Reproductive biology of Cyrtopodium punctatum in situ...

Reproductive biology of Cyrtopodium punctatumin situ: implications for conservation of an endangeredFlorida orchid

DANIELA DUTRA,*1 MICHAEL E. KANE,* CARRIE REINHARDT ADAMS* and LARRY RICHARDSON†*Department of Environmental Horticulture, University of Florida, PO Box 110675, Gainesville, Florida 32611, USA and†Florida Panther National Wildlife Refuge, US Fish and Wildlife Service, 3860 Tollgate Boulevard, Suite 300, Naples,Florida 34114, USA

Abstract

Cyrtopodium punctatum (Linnaeus) Lindley is an endangered epiphytic orchid restrictedin the USA to southern Florida. This species has been extensively collected from the wildsince the early 1900s, and today only a few plants remain in protected areas. As part of aconservation plan, a reproductive biology study was conducted to better understand theecology of this species in Florida. Cyrtopodium punctatum relies on a deceit pollinationsystem using aromatic compounds to attract pollinators. Nine aromatic compounds wereidentified as components of the fragrance of C. punctatum inflorescences, including twocompounds that are known to be Euglossine bee attractants. However, this group of beesis not native to Florida. Of the four bee species observed to visit C. punctatum flowers inthe present study, carpenter bees (Xylocopa spp.) are likely to be the main pollinators.Pollination experiments demonstrated that C. puntatum is self-compatible, but requires apollinator and thus does not exhibit spontaneous autogamy. In addition, the rates of fruitset were significantly higher for flowers that were outcrossed (xenogamy) than for thosethat were self-crossed. Thus, the species has evolved a degree of incompatibility.Examples of natural pollination and fruit set were observed during the present study(2007–2008), but the rates of reproduction were modest as a consequence of the low plantnumbers and possible changes in insect densities as a result of anthropogenic influences.

Keywords: breeding system, Cyrtopodium punctatum, Orchidaceae, pollination.

Received 11 March 2009; accepted 21 May 2009

Introduction

Several types of plant rarity exist and different naturalprocesses may affect species density and distribution(Rabinowitz 1981; Fiedler & Ahouse 1992). Species may beconsidered rare based on habitat specificity, local popula-tion size and/or the size of their geographical range.However, plants that were once common in their rangemay now be rare as a result of anthropogenic influences(Partel et al. 2005). The factors influencing rarity need tobe considered for an effective conservation plan to bedesigned for any target species. Pollination systems are

often linked to plant rarity given the dependence oninsect species for cross pollination as this relationship isrequired for seed production and fruit set to take place.Decreased abundance or loss of pollinators often results inincreased rarity of the plant. The Orchidaceae is a verydiverse family, and pollination systems may range frombeing very specific with a plant being pollinated by onlyone insect, to very generalized with a plant being polli-nated by many species (van der Pijl & Dodson 1966;Dressler 1981; van der Cingel 1995). A comprehensiveinvestigation of the pollination system is a prerequisite todeveloping a conservation plan for any rare species.

Cyrtopodium punctatum (L.) Lindl. is an epiphytic orchidfound in southern Florida, Cuba, Hispaniola, PuertoRico and the north-western Caribbean coast of SouthAmerica (Romero-González & Fernández-Concha 1999).

Correspondence: Daniela DutraEmail: [email protected] address: Botany Department, University of Hawai’i atManoa, 3190 Maile Way, Room 101, Honolulu, HI 96822, USA.

Plant Species Biology (2009) 24, 92–103 doi: 10.1111/j.1442-1984.2009.00242.x

© 2009 The AuthorsJournal compilation © 2009 The Society for the Study of Species Biology

In Florida, the species is listed as endangered (Coile &Garland 2003); populations have been depleted because ofcypress logging and over-collecting during the pastcentury. Early accounts in the literature to C. punctatum asabundant throughout south and south-west Florida, par-ticularly in the cypress swamps in the Big Cypress region(Ames 1904; Luer 1972). Photographs from the early 1900sshow wagons loaded with plants being extracted from thewild. Consequently, only a few plants still remain in verysmall populations within inaccessible protected areas,such as Everglades National Park, Big Cypress NationalPreserve and the Florida Panther National Wildlife Refuge(FPNWR). Over the past 10 years, observations of plantpopulations in the FPNWR have indicated very low ratesof capsule formation and seed production in the remain-ing plants (L. Richardson, pers. comm. 2005). Additionalinformation regarding the breeding strategies and polli-nation mechanisms of natural populations of C. punctatumin Florida is critical to develop strategies for the recoveryof the species.

The study of pollination systems, including breedingsystems, is critical in rare plant conservation because itmay influence the genetic diversity of populations(Hamrick & Godt 1996). Pollination aids the long-termsurvival of most species because population persistencedepends on the recruitment of new plants from geneti-cally diverse seeds (Dixon et al. 2003). Most of the perti-nent literature on the pollination of C. punctatum refers toobservations conducted in only one area of the species’range (e.g. Puerto Rico; Ackerman 1995). In different partsof its geographical range, a species may possess differentbreeding strategies, which is not uncommon for tropicalorchid species in Florida. For example, both Epidendrumnocturnum and Secoila lanceolata produce seeds by agamo-spermy in Florida, but use animal cross-pollination in themore southern part of their range (Catling 1987; Brown2002).

Pollinator species may also vary from one location in thegeographical range to another. Pollination ecotypes havebeen reported by Robertson and Wyatt (1990a) in Plantan-thera ciliaris. In the present study, the pollination ecologyof two disjunct populations was compared and theprimary pollinators were found to differ between thesites. Cyrtopodium punctatum is reported to be pollinatedby Euglossa bees (van der Pijl & Dodson 1966; Jeffrey et al.1970; Luer 1972; Dressler 1993). However, as Florida liesoutside the range of euglossine bees, C. punctatum’s pol-linator(s) in Florida may vary (Chase & Hills 1992). InPuerto Rico, C. punctatum is reported to be pollinated byCentris or Xylocopa bees (Ackerman 1995). Bees of thegenus Centris were reported by Dodson and Adams (inLuer 1972) to visit flowers of C. punctatum in Florida.However, these observations have not been fullydescribed or verified.

Fragrances produced by plants attract pollinators andmay function to select the type and number of visitors(Hills et al. 1972). Floral fragrances have been associatedwith deception pollination systems based on generalizedfood deception (Little 1983) and can be an extra tool in thedetermination of pollination systems. Different fragrancecompounds or groups of compounds may attract differentpollinators. A larger array of compounds may be benefi-cial in deceit pollination systems because a higher numberof visitors will be attracted to the plant.

In many orchid pollination strategy studies, capsuleformation is commonly observed as the final result of thebreeding system study without taking into considerationseed viability (Wong & Sun 1999; Lehnebach & Robertson2004). Other researchers have used seed number andweight as a further measurement for breeding systemdetermination (Robertson & Wyatt 1990b; Sipes & Tepe-dino 1995). However, this type of measurement producesuncertain results because of differences in capsule matu-rity. Sipes and Tepedino (1995) reported that fruit weightand fruit size were not correlated with seed number, andinstead used seed number as another measurementfor breeding system determination. As capsules of C.punctatum require approximately 11 months to mature,repeated capsule measurement over the course of capsuledevelopment may minimize this limitation.

Another tool used to evaluate breeding systems is tet-razolium seed viability testing (TZ) and it has been usedto assess the viability of European and North Americantemperate orchid species (Van Waes & Debergh 1986;Lauzer et al. 1994; Vujanovic et al. 2000; Bowles et al. 2002)and tropical epiphytic species (Singh 1981). Assessingseed viability and germinability is crucial for in vitropropagation of orchid species and their conservation andmay be a helpful tool to more precisely determine effec-tive pollination strategies for threatened and endangeredtaxa.

As a prerequisite to developing a comprehensive con-servation plan for C. punctatum, integrated field and labo-ratory studies were conducted with the objectives of: (i)determining the breeding system of C. punctatum throughcontrolled pollination experiments and the effects on bothcapsule formation and seed viability; (ii) determining thepollinator(s) of C. punctatum in Florida; and (iii) determin-ing the volatile potential attractant compounds present inC. punctatum flowers.

Materials and methods

Study species

In the FPNWR (Collier County, FL, USA), C. punctatumplants are found growing epiphytically on cypress (Taxo-dium distichum) and less often on cabbage palm (Sabal

R E P R O D U C T I V E B I O L O G Y O F C Y R T O P O D I U M P U N C T A T U M 93

Plant Species Biology 24, 92–103 © 2009 The AuthorsJournal compilation © 2009 The Society for the Study of Species Biology

palmetto) or old stumps of massive cypress trees that werelogged in the 1940s. The vegetation classification for thehabitat is cypress forest moderately to densely distributedcypress with an open canopy. The understory is com-posed of Fraxinus caroliniana, Cladium jamaicense andAnnona glabra, among others. In other areas of Florida(e.g. Everglades National Park), C. punctatum can also befound growing on Buttonwood trees (Conocarpus erecta) infull sun. The inflorescences of C. punctatum are up to 1 mlong and can have 30–40 flowers.

Breeding system determination

One population with 15 plants located in the FPNWRserved as the source of plants for the determination of thebreeding system. Six plants located approximately 30 mapart were used for the breeding system experimentduring two consecutive years (2006–2007 and 2007–2008).Seven breeding system treatments (Table 1) were appliedto five flowers of each plant. Flowers were randomlyselected on the inflorescences. A total of 245 flowers wereused in this study. A plant located approximately 1.61 kmfrom the breeding system study population served as thepollen donor. Pollen was applied to flowers within 1 h.

The breeding system determination methods wereadapted from Wong and Sun (1999). An extra treatment(induced xenogamy) was added based on Stewart (2007).The breeding system treatments are described in Table 1.Pollen used in the artificial xenogamy treatment wasobtained from a plant originating from a different popu-lation in the FPNWR. Inflorescences were produced inFebruary and were bagged before the flowers openedwith 95 mm nylon mesh to prevent pollination eventsprior to initiating the experiment (Fig. 1). Plants remainedbagged until signs of capsule formation were observed.The percentage of flowers forming capsules was recordedand capsule development was recorded bimonthly for1 year by recording capsule length, width and abortion.The breeding system study was conducted during 2006–2007 and 2007–2008. Data were analyzed using general

linear procedures and Waller–Duncan mean separation(a = 0.05) with SAS v. 9.1 (SAS Institute 2003).

Tetrazolium seed viability test

Seed capsules produced from the breeding system studytreatments (2006–2007) were collected on 23 February2007 and dried over silica desiccant for 70 days at23 � 2°C. The seeds harvested from each capsule col-lected were transferred to 20 mL scintillation vials. Thevials were then maintained in cold storage (-10 � 2°C).

Three 5 mg seed subsamples (75–100 seeds) from eachcapsule of each treatment were dispensed into a 1.5 mLmicrocentrifuge tube and homogenized with the aid of avortex shaker. The methods of tetrazolium viabilitytesting follow Lakon (1949). A pretreatment test con-ducted prior to experimentation indicated that a 15 mintreatment in a seed scarification solution of 5% CaOCl2

was ideal for C. punctatum seeds. Percentage seed viabilitydata were collected for each of the three subsamples percapsule by placing seeds that were suspended in waterdroplets into Petri dishes and the seeds were then scoredwith the aid of a dissecting microscope. Approximately100 seeds were counted per subsample. Embryos that con-tained any degree of red or pink coloration were scored asviable and white embryos were scored as non-viable(Fig. 2). Percentage viability was calculated for each rep-licate by dividing the number of stained embryos by thetotal number of embryos. Data were analyzed usinggeneral linear procedures and Waller–Duncan mean sepa-ration (a = 0.05) using SAS v. 9.1.

Asymbiotic seed germination test

Seeds from the capsules produced from each specificbreeding system treatment were sampled (5 mg sub-samples of 75–100 seeds). Seeds from the same treatmentwere then pooled and dispensed into 1.5 mL microcentri-fuge tubes. The seeds were surface sterilized in a solutioncontaining 5 mL ethanol (100%), 5 mL 6.0% (v/v) sodium

Table 1 Pollination treatments used to determine orchid breeding systems

Breeding system test Flowers bagged Treatment Pollen source

Control No None Open pollinationAgamospermy Yes Emasculate No pollinationSpontaneous autogamy Yes None Same flowerInduced autogamy Yes Emasculate Same flowerArtificial geitonogamy Yes Emasculate Different flower, same plantArtificial xenogamy Yes Emasculate Different populationInduced xenogamy Yes Emasculate Same population, different plant

Adapted from Wong and Sun (1999) and Stewart (2007).

94 D . D U T R A E T A L .

© 2009 The Authors Plant Species Biology 24, 92–103Journal compilation © 2009 The Society for the Study of Species Biology

hypochlorite and 90 mL sterile distilled water for 3 min,followed by three repetitive 30 s rinses in sterile distilledwater.

P723 Orchid Seed Sowing Medium was prepared fromconcentrated stock solutions using the formulation devel-oped by PhytoTechnology Laboratories, LLC (ShawneeMission, KS, USA) and adjusted to pH 5.8 with 0.1 mol/L

KOH prior to the addition of 0.8% TC agar (Phyto-Technology Laboratories, LLC) and autoclaving at117.7 kPa for 40 min at 121°C. The autoclaved medium(approximately 50 mL medium/plate) was dispersed intosquare 100 mm ¥ 15 mm Petri plates (Falcon ‘Integrid’Petri Dishes; Becton Dickinson Woburn, MA, USA). Thebottom of each dish was divided into 36, 13 mm ¥ 13 mm

Fig. 1 Breeding system experiment. (a) Flower, (b) artificial removal of pollinia, (c) capsule measurement, (d) Cyrtopodium. punctatumplant with pollination bag in natural habitat at the Florida Panther National Wildlife Refuge (Collier County, FL, USA).



cells (Fig. 2). Only the 16 cells in the middle of the platewere considered for inoculation because the cells at theouter edges of the plate were more susceptible to drying.Five of the 16 interior cells were selected randomly forsowing using a computerized random number generator.Surface-disinfected seeds were sown onto the surface ofthe sterile germination medium using a sterile bacterialinoculating loop. Plates were sealed with NescoFilm(Karlan Research Products, Santa Rosa, CA, USA) andincubated under 12/12 h light/dark (60 mmol/m2/s) pho-toperiod at 25 � 3°C. Approximately 82 seeds were sownonto each plate (mean seeds/plate = 82, mean seeds percell = 16.4). Five replicate plates were used for each breed-ing system treatment. Seed germination and protocormdevelopment stage percentages were recorded weekly for10 weeks. Seedling development was scored on a scale of1–5 (Table 2; modified from Stewart et al. 2003). Germina-

tion percentages were calculated by dividing the numberof seeds in each germination and development stage bythe total number of viable seeds in the subsample. Datawere analyzed using general linear model procedures andWaller–Duncan at a = 0.05 with SAS v. 9.1. Percentage

Fig. 2 Seed viability and asymbiotic germination. (a) Germinating seeds (stage 2). (b) Protocorm at stage 4 of development showing twoleaves. (c) Stained and unstained seeds from tetrazolium. (d) Falcon ‘Integrid’ Petri dishes showing cells.

Table 2 Seed developmental stages of Cyrtopodium punctatum(modified from Stewart et al. 2003)

Stage Description

1 Intact testa2 Embryo enlarged, testa ruptured (= germination)3 Appearance of protomeristem4 Emergence of two first leaf primordia5 Elongation of shoot and further development

96 D . D U T R A E T A L .


germination counts were arcsine transformed to normal-ize variation.

Pollinator identification

Pollinator observations were conducted during the 2007and 2008 flowering seasons on a population consisting of15 plants. Observations took place on plants in the popu-lation that had the largest number of inflorescencesduring both flowering seasons (12 and 14 inflorescencesin 2007 and 2008, respectively). Observations took placefrom 07.00 to 18.00 hours for the first 2 days (13 and 14March 2007). When peak visitation time was identified,observations took place from 10:00 am to 4:00 pm for theremaining days (15, 23–25 March 2007). In 2008, observa-tions took place on 18–19 March (10.00–15.00 hours) and26 March (11.00–15.00 hours). Flower visitors were pho-tographed, collected and identified, and their behavior inthe flowers was documented. Flower visitors capable ofpollinating the flowers were determined either by directobservation or by the detection of pollinaria on theirbodies. The insects collected were identified at the FloridaState Collection of Arthropods (Gainesville, FL, USA).

Fragrance analysis

Two inflorescences with 30 open flowers each from two C.punctatum plants were collected from the FPNWR duringthe 2007 flowering season. The inflorescences were col-lected while still attached to the pseudobulbs to preservethe flowers and taken to the US Department of Agricul-ture Chemistry Research Unit (Gainesville, FL, USA).The inflorescences were placed into a 15 cm diameter ¥40 cm tall glass volatile collection chamber. Volatileswere collected over three time periods: 9:00–12:00 am,12:00 am–5:00 pm and 5:00 pm–7:00 am. The volatile trapswere extracted with 150 mL methylene chloride. Gaschromatography-mass spectroscopy (GC/MS) analyses ofthe collected volatiles were carried out on an HP-6890 gaschromatograph coupled to an HP5973 mass spectrometer(Agilent, Santa Clara, CA, USA). The GC/MS peaks ofinterest were identified by comparing their mass spectraldata with data from three mass spectral libraries. Onlytwo inflorescences could be used because very limitedplant material from wild populations is available owing tothe endangered status of the species.

Results

Breeding system determination

Capsule formation occurred in four out of the seven treat-ments tested during both the 2006–2007 and 2007–2008seasons (Figs 1c,3). There was no evidence that the agamo-

spermy, spontaneous autogamy or open pollination(control) treatments resulted in capsule formation (Fig. 3).The induced geitonogamy (pollen from the same flower)and artificial geitonogamy (pollen from the same plant, butfrom different flowers) treatments resulted in the produc-tion of significantly fewer capsules (14.6 and 27.1% capsuleformation, respectively) than the artificial xenogamy(pollen from a different population) and inducedxenogamy (pollen from a different plant in the same popu-lation) treatments (48.9 and 73.9%, respectively) (Fig. 3).

Capsule widths and lengths were significantly differentamong the pollination treatments (Figs 4,5). During the2006–2007 growing season, the length of the capsulesformed following induced autogamy was significantlyless than the length of the capsules produced followingthe other treatments 240 and 300 days after pollination(Fig. 5). In the 2007–2008 season, the capsule length ofcapsules formed following both the induced autogamy(pollen from the same flower) and artificial geitonogamytreatments (pollen from the same plant, different flower)were significantly less than the length of capsules formedin the other treatments (Fig. 5). In addition, in the 2007–2008 season, the width of the capsules formed from theinduced autogamy treatment was significantly less thanthat in all other treatments throughout the year (Fig. 4).

Fig. 3 Percentage capsule set among seven breeding systemtreatments. Data from 2006 and 2007. Six plants and 30 flowersper treatment were used per year. Percentages with the sameletter are not significantly different. Control (no treatment, openpollination); agamospermy (emasculated flower, no pollination);spontaneous autogamy (no treatment, pollen from the sameflower); induced autogamy (selfing with pollen from the sameflower); artificial geitonogamy (selfing with pollen from thesame plant, different flower); artificial xenogamy (pollen fromdifferent population); induced xenogamy (pollen from a plant inthe same population).



Viability and asymbiotic seed germination tests

We found no differences among treatments for either seedviability or asymbiotic seed germination. In general, seedgermination was very low (induced autogamy, 2.9%; arti-ficial geitonomy, 1.9%; artificial xenogamy, 4.3%; inducedxenogamy, 4.7%). This was unexpected considering thehigh seed viability observed across treatments thatresulted in seed production, particularly induced auto-gamy (79.9%), artificial geitonogamy (67.6%), artificialxenogamy (87.2%) and induced xenogamy (79.1%;Table 3). However, capsules that originated from selfing(induced autogamy and artificial geitonogamy) had sig-nificantly lower seed germination (2.9 and 1.9%, respec-tively) than seeds that originated from outcrossing(induced xenogamy, 4.7%).


During the two flowering seasons (2007 and 2008), fourbee species were observed visiting the flowers of C. punc-tatum (Fig. 6). Apis mellifera was the most frequent visitorin 2007 (53 visits in 46.5 h; Table 4). However, A. mellifera

and Megachile xylocopoides were only observed during the2007 flowering season (Table 4). Both Xylocopa micans andXylocopa virginica were frequent visitors to the flowersduring both flowering seasons (Table 4). Flowers wereobserved without pollinia; however, no visitors wereobserved carrying pollen to or from flowers.

Fragrance analysis

Nine fragrance compounds were identified as beingproduced by C. punctatum flowers: benzaldehyde,myrcene, benzyl alcohol, Z-b-ocimene, E-b-ocimene,1,3,8-para-menthatriene, methyl salicylate, indole andE,E-a-farnesene. The relative abundance of these com-pounds was recorded (Fig. 7).

Discussion

Breeding system

Our results show that C. punctatum is not capable of repro-ducing autogamously and requires a pollinator for

Fig. 4 Effect of breeding system treatment on mean capsulewidth (� standard error). The interaction between the 2006–2007and 2007–2008 flowering seasons was significant.

Fig. 5 Effect of breeding system treatment on mean capsulelength (standard error). The interaction between the 2006–2007and 2007–2008 flowering seasons was significant.

98 D . D U T R A E T A L .


capsule set. Given that no capsules were formed fromthe agamospermy or spontaneous autogamy treatmentsimplemented in the field, we demonstrated that no spon-taneous self-fertilization occurs. Although a few capsuleswere formed from induced autogamy (selfing with pollenfrom the same flower; 17.2%) and artificial geitonogamy(selfing with pollen from the same plant, but from a dif-

ferent flower, 27.2%), significantly more capsules wereproduced following artificial and induced xenogamy(48.9 and 73.9%, respectively; Fig. 3). Species within theOrchidaceae are predominantly self-compatible; however,species normally require an insect vector to facilitatepollen movement (Dressler 1981). This is supported in C.punctatum; spontaneous autogamy treatments resulted in

Table 3 Seed viability, germination and development stage percentages generated from Cyrtopodium punctatum pollination treatments inthe 2006–2007 flowering season

Pollination treatmentSeed

viability (%)Total

germinationSeedling development stages (%)†

1 2 3 4

Control‡ 0.0 0.0 0.0 0.0 0.0 0.0Agamospermy‡ 0.0 0.0 0.0 0.0 0.0 0.0Spontaneous autogamy‡ 0.0 0.0 0.0 0.0 0.0 0.0Induced autogamy 79.9 ab 2.9 bc 17.9 ab 1.2 b 1.7 a 0.35 aArtificial geitonogamy 67.6 b 1.9 c 18.1 a 1.2 b 0.49 b 0.44 aArtificial xenogamy 87.2 a 4.3 ab 17.6 b 2.7 a 1.9 a 0.55 aInduced xenogamy 79.1 ab 4.7 a 17.5 b 4.0 a 1.1 ab 0.29 a

† See Table 2 for descriptions of the seedling developmental stages. ‡ Treatments produced no capsules. Six plants were used and 30flowers were used per treatment. Values with the same letter are not significantly different (a = 0.05).

Fig. 6 Bees observed visiting flowers. (a) Apis mellifera (honey bee). (b) Xylocopa virginica. (c) Xylocopa micans (female). (d) Megachilexylocopoides. (e) X. micans (male).



no capsule formation, but both induced autogamy andartificial geitonogamy treatments, where pollen wasmanually moved to simulate the action of the pollinator,resulted in capsule formation.

Other studies with epiphytic orchid species show adegree of self-incompatibility and the need for pollinators(Ackerman 1989; Jaimes & Ramírez 1999; Borba et al. 2001;Lehnebach & Robertson 2004). Lehnebach and Robertson(2004) and Borba et al. (2001) reported that capsules werenot produced following agamospermy and spontaneousautogamy treatments on the epiphytic orchid species thatthey examined, showing dependence on a pollinator.Some degree of self-incompatibility was also found forthese species (Borba et al. 2001; Lehnebach & Robertson2004). Significantly fewer capsules were formed from arti-ficial selfing than from induced outcrossing. Similarly,some degree of self-incompatibility exists in C. punctatumbecause fewer capsules were formed from selfing thanoutcrossing treatments.

Cyrtopodium punctatum uses deceit via aromatic com-pounds and visual signals to attract insect visitors. Ourfragrance analysis and compound identification showed afloral bouquet strategy in which a diverse array of attrac-tant compounds are produced. However, C. punctatum

flowers provide no food reward to floral visitors. Thereare implications of having a deceit pollination strategy.Neiland and Wilcock (1998) found that orchids that offerfood rewards (e.g. nectar) had double the probability ofsetting capsule than non-rewarding species across all geo-graphical areas. In North America, the rate of capsuleset was 19.5% for species that offer no nectar rewardcompared with 49.3% for rewarding orchids (Neiland &Wilcock 1998). A closely related species in Brazil, Cyrtopo-dium polyphyllum, also uses a deceit pollination system.This plant mimics the yellow flowers of a reward-producing unrelated species that occurs in the samehabitat. This species also has low capsule set as a result oflow pollinator visitation (Pansarin et al. 2008). Pollinatorlimitation has also been observed in other deceptiveorchids, such as Serapias vomeracea and Pogonia japonica(Matsui et al. 2001; Pellegrino et al. 2005).

Historically, populations of C. punctatum in Floridawere numerous and plants were abundant. In this situa-tion, a deceit pollination system with low capsule produc-tion was viable because the probability of capsuleformation was higher owing to plant abundance. How-ever, populations are now small and fragmented; thus,the deceit pollination system may minimize future repro-ductive success for the species as the likelihood of cap-sules being produced is so low. The open pollinationtreatment resulted in no capsule formation during boththe 2006–2007 and 2007–2008 growing seasons in thepopulation studied. However, two capsules were formedon two different plants during the 2006–2007 season ininflorescences not used in the breeding system study, sug-gesting that natural pollination does occur, albeit at verylow levels. The deceit pollination system alone may not bethe cause of the low capsule formation at the FPNWR. Adecrease in pollinator populations may also be affectingnatural capsule set. A more detailed study on the insect

Fig. 7 Fragrance analysis. (a) Relativeamounts of fragrance compoundsextracted from Cyrtopodium punctatuminflorescences.

Table 4 Insect visitors to Cyrtopodium punctatum

Insect visitors 2007 2008

Xylocopa micans 28 17Xylocopa virginica 15 2Apis melifera 53 0Megachile xylocopoides 2 0

Total number of observation hours for 2007 and 2008 were46.5 h and 14.2 h, respectively.

100 D . D U T R A E T A L .


populations of the area should be conducted, particularlybecause of the closeness of the populations to agriculturalfields that continually use pesticide applications.

The capsule size measurements (length and width)taken during the 11-month maturation period in bothyears indicated that the induced autogamy and artificialgeitogamy treatments resulted in capsules that wererelatively smaller than capsules produced from othertreatments (except for capsule length in 2006–2007).However, capsule size may vary from year to year owingto significant variation in environmental factors, such asrainfall and temperature, which can influence capsule for-mation and maturation over time. Our research suggeststhat breeding system studies should be conducted acrossmultiple growing seasons.

Seed viability and asymbiotic germination

Tetrazolium seed viability testing showed high seedviability in all breeding system treatments that formedcapsules in C. punctatum. However, germination was verylow across all treatments. This may be because of subop-timal in vitro cultural conditions for asymbiotic germina-tion. Light may inhibit germination of C. punctatum seeds(Dutra et al. 2009). By sowing seeds and initially incubat-ing them in the dark, germination could be significantlyimproved in this species.


During the 2007 flowering season, honey bees (Apismelifera) were the most common visitors to C. punctatumflowers. However, these bees were too small to removepollinaria and efficiently act as pollinators. In the 2008flowering season, honey bees were not seen visiting C.punctatum flowers. Xylocopa bees (X. micans and X. vir-ginica) were observed visiting the flowers, but removalor deposition of pollinia was not observed during thefield observation period, despite being large enough tofit inside the flowers. In Puerto Rico, C. punctatum isreported to be pollinated by Xylocopa bees (Ackerman1995). Pemberton and Liu (2008) observed two species ofCentris bees, visiting the flowers of C. punctatum in south-east Florida. However, these pollinator observations wereconducted at Fairchild Tropical Botanical Garden and inFort Lauderdale, Florida, in an artificial setting withplants of unknown origins. Observations conducted in agarden setting do not accurately reflect the real ecologicallinks because a mixture of exotic and native plants areplanted in unnatural arrangements causing insect pollina-tor densities to change. Pollination observations con-ducted in situ are more representative because it is wherenatural conditions prevail and where future plant reintro-ductions will take place.

Fragrance analysis

The compounds identified in the fragrance analysis of C.punctatum are considered relatively common pollinatorattractants produced by many other orchid species (Kaiser1993). However, two compounds identified, indole andmethyl salicylate, can be associated with the Euglossinipollination system (Williams & Whitten 1983). AlthoughFlorida is located outside the range of Euglossini bees, C.punctatum is reported to be pollinated by Euglossa bees inpart of its range (van der Pijl & Dodson 1966; Jeffrey et al.1970; Luer 1972; Dressler 1993). An introduced species ofEuglossa (E. viridissima) has recently become naturalized insouthern Florida (Skov & Wiley 2005). Pemberton andWheeler (2006) extracted and identified compounds col-lected by E. viridissima in southern Florida. Althoughmethyl salicylate and indole have been detected in manyother Euglossini species in the tropics (Ramírez et al.2002), these compounds were not found in the collectionstorage organs of E. viridissima in southern Florida (seePemberton & Wheeler 2006). It is possible that even if E.viridissima was found in the same location as C. punctatum,it would not be attracted by the types of compounds thatthe plant produces and would not affect pollination.

Conservation implications

Understanding a plant’s breeding system and pollinatordiversity can help conservation and management strate-gies because plant populations may be greatly impactedby a reduction in pollinator visitations (Kearns et al. 1998;Havens 1999). For example, Sahli and Conner (2006)found that pollinator visitation and not pollinator effec-tiveness is the main driving factor in determining pollina-tion importance. In a generalized pollination system, suchas that in C. punctatum, the pollinators that visit the plantsthe most are most likely to be the ones responsible forcapsule formation and not simply infrequent visitors.Xylocopa bees visited the plants during both seasons andwere responsible for most of the visits during 2008.Although A. melifera was the most frequent visitor in 2007,it did not visit the plants during the 2008 season.

Observations made during the present study and bybiologists over the past 10 years at the FPNWR indicatethat very few capsules were formed by C. punctatum. Pes-ticide use in nearby agricultural areas is another possibledirect cause of low capsule formation as pesticide use maydecrease pollinator populations at the study site. Similarly,habitat degradation may affect insect populations. Thearea has been impacted from hydrological changes thathave shortened the hydroperiod. Local insect populationdynamics should be studied for conservation purposesand land management in areas were C. punctatum occurs.

Our study indicates that sexual reproduction in C. punc-tatum is severely depressed, suggesting that recruitment is



highly unlikely. Consequently, the long-term viability/persistence of the existing population is at risk. Inresponse to a likely need for a reintroduction program,additional studies aimed at determining the populationgenetic diversity and structure of the existing populationsand developing efficient asymbiotic germination proce-dures for reintroduction are underway. Although geneticdiversity studies are being conducted, manual pollinationshould be conducted within populations to ensure thatseeds are being produced for in situ recruitment and exsitu propagation for future reintroductions.

Pemberton and Liu (2008) suggested that restoration ofC. punctatum could be aided by planting Brysonima lucida,a rare species in the Malpighiaceae that is restricted tosouthern Florida, which attracts Centris bees. However, B.lucida is not naturally found growing in the big cypressregion where natural populations of C. punctatum arefound (Wunderlin & Hansen 2004). We suggest that localrecovery plans for the species be developed given thehigh habitat variation (e.g. between Everglades NationalPark vs FPNWR) and fragmentation found in Florida. Rec-ommendations should not be made solely from the rela-tively few plants in cultivation and should be based onin-depth studies that will likely affect the survival of thespecies in the wild.

Acknowledgments

We thank Hans Alborn (USDA) for his help with thefragrance analysis, Jim Wiley (Florida Department of Agri-culture and Consumer Services) for his help with insectidentification and the Florida Panther National WildlifeRefuge staff for logistical support in the field. Specialthanks to Nancy Philman, Timothy Johnson, Scott Stewartand Philip Kauth. This project was supported by both theUS Fish and Wildlife Service and the American OrchidSociety.

References

Ackerman J. D. (1989) Limitations to sexual reproduction inEncyclia krugii (Orchidaceae). Systematic Botany 14: 101–109.

Ackerman J. D. (1995) An Orchid Flora of Puerto Rico and the VirginIslands. The New York Botanical Garden, Bronx.

Ames O. (1904) A Contribution to Our Knowledge of the Orchid Floraof Southern Florida. Contributions from the Ames BotanicalLaboratory I. E. W. Wheeler, Cambridge.

Borba E. L., Semir J. & Shepherd G. J. (2001) Self-incompatibility,inbreeding depression and crossing potential in five BrazilianPleurothallis (Orchidaceae) species. Annals of Botany 88: 89–99.

Bowles M. L., Jacobs K. A., Zettler L. W. & Wilson Delaney T.(2002) Crossing effects on seed viability and experimentalgermination of the Federal threatened Platanthera leucophaea(Orchidaceae). Rhodora 104: 14–30.

Brown P. M. (2002) Wild Orchids of Florida. University Press ofFlorida, Gainesville.

Catling P. M. (1987) Notes on the breeding systems of Sacoilalanceolata (Aublet) Garay (Orchidaceae). Annals of the MissouriBotanical Garden 74: 58–68.

Chase M. W. & Hills H. G. (1992) Orchid phylogeny,flower sexuality, and fragrance seeking. BioScience 42: 43–49.

Coile N. & Garland M. A. (2003) Notes on Florida’s Endangered andRare Plants Page, 4th edn. Botany Section No. 38. Bureau ofEntomology, Nematology and Plant Pathology, FloridaDepartment of Agriculture and Consumer Services, Divisionof Plant Industry, Gainesville.

Dixon K. W., Kell S. P., Barrett R. L. & Cribb P. J. (2003) OrchidConservation. Natural History Publications, Kota Kinabalu.

Dressler R. L. (1981) Orchids Natural History and Classification.Harvard University Press, Cambridge.

Dressler R. L. (1993) Phylogeny and Classification of the OrchidFamily. Cambridge University Press, Cambridge.

Dutra D., Kane M. E. & Richardson L. (2009) Asymbiotic seedgermination and in vitro seedling development of Cyrtopo-dium punctatum: a propagation protocol in an endangeredFlorida native orchid. Plant Cell Tissue and Organ Culture 96:235–243.

Fiedler P. L. & Ahouse J. J. (1992) Hierarchies of cause: toward anunderstanding of rarity in vascular plant species. In: Fiedler P.L. & Jain S. K. (eds). Conservation Biology: the Theory and Prac-tice of Nature Conservation, Preservation, and Management.Chapman and Hall, New York, pp. 23–47.

Hamrick J. L. & Godt M. J. W. (1996) Effects of life history traits ongenetic diversity in plant species. Philosophical Transactions ofthe Royal Society in London 351: 1291–1298.

Havens K. (1999) Pollination biology: implications for restoringrare plants. Ecological Restoration 17: 216–218.

Hills H. G., Williams N. H. & Dodson C. H. (1972) Floral fra-grances and isolating mechanisms in the genus Catasetum(Orchidaceae). Biotropica 4: 61–76.

Jaimes I. & Ramírez N. (1999) Breeding systems in a secondarydeciduous forest in Venezuela: The importance of life form,habitat, and pollination specificity. Plant Systematics and Evo-lution 215: 23–36.

Jeffrey D. C., Arditti J. & Koopowitz H. (1970) Sugar content infloral and extrafloral exudates of orchids: pollination, myrme-cology and chemotaxonomy implication. New Phytologist 69:187–195.

Kaiser R. (1993) The Scent of Orchids: Olfactory and Chemical Inves-tigations. Elsevier, Amsterdam.

Kearns C., Inouye D. & Waser N. (1998) Endangered mutualisms:the conservation of plant–pollinator interactions. AnnualReview of Ecology, Evolution and Systematics 29: 83–112.

Lakon G. (1949) The topographical tetrazolium method for deter-mining the germination capacity of seeds. Plant Physiology 24:389–394.

Lauzer D., St-Arnaud M. & Barabe D. (1994) Tetrazolium stainingand in vitro germination of mature seeds of Cypripedium acaule(Orchidaceae). Lindleyana 9: 197–204.

Lehnebach C. A. & Robertson A. W. (2004) Pollination ecology offour epiphytic orchids of New Zealand. Annals of Botany 93:773–781.

Little R. J. (1983) A review of floral food deception mimicrieswith comments on floral mutualism. In: Jones C. E. & Little R.

102 D . D U T R A E T A L .


J. (eds). Handbook of Experimental Pollination Biology. Van Nos-trand Reinhold, New York, pp. 294–309.

Luer C. A. (1972) The Native Orchids of Florida. New York BotanicalGarden, New York.

Matsui K., Ushimaru T. & Fujita N. (2001) Pollinator limitation ina deceptive orchid, Pogonia japonica, on a floating peat mat.Plant Species Biology 16: 231–235.

Neiland M. R. M. & Wilcock C. C. (1998) Fruit set, nectar reward,and rarity in the Orchidaceae. American Journal of Botany 85:1657–1671.

Pansarin L. M., Pansarin E. R. & Sazima M. (2008) Reproductivebiology of Cyrtopodium polyphyllum (Orchidaceae): a Cyrtopo-diinae pollinated by deceit. Plant Biology 10: 650–659.

Partel M., Kalamees R., Reier U., Tuvi E. L., Roosaluste E., VellakA. & Zobel M. (2005) Grouping and prioritization of vascularplant species for conservation: combining natural rarity andmanagement need. Biological Conservation 123: 271–278.

Pellegrino G., Gargano D., Noce M. E. & Musacchio A. (2005)Reproductive biology and pollinator limitation in a deceptiveorchid, Serapias vomeracea (Orchidaceae). Plant Species Biology20: 33–39.

Pemberton R. W. & Liu H. (2008) Potential of invasive and nativesolitary specialist bee pollinators to help restore the rarecowhorn orchid (Cyrtopodium punctatum) in Florida. BiologicalConservation 141: 1758–1764.

Pemberton R. W. & Wheeler G. S. (2006) Orchid bees don’t needorchids: Evidence from the naturalization of an orchid bee inFlorida. Ecology 87: 1995–2001.

Rabinowitz D. (1981) Seven forms of rarity. In: Synge H. (ed.). TheBiological Aspects of Rare Plant Conservation. Wiley, New York,pp. 205–217.

Ramírez S., Dressler R. & Ospina M. (2002) Abejas euglosinas(Hymenoptera: Apidae) de la Región Neotropical: listado deespecies con notas sobre su biología. Biota Colombiana 1: 7–118.

Robertson J. L. & Wyatt R. (1990a) Evidence for pollinationecotypes in the yellow-fringed orchid, Platanthera ciliaris. Evo-lution 44: 121–133.

Robertson J. L. & Wyatt R. (1990b) Reproductive biology of theyellow-fringed orchid, Platanthera ciliaris. American Journal ofBotany 77: 388–398.

Romero-González G. A. & Fernández-Concha G. C. (1999) Noteson the species of Cyrtopodium (Cyrtopodinae, Orchidaceae)from Florida, Greater Antilles, Mexico, Central and northernSouth America. Harvard Papers in Botany 4: 327–341.

SAS Institute (2003) SAS for Windows, version 9.1. SAS Institute,Cary.

Sahli H. F. & Conner J. K. (2006) Characterizing ecologicalgeneralization in plant–pollination systems. Oecologia 148:365–372.

Singh F. (1981) Differential staining of orchid seeds for viabilitytesting. American Orchid Society Bulletin 50: 416–418.

Sipes S. D. & Tepedino V. J. (1995) Reproductive biology of therare orchid, Spiranthes diluvialis: breeding system, pollination,and implications for conservation. Conservation Biology 9: 929–938.

Skov C. & Wiley J. (2005) Establishment of the Neotropical orchidbee Euglossa viridissima (Hymenoptera: Apidae) in Florida.Florida Entomologist 88: 225–227.

Stewart S. L. (2007) Integrated conservation of Florida Orchi-daceae in the genera Habenaria and Spiranthes model orchidconservation systems for the Americas (PhD dissertation).University of Florida, Gainesville, FL.

Stewart S. L., Zettler L. W., Minso J. & Brown P. M. (2003) Sym-biotic germination and reintroduction of Spiranthes brevilabrisLindley, an endangered orchid native to Florida. Selbyana 24:64–70.

van der Cingel N. A. (1995) An Atlas of Orchid Pollination: Euro-pean orchids. Balkema Publishers, Rotterdam.

van der Pijl L. & Dodson C. H. (1966) Orchid Flowers: TheirPollination and Evolution. University of Miami Press, CoralGables.

Van Waes J. M. & Debergh P. C. (1986) In vitro germination ofsome Western European orchids. Physiologia Plantarum 67:253–261.

Vujanovic V. St, Arnaud M., Barabe D. & Thibeault G.(2000) Viability testing of orchid seed and the promotionof colouration and germination. Annals of Botany 86: 79–86.

Williams N. H. & Whitten W. M. (1983) Orchid floral fragrancesand male Euglossine bees: methods and advances in the lastsesquidecade. Biological Bulletin 164: 355–395.

Wong K. C. & Sun M. (1999) Reproductive biology and conser-vation genetics of Goodyera procera (Orchidaceae). AmericanJournal of Botany 86: 1406–1413.

Wunderlin R. P. & Hansen B. F. (2004) Atlas of Florida VascularPlants. Institute for Systematic Botany, University of SouthFlorida, Tampa. [Cited 15 June 2009.] Available from URL:http://www.plantatlas.usf.edu/



Reproductive biology of Cyrtopodium punctatum in situ...

Documents

Transcript of Reproductive biology of Cyrtopodium punctatum in situ...