : ' . I number.9f~eeds./fruit ean 1.16I angeof seed$/ifruit 1-21

I :[1[. propull ans 1978 data only. i

0\ \0 Table 19. il Comparison of ,data between sites measuring the natural level of seedprqduction in I. propullans· and I. a'~idum in 1977 and 1978.

Site I Site II Si te II S te IV Total

Ii I. propullians - 1977 numb~r bagged/number retrieved 7/7 10/10 17/17 %wi1h fruit. • I 0% 10% 5.8% meanilnumber of seed~/fruit 1 1 rang~ of seeds/fruit 1 1 .I. propu11 ans - 1978 number bagged/number retrieved 10/10 12/10 14/14 36/34 % with fruit 0% 30% 36% 17.6% mean., number of seeds/fruit 1.66 1 1.33 rang~ of seeds/fruit 1-3 1 1-3

E•• a 1ibidu~ -1977 - numb;r ~~gged/number!.r~tr.i eved 8/8 5/5 13/13 %wi\h fruit> ••••• ,[ 62.5% 80% 69.2% meani [number ofseedSVfrh.li t . 3.6 3.7 3.66 rang' Of seeds/ft'1ui~ . . 1-6 1-5 H6 I

E!. albidu' - 1978 .•.•••••••• f" ! - numb~r bagged/number i r~trie~ed 10110 10/9 3Pl29 %with fruit 70% i1 66.6% 68;8% mean.! number of· seeds/frUi t 9.6! 6.5 618% rang~ of seeds/fruit 5-121 2.-11 ~.LI2 I

!I !

'-So Table 20. Summary of results determining the natural level of fruit and seed set of I. propullans and E. albidum in 1977 and 1978, pl~.s the annual seed crop ofjeach species.

E. propullans E. albidum nu ber marked/number retrieved 53(53 I % I ith fruit 1p% me' n number of seeds/fruit I 11.3 ra' ge of seeds/fruit 1~3 I es: imated annuals~ed crop .3/flowering v.

I

...... ""'J 72

the values obtained from intraclonal crossestsuggestingthatpoJJjnators are effective in transferring pollen within clones ofE. aTbidUm.

Seed Dispersal

Th~ outgrowth present on both theE. propullal'ls al'ld=L==albldurrcseed is illustrated in Fig. 26. This structure has been suggested as a means

'ito attract ants twhich would then

'back to their nests (Gates t 1943). E. propu11 ans seeds fed upon the outgrowth the ants did not move or carry seeds interval.

Because the ants were environment t theirbehavior may notbe·inclicat-ive-of--how__tbe-v---benave nature. 73

Jig. 26. Seeds of f.. propullans (left) and E. albidum:..c(r5ght) showing the outgrowth present on each of the seed coats. . 74

DISCUSSION

Morphology

are ciistinguishable bas~d on the mor- . albidum. The differences in forlhe six characters measured (Table 21). ma.YI"'~presentferti Tehybrid plants from the propull ans. Some of theE. propull ans oser to the population of the inter- mediate size class,sugges that these plants may of the hybrids"to the The relative rarity of ate plants further suggests-that the intermed-i-a-1;e-s---a-Y'e~n0t--f~q\;le·ntly formed. The putative hybrid which is suggested from t~e morphological analysis will require chemotaxonomic investigations to further define the origin of this taxa. Morphologically, f. albidum and E. propullans (depicted as a dwarfed 1. albidum) are two distinct species. the characteristics of the three species of the white-flowering Erythronium of eastern North Ameri ca (f. propull ans, f. a1bi dum, and f. mesochoreum) are compared in Table 22.

Demographics and Phenology

Within Rice and Goodhue Counties, popula.tiOns of E. propullans and E. albidum grow in relatively undisturbed, climax, maple-basswood forests """"'''ii,~",,,.jt,'ti=l.W''i{,q:.:~;:;.~=~·~.ll!1ll1·fU!l~'Jm/ll!ll1!m1ll!S:~tIIl..~~iIY_,~:lil~~'U;·~-~~"""~·---""--"--·"·--~---·~-·~~~~'

Fabll~ 21.Comp.a~i$On between I"propullans, the intermediate size ela~s and E. albiiPumq ,for 6 morphological characters. i!

E. propull ans intermediate size class E. albidum meani flower 1ength 1.19 em 1. 90 em 2.78 em meani peduncle length I ', .. ".,' 6.48 em 9.78 em 12.25 em meanll ea,fl ngth 9.06 ~m 11.74 em 12.22 em 7 I 1ea fi pedunel e' rat!ib 1. 5: 1 1.2:1 1:1 i mean!, leaf >wi dth 2 02

11 I T :' i 1distanee from t~e origin of,the stet offshoot to !the apex of the peduncle

!, I, Table 22. ~omparison of the morpho ogyeco!logy~ and c~ro~osome numberofE!.pt-opul1al1ls,E!. _._. __ ... : and E. mesoqhoreum., .

I. prdpullans E. albidum

i flower length .8-1. 5 em 2.1-4.3 em I perianth shape completely reflexed completely reflexed usually straight leaf width 1.2-2.8 cm .7-4 cm 1-2 cm leaf traits mottled u ually nonmottled lateral steill offshoot present! absent i prop~gation Ilof I-leaved forms 1 to 3 runrers 1 1 dropper habitatpre~erence shaded ravrnes pr~1ries, pastures, I I, dry woods diploidchrgmo$omenumber low 4o.:b 22 :1· • I ! i. ..;,i1 lE. mesocho~eum data and (l95h and RObertson! '1 . I i I

...... 0'1 77

In general, E. albidum grows.~ abundantly adjacentt() rivers or streams. • ______.....•..••_•. ,. __'0. .'.,.•,.,. .,'•••_ __• _ ...•...... __ .."'. .• __ .". " __ ._ on south facing slopes while I. propullans is usually restricted to north facing slopes of ravines. Individual clones of E. albidum are generally larger in area than I.

propullans in colonization. I (as Site I), the population of I. pullans populations, however, the per clone is declining, although the number of clones is stable. Because I .. propa 11 ans is unable to thri ve i ngt~tllrQ~_~L~reas, efforts_1()_Jllilin­ tain suitable habitat should give careful consideration-to-the successional stage of the forest. Other herbaceous plants, such as Trillium nivale, Dentaria laciniata, Hepatica acutiloba, and , may serve

~-'"""""~~.,,~,,,,,,-_.-~ as indicators of appropriate habitat in which I. propullans may grow. The phenology and life cycle of I. propullans and E. albidum are similar. Both species emerge, flower, initiate runner growth, form fruit, and senesce synchronously. Because I. a1bi dum and E.propullans are vernal species .which complete their above ground life cycler 1rr a limited amount of time, staggered blooming periods which couldinslIreminimalhybridi­ zation andbackcrossing does not occur. Phenologicaliis()latjon~which might strenghten reproductive isolation between the two species, therefore,

".__• _ _ • "__·o~~·. ----~,._------does not exist. The ability of I. albidum and I. propullans to remain momphologically distinct, with the excep!!onofa few intermediate plants, suggests, however, that some degree of reproducttveisolation must exist between the two species. 78

Reproducti ve Capaci ty .. ..~~._--~_._--_. -- .. - The capacity for sexual reproduction is greater in E. albidum than in ~. propullans. The amount of pollen produced per E. albidum flower \'las six to eight times greater than An aver- age of 57% of E. a1bidum pollen wa~; staiin_~~~_~~_==~~:~:i~!!lJ~~~~~~~:~~'~~:~~ .2n

Eo propullans. The average ~. ~~~. T II ower naS~~OOVlll e~s, has only 5.3. The contrast

~eed crop of I. albidum is 4 seeds per flowering individual in E. propullans. The pollen-ovule ratio may be used a particular p1ant. This prediction is based ._ . tio increases as the plant nr(,nrpc::c~pc:: situation to an outcrossing (allogamous) si

'i', '" An outcrossing index devised by Cruden (1977) utilizes pollen... ovuleratios

~_.._ .. -"'-==,,,,,,,~.m"_._. as indicators of breeding systems and correlatespollel1...ovuleiratioswith success;onal stages of habitat. Using this index ,bothE.propul Hms and

E. albidum would be predicted to be facultatively xenogamoLl$::speCi~$ (outcrossing and regularly self-compatible) and exist in Jateisuccessional stage habitats (habitat undisturbed) and flower when pollil1a'tQrsare un­ reliable (early vernal insects). The I. albidumreproductivestrategy is correctly predicted, however, I. £!:.QP.ullans is extremelyxenogamous, pro­ ducing seeds only when crossed with I. albidum.

Vectors of Pollen Transfer I. albidurn and I. propullans are bell-shaped blossoms, typically suited to a "crawling-in" behavior of bees (Faegri and van der Pilj, 1971). 79

Both species were visited primarily by Andr~nacarlini Ckll., small so·l­ itary bees which were observed to transfer pollen between clones of both species. According to Linsley (1958) members of the genus Andrena are 01 i go­

1ectic bees, i.e., they di sp1ay fl ower specifi~i~~-a:!1~constancy and are physiologically synchronized with visit. The significance of oligol competition for flowers between more efficient pollination appropriate pollen source is to alternate po 11 en sources. This type of plant-pollinator relationship has been exemplified in the pollination case studies conducted within the genus Erythronium U1ich- ener and Rittenmeyer, 1956; Utech and Kawano, 1976a). Andrena carlini,

...... _,.- -~~~<~,~,~"."_•.• _~" .. ".,, the pollinator of Erythronium albidum and~. propullans, \'/as observed to be largely dependent of Erythronium for pollen, although it was also ob­ served to collect nectar from Dentaria laciniata and Claytonia virginica. Whereas Erythronium albidum and/or ~. propullans may be necessary for the survival of this bee, the absence of Andrena carlini would certainly ex­ clude the only reliable vector for the transfer .of pollen for E. albidum and I. propullans. The pollination and foraging behavior of Andrena carlini is well suited to both the I. albidum and I. propullans flower, even though I. propullans is much smaller in size. Sufficient amounts of pollen adhere to the ven­ tral surface of the insect, promoting the transfer of pollen from clone to clone. The preference of Andrena carlini for I. albidum probably

• 80 reflects the larger size and conspicuousness of the flower as well as the relative abundance of pollen and nectar, giving E. albidum a competitive edge for pollinators over E. propullans. However, where large clones of E. pr'opull ans occurred wi th few f. a1bi dum evi dent, the poll inator vi sited

E. propullans flowers at the exclusion of E. alb"i"?!Jm. This suggests that the bees are under strong resource allocation regimes.

Collectively", the results for seed production in f. albidum is highest 23). When flowers were exposed to their own the same clone, far fewer seeds with pollen from another clone within the same site or from another site.

E. albidum is facultatively outcrossing, adapted for outcrossing but able to self-fertilize,'" as predicted from its pollen..ovule ratio. Seed production in [. albidum by self-fertilizatfonwas found to be most successful when the pollen of [. albidum reached the stigma within 24 hours after the anthers dehisced. In 1978, .whenflowe.rs. .w~rebagged and hand-pollinated, 50% produced fruit with a meanof2.3seeds per fruit. In 1977, when bagged flowers were not hand pollinateCl,frllitand seeds did not form, suggesting that some physical mechanism is required for the pollen to reach the stigma. Two such mechanisms were observed. First, the foraging activity of Andrena carlini often bent the style and stigma toward the making contact with it. Second, the daily opening and closing of the flower brought the stigma in close proximity to and often touching the outer longer whorl of anthers. The latter mechanism can be of significance to I. albidum as it can then produce seeds by self-fert- 1\ .... " Table 23. Summar~1 of results obt~ined from breedi~g experiments in 0[. propullarisand .[."albidum.

INBREEDING 4 • OUTCROSSING apomixis seTfing intraclonalinterclonal interpopulational interspecific control

E. albidum %with fruit 0% 50%1 40% 47% 3% 73% 69% mean number of seeds/fruit 2.331 4.·2 12.5 9 7.7 5.8 i range of seeds/\l'ruit 1_61 1-2 3-23 1-24 1-20

E. propullans %with ifruit 0% 0% 0% 0% 20%1 mean number of ~eedS/fruit 1.66 ! range of seeds/fruit 1-2

11978 data only 82

ilization even if pollinators are not present. However, the traits of protandry and stamen dimorphism limits the extent to which I~ albidum can produce seeds from this mechanfsm. The results of crosses between the two species were striking; I. al­ bidum readily produces seeds with is only capable. of producing seeds when polli lack of seed production by I. propullans when s:uggests that I. propullans cannot produce reproduction. The scarci of hybrid plants class) may be a result of pollinator constancy or a tality of seeds of hybrid origin. Viability of ship and eventual recruitment, interspecific hybridization, needs additional

Evolutionary Relationships ._~~~~ The pattern of distribution in the white.ErythY'onfUmiofieasternNorth America5where the tetraploid species are distributecll1ort b·9fthediploid species, is similar to that reported by Parks and Hardin (1963) for the yellow Erythronium species of eastern North Arnerica.Theyhypothesized that the yellow tetraploid, I. americanumwttbaicliplofd chromosome number of 48, is an allotetraploid formedbythellatural hybridization of two diploid species, E. rostratum (2N=24) and I. umbilicatum (2N=24). Hybridization probably took place in Alabama (the region with the most yellow-flowering species of Erythronium) during the Pleistocene Epoch; the tetraploid then migrated northward following the retreat of glaciation to its present distribution (Parks and Hardin, 1963). Simflarly, thisscenerio might be applied to the white flowering taxa. 83

Species BecauseE. propullans is maintainea exc+uslvelY oyvegetatlve repro­ duction, all progeny are presumably genetically identical. If indeed the entire species lS genellcally uniform, then the inability of I.propullans. to successful1ybreedintraspecifically may be explained by suggesting that the entire species is one self-incompatible clone distributed in spate"."='To test thi's'''hypothes i s ,chelllotaxoHomicteehrri,Etues,particul arly isozyme analysis, are being applied to determine the extent of genetic variability within I. pr6pullans, a fac'torof lmp()rtance to the persistence of rare plants (Stebbins, 1942).

l' :1 84

A natural low. s.eed set in .s. a1hi dum appears to be the rule. The pers.istence of E. propullans is insured by the. multipltc<.ltion of bulbs in

lieu of reproduction by seeds. Critical seed characteristics are trans­

ferred to the bul bs, such as (1) the abil ity to store starch during spri ng . "growth, (2) the ability to provide protection for the--E1evele~~ing:snoot­

withi n, and (3) the abil ity to overwi nter and endure freezi ng temperatureS.

The emphasis on bulb reproduction may be viewed as a successful adaptation for ephemeral, vernal sporophytes. For example, bulb formation does not

rely upon pollinators which are unreliable in tion of bulbs in spring insures that suc:ceSS1"U petuated and any genetic heterozygosity will not be

i nb~eedi ng of re1ati ve ly i so1ated nnr)l/li'I t;-nrls·:.--I-InIAIPv@·r~:--t~!'lp'-nt·orll-H,·t;fm

of seeds by sexual reproduction; whether by recombi

is the.. only means by which new genetic combinations mayby

in .s. proplillans, mi ght reduce the prospects 1nnrl_r,o\"m the species. Lacking continual input of genetic

population of mature plants, the species environments or to colonize new habitats. An ." " ...... species-could be eradicated if the The "Big Woods" maple-basswood forest in thrives is but a remnant of its original post-glacial range. This restrictiveness

Woods II is conti nui ng today. under~.h~.p~~ss~res _~!'.ought on by agri­ cultural and urban development. Therefore, the persistence of this species

is in doubt •. The species will become extinct-unless its critical habitat

is protected. 85

.~ ~ CONCLUSIONS

BQth I. p'ropullans and I.albidum were pollinated by Andrena carlini

o • Ckl1.,an oligolectic solitary bee, which__\'Las_Qbser'y_ed t&~j~ransferp()Jlen between either clones of _I. a1hi dum, or I. propullans. The pendant-shaped blossoms of E. albidum and .s. propullans were well suited to its pollina­ tion behavior. Both species are protandrous and have dimorphic stamens to promote outcrossing. The pollen-ovule ratio and breeding experiments conducted within the species indicate that E.albidumisfaC:ultatively outcrossing and able to produce seed when crossed with .s. propullans. ' .s. propullans, however, is xenogamous, not self-compatible and able to produc~_s~!=d when crossed with~lbidum. A12\'1l']Rt()c:I~~cljx~<::a- pacity, a low level of natural seed production, and an inability to sexually reproduce with members, of its own species ,suggests---thatc-thepersisten of .s. propullans is maintained only by the multiplication of bulb 86

Of the two s.pecies, I. propullans isles.sabundant in number of individuals per clone and number of clones per site. This, coupled with the possibility of agricultural or urban development within its populations, and the lack of input of genetic variability required to adapt to a dras- tica lly changed environment, foretell sthe exti-rpa-ti0fl~~0~f i-t-s populations. Unless its critical habitat is to become extinct. 87

LITERATURE CITED

Applegate, E. 1. 1935. The genus Erythronium: blltional study of the western North American 58-113.

Blodgett, D. H. 1894. On the deve1oplTlent of the bul b bftheadder I s tongue Bot. Gaz. 34:61-67. Blodgett, D. H. 1900. Vegetative reproduction and multiplication in Ery- throniurn. Bull. Torr. Bot. Club 27:305,.315~ . Blodgett, D. H. 1909. The stem offshoot in Erythroniumpropullans Gray Johns Hopkins University Ci rcul ar3",5:36.~39~._.~._.~._·_~~~~_ Blodgett,.D.H. 1910. The origin and development bfbt.llos in the genus Erythronium. Bot. Gaz. 50:340-373. Carlson, C. R.,.J. F. Cummins, R. J. Edwards, G. F. Harms, J. J. t4urray, R.E. Rolling, R. Severson and E. K. Woodworth: 1975. Soil survey of Rice County; Minnesota. Uni tedStat·€S'~Bepo:rtmen'l;··ofAgri cul ture Soil Conservation Service in cooperation with the University of Min­ nesotaAgricultural Experiment Station. 118 pp. Cave,M-; 1966. In: Documented chromosome numbers of plants • Madrono 18:245-246. Cooper, D. C. 1939. Development of megagametophyte in Erythronium al- bidum. Bot. Gaz. 100:862-867. -- Cruden, R. W. 1976. Intraspecific variatfonil'lpollen-ovule ratios and nectar secretion - preliminary evidence of ecotypic adaptation. Ann. Missouri BoL Gar. 63: 277 -289. . Cruden, R. W. 1977. Pollen-ovule ratios: a conservative indicator of breeding systems in flowering plants. Evolution 31:32-46. Daubenmire, R. F. 1936. The "big woods" of Minnesota .. Ecol. Monogr. 6:233;;;268 ~ . Faegri, K. and van der Pijl. 1971. The principles of pollination ecol­ ogy. 2nd ed. Pergamon. New York. 291 pp. 88

Gates, B. 1943. Carunculate seed dissemin~tioD~~.YantS. Rhodora 45: 438-445. Graff, P.- W. 1916. The stamens of Erythronium americanum. Torreya 16: 180-182. Haque, A. 1951. The embryo sac of Erythronium americanum. Bot. Gaz. 112:495-500.

Ireland, R. 1957. Biosystematics of O1fKam;as~ ~~:.....:..::= choreum. M.S. Thesis, Universi . I

Knerr, E. B. 1891a. Erythronium ~~~~~.

Knerr, E. B. 1891b. Erythronium ~~~~~. 13:20-21. Linsley, E. G. 1958. The ecology of solitary bees. Hilgardia 27:543-599.

Marschner, F. J. 1930, Map of the origional vegetation of Minnesota. Office of Agricultural Economics, U.S. Department of Agriculture.

Meads, M. E. 1893. The range of val"iatio~n~_in_$~pecies of Erythronium. Bot. Gaz. 18:134-138. Michener,.C.and.C Rittenmeyer. 1956. The ethology of Andrenaer,ythronii with comparat ve data on other speci es (Hymenoptera, Andrenidae); University of Kansas Sci. Bull. 37:645-684. Mitra;J ~ 1965. An improved method for~pl-a-nt-ehr0mes·emepreparation. Nucleus 8:179-182. Parks, C. and J. Hardin. 1963. Yellow of the eastern United States. Brittonia 15:245-259. Pickett, F. L. 1917. The length of Erythronium stamens. Torreya17: 58-60. RadfoY'd,j\., W. Dickison, J.Massey and C. Bell. 1974. Vascular.plant systematics. Harper and Row, New York. 891pp. Rickett, H. W. 1937. Erythronium mesochoreum. Rhodora 39:101-105. Robertson, K. 1966. The genus Erythronium in Kansas .. Ann. Missouri Bot. 53:197-204. Rosen~ahl, C. O. 1919. Variations in the flowers of Er,ythronium pro­ pullans Gray. Torreya 19:43-47. Schaffner, J. 1901. A contribution t()theTlfe history and cytology of Erythronium. Bot. Gaz. 31:369-388. 89

Sernander, R. 1906. Entwurf einerMonographiederEuropaischen.Myrme­ choren. Kengl. Svenska Ventenskap, Handl. 41:1-410. Smith, F. 1955. Megagametophyte deVelopment fl1fiVe m spetieS of Ery'" thronium. Amer. Jour. Bot. 42:213-224. Stebbins, G. L. 1942. The genetic approach to problems of rare and endemic species. , Madrono 6:241-258. Stebbins, G. L. 1971. Chromosomal evolution in higher plants. Edward Arnol dttd., London. 216 pp. Struik, G. 1957. The distribution and reproduction of some herbaceous plants in a maple forest. M.S. Thesis, University of Wisconsin. Swanson, S. D. ~nd S. H. Sohmer. 1975. The biology of Podophyllum pel­ tatum L. (Berberidaceae), the may appTe;~m~I.mmThemtransfer of pollen and success of sexual reproduction. BulL Torr. Bot. Club 103:223-226. Utech, R. and S. Kawano. 1975a. Biosystematic studies in Erythronium (Liliaceae-Tulipeae). 1. Floral biology of E. japonicum Decne. Bot. Mag. Tokyo 88: 163-176 .. Utech, R. and S. Kawano; 1975b. Biosystematic studies in Erythronium (Liliaceae-Tulipeae). II. Floral anatomy of'E. japonicum Decne. Bot. Mag. Tokyo 88:177-185. Utech: 'R. and S. Kawano. 1976. (Li 1iaceae-Tul i peae). II I. SomatimCmmmmm~~_2~ ~_~..:,_mmm:m.mm:::mm ~=-:::'-:-:-";";:;;"" Decne. Cytologia 41:749-755. Wright, H. E. Jr. 1956. Sequence of glaciation in eastern Minnesota. Geol. Soc. Amer. Guidebood, Minneapolis Meeting. 90

Appendix 1. I~ propullans herbari.um s.peci.mens.examJne

Minnesota, Rtce County, Faribault (type locality), St. Mary's Preparatory School: May 1871, Hedges. sn. (G) TYPE SPECIMEN, duplicate at NY; F. S. Beane sn . (MIN); May 1944, K. Orwi gsn.~L~,J1I10 ;]MayJ953, Rosendahl, Mayla, Jalm and Dobies 7962 (HIN). Shattuch Pt7eparatory School,

Faribault: 22 April 1977, Banks , . Faribault: May 1890, F. S. Beane (G); 1872, Darlington (NY); May Darlington sn. (NY); M. Hedges ' 1. Martindale sn. (NY); 1891, Patterson sn

so. (F, ~UN, MO, NY, PH); May J. H. Sandberg sn. (F, G, MICH, Aiton $n. (F, MIN, MO, NY); 14 r1ay 1961, T.

T. ~10 1 (M J. r ey 1041 IN); H. S. sn. U1I.N)'~~~~~------'---"-"'~'. .. Minnesota Goodhue County, Kenyon: 30 April 17 1 1946, J. W. Moore and N. L. Huff 18443 (G, MIN); 6 t1ay 1945, C. O. Rosendahl 7722 (MIN, MO); 6 May 1975, Sohmer 9738, 9767 (UWL); 10 May 1975, Sohmer 9769, 9770 (UWL). Cannon Falls·: April 1895, G. B. Aiton (MIN,

NY); E. 01 Rosendahl, E. E. Dutton and E. T. Nielsen 2934 (mN); 10 May 1919, C O. Rosendahl and F. K. Butler 3786 (MIN); 3 Hay 1918, C. O. Rosendahl and F. K. Butler 3451 (G, MIN, NY); J. L. Scofield 1015 (MIN).

ZumbrQta~:.MClY 1892,A.~_~. Ballard (G, rn N, NY). Minnesota: Darlington sn. (tUN).