<<

r9901 NAUMAN: INTERGENERICHYBRIDIZATION

Principes,34(4), 1990, pp. l9l-198

lntergeneric Hybridization between and (Palmae:)

CrmroN E. NeulreN Fairchild Tropical Garden, lO9Ol Old Cutler Road, , Florid.a 33156t

Assrnecr assertionthat the representan inter- generic hybrid betweenthese two . Unusual plants of what initially appeared to be a species of Thrinax were discovered in the Lower Hybrids among vascular plants are suf- Keys. The plants were observed to produce ficiently well known that several criteria abundant ,but anthesis and fruit set can be used to establish a hybrid origin. did not occur. The sterility, apparent intermediate- Among these,the most important criterion ness in several leaf characters, and consistent co- is that the putative hybrid occurrencewith (Jacq.) Bai- exhibits mor- ley and Thrinax morrisii H. A. Wendl. suggesteda phological intermediateness between its hybrid, Three criteria were evaluated to test the parents, usually in several characters. hypothesis of hybridity: morphological intermediate. Seven additional criteria were listed by ness, occurrence in the geographical zone of overlap Gottlieb (1972\. Since Gottlieb was con- of the putative parent species, and occurrence in ecologically intermediate habitats. Fifteen of the 22 cerned primarily with the confidence of characters examined exhibited intermediateness, three detecting a hybrid origin for stable, pre- exhibited questionable intermediateness, and five sumably reproducing entities, and since exhibited apparent heterosis. Field studies indicated these plants do not produce functional the putative hybrids to occur only with both parental flowersothree of these criteria cannot speciesand only where the parents occur in immediate be proximity to each other in ecologically intermediate appliedto the presentstudy. These are the habitats. presence of partial F, fertility, the pres- ence of unusual amounts of interpopula- Unusualplants of what initially appeared tional variability in the hybrid, and occur- tobe Thrinax morrisii H. A. Wendl. were rence in more recent geologicalformations discoveredby a local naturalist (Ann Wil- than the parent species.The relatively long liams) on No Name Key, Florida. The generationtime in Coccothrinax and Thri- plants were subsequentlyfound in several nax, and the logistic difficulties of crosses additional locations on neighboring Big Pine for experimental synthesis of the hybrid Key (Fig. I). During several years of (Read 1975), are sufficiently prohibitive observation,none of theseplants produced that this criterion cannot be evaluated.An flowers, but abortive inflorescenceswere additive chemicalprofile for marker paren- produced in abundance. This production tal compounds is also typical of hybrids of abortive inflorescencesand an apparent but is here consideredessentially equiva- intermediatenessand co-occurrence with lent to morphologicalintermediateness for' Coccothrinax argentata (Jacq.) Bailey practical comparisons.Other criteria relat- (fig. 2) and T. morrisii (Fig. 3) led to the ing to polyploid hybrids, e.g., chromosome number, pairing behavior, etc., could not I Present address:30I Hesler Biology Bldg., Univ. be applied. Previous chromosome counts of Tennessee.Knoxville. TN 37996-I I I0. of T. morrisii and of C. argentata (Read. r92 PRINCIPES lvor. 34

at the herbarium of Fairchild Tropical Gar- den (FTG).

Results Intermediatenessand apparent hetero- sis is indicated for all of the characters of gross morphology examined. The scatter diagram (Fig. 4) illustrates severalofthese intermediate characters. Other characters showing this pattern include: presenceor absenceoftransverse veins, segmentshape ratio (distance to widest point along the segment,/segmentwidth), greater palman length (distance from the hastula to the sinus between the two central segments), degree of segment splitting, frequency of split petiole bases,and shapesofthe adax- ial and abaxial hastulas. The ranges of character state expressionin three char- acters, lamina outline, lamina folding, and segment habit are considerable,and dif- l. Habit of the putative hybrid ferences are in the form of frequenciesof characterstates. Heterosis is indicated for five charac- 1963, 1964, 1965, I97 S;Venl

2. Habit of Coccothrinax argentata. 3. Habit oI Thrinax morrisii.

argentata by a mesophyll of both palisade mer location on No Name Kev (Monroe and spongy mesophyll cells. The putative Co., Florida). The putative hybrids only hybrid exhibits a mesophyll of largely pal- occur in mixed populationsof C. argentata isadecells and a more or lessdistinct spongy and T. morrisil, and in all casesless than Iayer. Fibers surrounding secondary veins five meters from plants of both parental in C. argentata extend from the abaxial species.The putative hybrids are not known to the adaxial epidermis, interrupting the from areas or habitats occupied by only hypodermis both abaxially and adaxially. one of the parental species. The mixed In T. morrisii, these same fibers rarely nature of the parental populations can be extend acrossthe entire width of the meso- taken as evidence of ecological interme- phyll and generally do not interrupt the diateness.Coccothrinax argentata occurs hypodermis. The putative hybrid exhibits in coastal pinelands or strand habitats in a condition closer to that seenin C. argen- the lower and along the east- tata,bul the fibers generally do not inter- ern coast from southern Dade Countv. rupt the hypodermis. Florida, north to Lake Worth in Palm Field studies verified the putative Beach County (Fig. 6). Thrinax rnorrisii hybrid's occurrence in the geographical occurs in wooded areas, hammock edges range of overlap and in habitats ecologi- and in mangrove and tropical strand hab- cally intermediate between the parental itats from Key Largo southward (Fig. 6). species.Hybrids are known from22loca- Coccothrinaxargentata co-occurswith I tions on Big Pine Key, and from one for- morrisii only in the slash pine flatwoods t94 PRINCIPES [Vor. 34

Table 1. Means and rdnges of expressionin characters examined/or Coccothrinax argentata, Thrinax morrisii, and their putatiae hybrid. ("mostly" meansmore than 807o "f ,tt" t"-tt"+

Character C. argentata Hybrid T. morrisii

1. Petiole base split no mostly yes, rarely no yes 2. Petiole apex width 1.40(0.9-r.9) r.72(r.2-2.0) r.47 (0.6-2.0) (cm) 3. Abaxial hastula straight to broad-round- straight to obtuse-round- rounded, deltoid apical shape ed, apical tooth pres- ed, apical tooth pres- tooth mostly absent ent or absent ent or absent 4. Adaxial hastula obtuse to rounded trian- obtuse to acute rounded, obtuse rounded, apical shape gular, apical tooth apical tooth present tooth mostly absent present 5. Adaxial hastula 0.6 (0.5 0.8) r.2(r.0-r.s) 0.7(0.s-0.9) length (cm) 6. Abaxial lamina indu- dense moderate scattered & few ment density 7. Wax dots present absent few to absent presenr B. Transverse veinlets absent obscure but present obvious 9. Segment number 3s.0 (29-42) 44.4 (39-48) 49.2(4s-s6) 10. Lamina outline 3/+to )I orbicular Yr to )1 orbicular I to >1 orbicular Il. Lamina form flat or with I raised mostly with several mostlymore than I group of segments groups of raised seg- groupof raisedseg- menTs ments 12. Segment habit arching to mostly droop- mostly drooping or arch- straightto arching ing or pendent ing-pendent 13. Central segment 5.72 (s.2 6.4) 8.r8(7.s-9.0) 6.77(6.0-7.9) length (dm) 14. Central segment 2.34(2.r-2.6) 3.52(3.0-3.7) 3.93(3.s-4.6) maximum width (cm) 15. Distanceof maxi- 2.06 (1.7s-2.23) 3.49(2.92-4.4s) 3.34(2.89-4.08) mum segment width from hastula (dm) 16. Shaperatio charac- 2.80 (2.37-3.20) 2.4(1.68-2.93) 2.04(r.9r-2.38) ter I3./character 15 17. Greaterpalman r.i3 (0.80 r.91) 2.97(2.67-3.s4) 3.06(2.75-3.68) length (dm) 18. Lesserpalman 1.07(0.52-1.8?) 2.8(2.34-3.37) 2.74(2.49-3.26) length (dm), second sinus either side of the central segments 19. Degree of splitting 16.68(r0.8-36.9) 39.22(33.0-51.6) 4s.r (40.7-47.2) character 17 x 100,/characterl3 20. Abaxial fiber bundle bundles contained entire- bundles not contained bundlesoccasionally in- intrusion into the ly within the hypoder- within, but interrupt- terruptingthe hypo- hypodermis mis ing the hlpodermis dermis 2I. Secondaryvein fiber bundle fibers extending bundle fibers extending bundlefibers not extend- bundles extending across ennre meso- across entire meso- ing across entire me- across and inter- phyll and interrupting phyll and not inter- sophyll, nor interrupt- rupting the hypoder- the hypodermis rupting the hypoder- ing the hypodermis mis mis 22. Mesophyll of all pal- mostly equal layers of mostly unequal layers of all palisade cells present isade cells spongy and palisade spongy and palisade cells present cells present r99ol NAUMAN: INTERGENERICHYBRID 195

60

tl I t L aL M O Coccothrinoxorgenlolo 6 A Hytria o ! Thrinox llE30 = z. no indument Wox Dol Densiiy ( dense O c o moderote O E J CD sporse o o o u, 20

Moximum Segment Width (cm) 4, Scatterdiagram for fou charactersshowing intermediateness in the putativehybrid between Coccothrinax argentata and.Thrinax morrisii.

and bordering habitats of the lower Florida tram) Small, palmetto (Walter) Keys, these presumably represent ecolog- Lodd. ex J. A. Schultes,and 7. radiata ically intermediate habitats. Lodd. ex J. A. & J. H. Schultes.The first two species differ dramatically from the Discussionand Conclusions hybrids in a number of morphologicalfea- tures and are very unlikely to be part of Of the 22 morphological characters the hybrid's parentage., examined, 14 exhibited intermediateness however, becauseit is congeneric with ?1 in the putativehybrid (6-12, L4, 16, I7, morrisii, represents a possible parent. 19-22, Table l), five exhibitedapparent Thrinax radiata is not abundant on Big heterosis(2, 5, 13, 15, l8), and three Pine Key, being representedby only four exhibitedquestionable intermediateness (1, individualsin a singlelocation. At this loca- 3, 4). Field studieshave shownthe putative tion, the plants are growing in a roadside hybrid to occur in the geographical and area where there is evidence of dumping ecologicalranges of overlap for the paren- and the plants may represent accidental tal species. introductions. The species does not co-' Although the evidencefor the presence occur with the hybrids anywhere on Big of hybrids is convincing, another possible, Pine Key or No Name Key. The hybrid but unlikely, parentage exists for the is also absentfrom locationswhere T. mor- hybrids. Three other palm species occur risji and T. radiata do co-occur such as on Big Pine Key, repens (Bar- Big Coppitt Key. Therefore, T. radiata 196 PRINCIPES lVoL. 34 Thrinoxmorrisii

Hybrid

P

IOOpm

Coccolhrinoxorgentofo

P r9901 NAUMAN:INTERGENERIC HYBRID

doesnot meet the criterion of geographical co-occurrencewith either C. argentata or T. ntorrisii. Additionally, morphological characters of the hybrid do not indicate T. radiata to be one of its parents whether T. morrisii or C. argentala is taken to be the secondparent. For example,T. radiata has glabrousabaxial leaf surfaces,and the hybrids suggesta parent with a denseabax- ial indument. Additionally, the characters of greater palman length, segment num- ber, and central segmentmaximum width contradict a hypothesiswith Z. radiata as oneof the parentalspecies (see Read I 975 for more details on T. radiata). Compared with other proposed inter- HYbrid (cf. generic palm hybrids Balick et al. 6. Florida distributionsfor Coccothrinaxargen- I9B7), theseplants are unusualin at least tata, Thrinax morrisii, and their putativehybrid. two respects.First, they are apparently sterile, the inflorescencesabort (i.e., they do not fully extend and the flowers do not be needed to evaluate this alternative to emergefrom the peduncular bracts), while hybridity further. hybrids between Attalea and Orbignya This study has provided support for the and others in the Attaleinae (Balick et al. hypothesis of hybridity. The plants are l9B7) are fertile. Second, the putative morphologically intermediate between their hybrids exhibit a morphology more like one proposed parental species in a number of parent than the other, in this case T. rnor- morphological characters, and occur wher- risii. ln addition to the characters listed ever the two parental species co-occur in in Table l, the inflorescencearchitecture ecologically intermediate habitats. How- and indument are more like those of Thri- ever, the data have also suggested that na.x tharnCoccothrinax. This suggeststhat these plants represent stressed individuals these plants may only represent stressed of T. morrrisil, though only by not refuting individualsof T. morrisii. However,in other this alternative. More investigation of these populations of T. m,orrisil no individuals plants will be needed to establish their ori- have been found that possesscharacter gin. statestypical of the putative hybrid, e.g., the denser abaxial indument. Additionally, Acknowledgments typical plants of T. morrisii often occur less than two meters from the putative I thank D. Holle (Refuge Manager, hybrids and show no signs of stress. Thri- National Key Wildlife Refuge, Big nax morrisii, however, exhibits a wide Pine Key, FL) for permission to sample degree of morphological variation (Read on refuge property, Dr. R. W. Sanders 1975) and a detailed analysis throughout (Fairchild Tropical Garden, Miami, FL) for its range and in a variety of habitats will comments and suggestionson the manu-

(- 5. Diagramatic lamina cross sectionsfor Thrinax morrisii, Coccothrinax argentata, and the putative hybrid. Shadedareas represent fiber and vascular bundles. E : epidermis, H : hypodermis. Vertical bars represent the extent of palisade (P) and of spongy (S) mesophyll. Scale bar applies to all three sections. I98 PRINCIPES [VoL. 34 script, and especially Ann Williams (Big READ,R. W. 1963. Palm chromosomes.Principes Pine Key, FL) for bringing the hybrids to 7: 85 88. 1964. Palm chromosomes.Studies facil- my attention and for help with various itated by pollen culture on a colchicinelactose aspectsof the field work. medium. Stain Technol. 39: 99-106. 1965. Chromosomenumbers in the Cor- LrrrRerunn Cnnl yphoideae.Cytologia 30: 385-39I. 1975. The gents Thrinax (Palmae: Cor- Beucx,M. J., A. B. Axnrnson, eno J. T. nnMnort- yphoideae).Smith. Contr. Bot. No. 19, pp. 1- Ros-Cosr,r. I987. Hybridizationinthe Babassu 98. Palm Complex. II. Attalea compta x Orbignya VENKATASUBBAN,K. E. 1945. Cytologicalstudies oleifera (Palmae). Brittonia 39: 26-36. in Palmae. Part I: Chromosomenumbers in a GorrLIEB, L. D. 1972. Levels of confidencein the few speciesof palms of British India and Ceylon. analysisof hybridization in plants. Ann. Missouri Proc. Indian Acad. Sci. 22: 193-207. Bot. Card. 59:435-446.

CLASSIFIED

PALMS-Indoor-Cold Hardy-Tropical. Over 240 species.Air mail shippedin original container(no bare rooting). Grown in temperate and sub-tropicalzones in our own essentially sterile palm mix. We are certified to ship to all states requiring inspections. Catalog $6 (refundable).An excellentcondensed reference. THE GREEN ESCAPE,P.O. Box 1417, Dept. P, Palm Harbor, Florida 34682.

PALM SEED FROM BORNEO. Many speciesavailable. For completeIisting write to B. L. & G. P.O. Box 1183, Kuching, Sarawak,East Malaysia.Fax. 082-256839.

LETTERS

Dear Friends, attended over the years have palm-related personalized license plate messages on I inadvertently omitted acknowledging them. For example, the license on my tn the Cham.aedoreaarticles published in "PALM pickup reads NUT." Principesin 1990 Inge Hoffman'ssupport I am making a collection of pictures of of my work with this group of palms. First such license plates, and invite all readers as Director of the Seedbankand later on with palm-related messages on their plates her own, she enthusiasticallysupported my to send me a snapshot for inclusion in a field work in and Central America. pictorial essay I am planning for Prin- I apologizefor this oversight and extend cipes. my sincere,although belated, appreciation for her support. Genv Wooo Donero R. Hornr

Dear Editor, I have noticed that many of the vehicles at various palm society functions I have