sign in sign up
<<
Home , Entomophily, Myristica, Myristicaceae, Myristica fragrans

Floral Biology of insipida (), a Distinctive Syndrome Author(s): Joseph E. Armstrong and Anthony K. Irvine Source: American Journal of , Vol. 76, No. 1 (Jan., 1989), pp. 86-94 Published by: Botanical Society of America, Inc. Stable URL: http://www.jstor.org/stable/2444777 Accessed: 06-09-2016 16:15 UTC

REFERENCES Linked references are available on JSTOR for this article: http://www.jstor.org/stable/2444777?seq=1&cid=pdf-reference#references_tab_contents You may need to log in to JSTOR to access the linked references.

JSTOR is a not-for-profit service that helps scholars, researchers, and students discover, use, and build upon a wide range of content in a trusted digital archive. We use information technology and tools to increase productivity and facilitate new forms of scholarship. For more information about JSTOR, please contact [email protected].

Your use of the JSTOR archive indicates your acceptance of the Terms & Conditions of Use, available at http://about.jstor.org/terms

Botanical Society of America, Inc. is collaborating with JSTOR to digitize, preserve and extend access to American Journal of Botany

This content downloaded from 132.198.8.49 on Tue, 06 Sep 2016 16:15:01 UTC All use subject to http://about.jstor.org/terms Amer. J. Bot. 76(1): 86-94. 1989.

FLORAL BIOLOGY OF MYRISTICA INSIPIDA (MYRISTICACEAE), A DISTINCTIVE BEETLE '

JOSEPH E. ARMSTRONG AND ANTHONY K. IRVINE

Department of Biological Sciences, Illinois State University, Normal, Illinois 61761; and CSIRO Tropical Forest Research Centre, PO Box 780, Atherton, Queensland 4883, Australia

ABSTRACT

The floral biology and pollination of Myristica insipida were studied in two different rain forest communities in Queensland. Floral morphology of M. insipida resembles that of M. fragrans in virtually all respects. The majority of female were receptive 48-72 hr. Male flowers were shorter-lived, functional for 12-48 hr. Both male and female flowers opened daily between 1800 and 2200 hr, but the activity of floral visitors did not begin until the next morning. The inconspicuous, creamy-white to light-green flowers had a strong, pleasant "floral" fragrance. The male flowers only offered as a reward, and female flowers, offering no reward, were judged to function by automimicry. A taxonomically diverse array of small, pollen-foraging were the effective , although thrips and were common floral visitors. In almost all respects, the beetle pollination syndrome of Myristica differs from the cantharophily of most other primitive angiosperms.

WITHIN THE there exists a di- represent a highly specialized pollination syn- versity of floral forms and pollination syn- drome involving constancy and re- dromes, although beetle pollination, canthar- wards of food, predator protection, breeding ophily, is common in many of these families site, and brood substrates (Gottsberger, 1974, (Gottsberger, 1974, 1977; Thien, 1974, 1980; 1977; Thien, 1974, 1980; Beach, 1982; Pell- Endress, 1986; Bernhardt and Thien, 1987). myr, 1984, 1985; Pellmyr and Thien, 1986; The small, inconspicuous, unisexual flowers of Tang, 1987). the Myristicaceae are not like the large, many- A study of cultivated indicated that parted, spirally-arranged flowers usually as- the male flowers only offered pollen as a re- sociated with beetle pollination, and thought ward, that the female flowers offered little or of as "primitive" and "typical" in the Mag- no reward, functioning by automimicry (Arm- noliales. , the nutmeg of strong and Drummond, 1986). The small, - commerce, has been reported to be pollinated tle pollinator did not engage in breeding activ- by small , , or possibly wind ities or destructive feeding on floral parts. This (Deinum, 1949; Mcllroy, 1967; Purseglove, is similar in many respects to the beetle pol- 1968; Cobley, 1976), but was found to be pol- lination reported for Drimys brasiliensis and linated by a small beetle (Armstrong and Diospyros pentamera (Gottsberger, 1977; Drummond, 1986). Gottsberger, Silberbauer-Gottsberger, and Eh- Cantharophily is a diverse pollination syn- rendorfer, 1980; House, 1985). There is some drome and even among primitive angiosperms chance that the flowers or flowering of culti- different aspects of beetle- interactions vated nutmeg may have been altered through have been identified (Gottsberger, 1977; Thien, conscious or unconcious selection during their 1980). Several studies have demonstrated that domestication. Further, in a plantation setting, the "primitive," magnoliacious flowers polli- pollinators and pollinator activity may differ nated by the haphazard, "mess-and-soil" ac- from those in a natural, rain forest community. tivities of beetles (sensu Faegri and Pijl, 1971), In order to corroborate and further document this unusual type of beetle pollination, we stud- ied a species of Myristica in its natural rain forest communities. Submitted for publication 29 June 1987; revision ac- Myristica insipida is a subcanopy that cepted 13 May 1988. occurs in several different rain forest com- The first author wishes to thank the staff of the CSIRO Tropical Forest Research Centre for their assistance and munities in Queensland. We studied the floral support. We are grateful for the discussions with P. Bern- biology and pollination of this nutmeg species hardt and 0. Pellmyr, and the comments of anonymous in two different rain forest communities, a low- reviewers, that contributed to this manuscript. J. Bouse- land study site at Little Pine Creek (LPC) and man, Illinois Natural History Survey, kindly provided family identifications of the beetles. This research was an upland study site at Curtain Fig State Forest supported by NSF Grant INT-85 13473. (CF), described in the preceding paper (Arm-

86

This content downloaded from 132.198.8.49 on Tue, 06 Sep 2016 16:15:01 UTC All use subject to http://about.jstor.org/terms January 1989] ARMSTRONG AND IRVINE-MYRISTICA FLORAL BIOLOGY 87 strong and Irvine, 1989). Specifically, our stud- , over 85% of each day's flowers open be- ies were designed to answer the following ques- tween 1800 and 2200 hr. The remaining flow- tions about floral function and pollinators: 1) ers reach anthesis during the remainder of the What are the flowering patterns and longevities night. No new flowers open from 400 to 1800 of the male and female flowers? 2) What are hr. the floral attractants and rewards? 3) What are Male flowers are relatively short-lived and the floral visitors and pollen vectors? The sea- show visible signs of senescence after 12 hr sonal flowering, sex ratios, spacing, and repro- with 40% abscising after 12-24 hr and the re- duction were studied also and the results re- maining 60% abscising between 24-48 hr (Ta- ported in the preceding paper. ble 1). Newly-opened male flowers have creamy-white to light-green perianths, and METHODS- Floral longevity was determined creamy-white, turgid androecia. After 12 hours by making daily observations of marked in- the androecia start to shrivel and turn a pink dividual flowers on conveniently located and to light-brown color. The pollen was viable accessible trees. Observations of the frequency, and could successfully pollinate female flowers duration, and diversity of floral visitors were throughout the life ofthe male flowers. In many made hourly over a 24-hr period on several cases flowers still borne on a tree largely were occasions. Changes in floral appearance, odor, depleted of pollen. Fresh male flowers produce presence or quantity of secretions and pollen a strong fragrance, which wanes as the flowers were noted. Visual changes in the female flow- senesce. However, even senescent and abscised ers were experimentally coordinated with stig- flowers produce some fragrance. We judged ma receptivity by pollinating flowers of known male flowers to be functional as long as they age and visual condition with a 200+ pollen contained pollen and/or remained on the tree. load from distant intrapopulation donor trees. The female flowers are functional longer than Floral visitors were captured and examined male flowers with 69.5% lasting longer than 48 microscopically for pollen loads. Voucher hr and a few lasting over 72 hr (Table 1). The specimens were sent to the CSIRO Division of perianths, pistils, and stigmas are colored a Entomology in Canberra. Visited flowers were light-green to greenish-white to creamy-white examined for evidence of damage or visitors' while fresh. Fresh stigmas always appear moist reproductive activities. Fresh, functional flow- but show no accumulation of liquid. The fe- ers were stained with 0.1 % neutral red for twen- male flowers are very fragrant when fresh and ty minutes and then rinsed with distilled water they remain very fragrant until senescence be- to identify any secretory tissues (Vogel, 1962). gins. The stigma and tips of the tepal lobes of postpollinated and senescent flowers turn a RESULTS-Floral morphology-The flowers dark, red-brown color (Fig. 4). The color change of Myristica insipida are identical to those of of the stigma was highly correlated with loss M. fragrans in essentially all respects (see Wil- of receptivity. Light-colored stigmas are re- son and Maculans, 1967; Armstrong and ceptive continuously until the beginning of col- Drummond, 1986; Armstrong and Tucker, or changes that mark the onset of senescence. 1986, for details of floral morphology). Both The stigmas of hand-pollinated flowers turned male and female flowers are borne on axillary dark within 12-24 hr post-pollination. inflorescences (Fig. 2) and have a 3-merous, syntepalous perianth subtended by a minute Floral attractants and rewards-Both male bracteole (Fig. 3-6). Both male and female and female flowers produce a similar fragrance flowers are small, although differing slightly in that is a slightly heavy, sweet, pleasant, "floral" size and shape. Female flowers were 4-5 mm odor. Individual female flowers seemed to be tall and 3-4 mm in diameter with a vase shape more strongly scented than individual stami- that was widest at the bottom (Fig. 3, 4). Each nate flowers, but because of the larger number has a single pistil containing a single of flowers (Armstrong and Irvine, 1989), male ovule and bearing a sessile, bilobed stigma. trees were more fragrant than female trees and Male flowers have a fairly uniform columnar could often be smelled several meters away. shape 5-7 mm tall and 2-3 mm in diameter We judge floral fragrance to be the primary (Fig. 5, 6). Each flower has a columnar recep- attractant because the small, dully-colored tacle upon which there are 6-9 laterally sessile flowers produce a very weak visual image (Fig. anthers. 1). No is produced by either male or fe- Flowering phenology and longevity-The male flowers. On several occasions receptive flowering of Myristica insipida follows a very female flowers were found with a liquid filling regular daily pattern. On both male and female the space between the pistil and the perianth,

This content downloaded from 132.198.8.49 on Tue, 06 Sep 2016 16:15:01 UTC All use subject to http://about.jstor.org/terms 88 AMERICAN JOURNAL OF BOTANY [Vol. 76

Fig. 1-6. Flowernag twigs, female flowers, and male flowers of Myristica insipida. 1. One of the authors (AKI) examining a male tree in full flower. This illustrates the visual impact of the inconspicuous flowers at a distance of 3 m. 2. Flowering twigs from a male tree showing several axillary inflorescences. Inflorescences on female trees are similar, but with fewer flowers. Bar = 1 cm. 3. Indeterminate axillary inflorescence showing one senescent (center) and two receptive, vase-shaped female flowers. Bar = 5 mm. 4. Apical view of receptive female flower showing stigma occupying the perianth opening. Stigma within senescent flower has turned a dark red-brown. Bar = 5 mm. 5. Indeterminate axillary inflorescence showing a functional, 12-hr-old male flower with androecial column exposed by the three recurved perianth lobes, and several columnar flower buds. Bar = 5 mm. 6. View into the perianth opening of a male flower showing the androecial column within. Bar = 5 mm.

but refractometer readings indicated no dis- lining the pollen locules took up the stain. The solved sugars. Open female flowers, no matter apex of the androecium, the androecial col- what their orientation, were found to take up umn, and perianth remained unstained. In the water by capillary action when hit by drops of absence of all other secretory functions, the water from a pipette. Checks with weather data stain-retaining tissues probably function as os- indicated that all observations of liquid in fe- mophores producing the floral fragrance. male flowers followed recent rainfalls. No ac- With no evidence of feeding on other floral cumulation of liquid was ever observed on a structures, pollen is the only apparent food stigma. reward offered to floral visitors of male flowers. The stigmas and tips of the tepal lobes in There was no evidence that floral visitors used female flowers retained the neutral red stain flowers as a protected breeding site or floral indicating the location of secretory tissue. In tissues as a brood substrate. Female flowers male flowers only the tissues surrounding and produce no perceivable reward. The perianths

This content downloaded from 132.198.8.49 on Tue, 06 Sep 2016 16:15:01 UTC All use subject to http://about.jstor.org/terms January 1989] ARMSTRONG AND IRVINE-MYRISTICA FLORAL BIOLOGY 89

TABLE 1. Longevity of male and female flowers

Duration (hr)

12-24 24-48 48-60 60-72 72+ Total

Male flowers N 45 66 0 0 0 111 % 40.5 59.5 0 0 0 100

Female flowers N 9 23 46 23 4 105 % 8.6 21.9 43.8 21.9 3.8 100

of female flowers closely encircle the stigma (Fig. 4) and restrict entry of most floral visitors.

There was no evidence of any feeding on stig- Fig. 7. Beetle visitors to flowers of Myristica insipida. matic or perianth tissues. Some of the floral From left to right; top row, Short-snouted, wide-eyed wee- visitors probed the moist stigma, but did not vil; Larger, short-snouted weevil; Narrow-waisted Niti- remain long. dulid beetle; Long-necked beetle; bottom row, Long-snout- ed, narrow-eyed weevil; Tortoise-shell weevil; Short, stubby Nitidulid beetle; Black Staphylinid beetle. Scale shows 1 Floral visitors-Newly-opened flowers are mm divisions. functional all night, but remain unvisited until the following dawn when the activity of floral visitors commences. During the day several species of small beetles (2-3 mm in length) visit Only one species of floral visitor, the most the flowers (Fig. 7; Table 2) with peak activity common weevil (short-snouted, wide-eyed), in the late morning. Although weevils (Cur- was observed visiting a female flower with a culionidae) were the most frequent Coleopter- pollen load (N = 10; Table 2). an floral visitor, beetles in Nitidulidae and Sta- Very small ants frequently were observed phylinidae were also common floral visitors. moving along branches to visit female flowers, There were no differences in visitation pat- but they were not observed visiting male flow- terns, seasonal or daily, among the beetle floral ers nor did they ever have pollen loads. Larger visitors. Several different beetle species could tree ants, black and green, were found on sev- be found visiting a tree simultaneously. The eral nutmeg trees, and while they would defend Coleopteran floral visitors were more taxo- their tree against invading botanists, they were nomically diverse at the upland Curtain Fig not floral visitors. (CF) study site than at the coastal Little Pine Very small thrips were the most frequently Creek (LPC) study site (Table 2). observed floral visitors, particularly in male With the exception of the Staphylinid beetle, flowers, where they would spend considerable all ofthese Coleopteran floral visitors are strong, time within the flowers, often for the duration rapid flyers. Beetles' visits to flowers are sol- of the entire observational period lasting sev- itary and relatively brief. Beetles visiting male eral hours. Other than the ever-present thrips, flowers stayed 30 sec to several minutes, unless we observed no nocturnal visits to nutmeg disturbed. They orient their ventral surface flowers. Thrips removed from male flowers had along the long axis of the androecial column an average pollen load of 2.1 grains, but thrips parallel to the microsporangia placing their captured from female flowers (N = 18; Table mouth parts in a position to feed on pollen 2) never had a pollen load. The thrips did not grains. The weevils all have relatively short demonstrate any movements or behavior that snouts (Fig. 7). Visits to female flowers were would suggest that they are capable of regular very brief, rarely lasting more than 10-15 sec. inter-tree movements. Beetle visitors to female flowers stand astride Dipteran larvae were found within several the stigma and probe the stigmatic cleft or the flowers of both sexes. Male flowers that con- space between the perianth and the pistil before tained these larvae were fluid-filled, generally taking wing and leaving. devoid of pollen, and usually senesced in less Most of the beetles were captured just before than 24 hr. The female flowers were also fluid- or after visiting male flowers. The beetles had filled, but there was no significant damage to an average pollen load of 5.4 grains (Table 2), the pistil and the stigma appeared receptive. which was carried mostly on their ventral side. The light trichome tomentum on the pistil was

This content downloaded from 132.198.8.49 on Tue, 06 Sep 2016 16:15:01 UTC All use subject to http://about.jstor.org/terms 90 AMERICAN JOURNAL OF BOTANY [Vol. 76

TABLE 2. Floral visitors captured at Curtain Fig (CF) and Little Pine Creek (LPC) study sites and their average pollen loads (APL)

Number captured

Visitors CF LPC Total APL

Coleoptera Curculionidae Short-snouted, wide-eyed weevil 30 6 la 91 3.6 Larger, short-snouted weevil 11 - 1 c 21.8 Long-snouted, narrow-eyed weevil 8 7 15c 12.2 Tortoise-shell weevil 8 - 8c 9.7

Nitidulidae Narrow-waisted beetle 10 1 Oc 4.6 Short, stubby beetle 5 - 5c 6.8

Staphylinidae Black beetle 8 23b 31 2.4

Undetermined family Long-necked beeetle 1 - Ic 3.0

Totals 81 91 172 5.4

Hymenoptera Small - 24 24e 0.0

Thysanoptera Thripsd 15 27 42c 2.1 - 18 18e 0.0

a N = 4 captured from female flowers, APL = 1.0. b N = 6 captured from female flowers, APL = 0.0. c Captured from male flowers only. d Most frequent floral visitors, 50-70% of male flowers. e Captured from female flowers only.

usually gone and the pistil often showed shal- sects have been reported as floral visitors of low, superficial damage. The larvae fall to the Neotropical Myristicaceae (Bawa et al., 1985), ground within the senescent flowers. A small and while undoubtably accurate, a similarly dipteran was observed atop mature flower buds, conducted survey would probably yield the perhaps in the process of oviposition. We same report for Myristica fragrans and M. in- judged this dipteran to be a floral parasite. sipida. The lack of pollinator diversity ob- served in a plantation of nutmeg in India (Arm- Visitation rates-Over 95% of all recorded strong and Drummond, 1986) could be the visits to the flowers of M. insipida were result of a loss of rain forest habitat to support to male flowers. Based on observed diurnal a diverse population of pollen-foraging beetles. visits per flower-hr (1 flower observed for 1 hr In most respects the floral biology and pol- equals 1 flower-hr), beetles were over four times lination syndrome of Myristica insipida is as likely to visit a male flower (0.056 visits/ identical to that of M.fragrans (Armstrong and flower-hr) as a female flower (0.013 visits/flow- Drummond, 1986). One significant difference er-hr). Therefore, a female flower of average is that M. insipida flowers nocturnally, exactly longevity (60 hr) would be expected to receive like M. fragrans, but has only diurnal floral 0.39 visits during its functional life at the ob- visitors. The Anthicid beetle pollinator of M. served rate of diurnal visitation. A male flower fragrans is completely nocturnal. It seems that of average longevity (36 hr) would be expected M. insipida has retained nocturnal flowering to receive one visit during its functional life. even though there is no nocturnal insect fauna in Queensland to act as pollinators. Both male DISCUSSION-Floral biology-The flowers of and female flowers of M. insipida function Myristica insipida attract ants, thrips, and a somewhat longer than those of M. fragrans, variety of beetles. While the most common and which probably represents an adaptation to easily observed floral visitors are thrips, only better accommodate the delayed diurnal ac- the beetles possess the characteristics of an ef- tivity of floral visitors. The physiological re- fective pollinator. Thrips and diverse small in- sponse necessary to produce a precise start of

This content downloaded from 132.198.8.49 on Tue, 06 Sep 2016 16:15:01 UTC All use subject to http://about.jstor.org/terms January 1989] ARMSTRONG AND IRVINE-MYRISTICA FLORAL BIOLOGY 91 anthesis can be considered relatively refractory of the tree's display, the reward quality of the to change in comparison to a gradual delay of individual would decrease and cause reward- senescence necessary to increase floral longev- sensitive floral visitors to seek a more reward- ity. When plant-insect interactions in polli- ing tree. Female flowers must last long enough nation systems routinely are represented as ef- to reasonably assure fertilization, and in this ficient products of natural selection, examples species female flowers function about twice as such as this serve as a reminder that evolu- long as male flowers. tionary processes are only as efficient as nec- There may be additional advantages to sep- essary to assure the reproduction of the organ- arating male and female functions in unisexual isms involved. flowers. Male flower and produce pol- The results of this study support the inter- len earlier and in greater quantity without the pretation that the female flowers of Myristica cost of producing excess pistils and ovules provide no pollinator reward and function by (Armstrong and Irvine, 1989). In flowers of automimicry. The perianth closely encircling Myristica the larger perianth opening of male the stigma excludes the beetles from the floral flowers allows small floral visitors access and interior. During brief visits to female flowers, provides an enclosure while they forage for beetles may probe the stigmatic surface, but pollen (Fig. 6). The smaller perianth opening there is no evidence that they receive any sub- of female flowers tightly encircles the stigma stantial reward. (Fig. 4). Only the smallest floral visitors (ants The higher visitation rates to male trees can and thrips) could enter the female flowers, but be attributed to their greater attractiveness. these were not judged to be effective pQllen Male trees produced 2.14 times as many flow- vectors. The beetle visitors attempting entry ers as female trees (Armstrong and Irvine, to the female flowers are accurately positioned 1989), but the visitation rate to male flowers atop the stigma. When flowers function by de- was over four times greater than for female ceit, the reduced duration of visits requires flowers. Larger floral displays have been dem- such a mechanism to efficiently transfer pollen onstrated to increase the removal of pollen and to the stigma (Little, 1983). Unisexual flowers contribute to a directional pollinator move- may have evolved partly because of the greater ment (Willson and Rathcke, 1974; Willson and efficiency and effectiveness of separating such Price, 1977; Queller, 1983). Even though bee- "conflicting" male and female functions (Lloyd, tles are more frequent visitors at male trees, 1982). the frequency of intertree movements of these small beetles is unknown. Although the pollen Cantharophily, the beetle pollination syn- loads observed on floral visitors were low, the drome- Beetle-pollinated flowers of primitive results ofpollination manipulations suggest that angiosperms can be grouped into three types if half of the average pollen load carried by (Thien, 1980), but all three groups possess sim- floral visitors was deposited on the stigma dur- ilar features including large, flattened ing a visitation, the observed 1% fruit pro- and staminodes that can best be interpreted as duction of open-pollinated flowers could be specializations for beetle pollination (Carl- accounted for (Armstrong and Irvine, 1989). quist, 1969; Endress, 1984b). These flowers are Given the pollinator activity we observed, three generally large with numerous, spirally-ar- out of every five female flowers will never be ranged parts that often form an enclosure. If visited. We interpreted the reproduction of the flowers are small, they are aggregated into Myristica insipida to be pollination limited large inflorescences that function like a many- (Armstrong and Irvine, 1989). parted flower. The rewards offered typical bee- tle pollinators include food, protection from Dioecy andfloralfunction -Unisexual flow- predators while within the flower (or inflores- ers separate male and female functions. The cence), a location for mating activities, and, in separation of sexual functions allows male and many cases, a brood substrate (Grant, 1950; female flowers to have different floral longev- Faegri and Pijl, 1971; Thien, 1974, 1980; ities, and female flowers generally last longer Gottsberger, 1977; Beach, 1982; Endress, than male flowers (Primack, 1985). The inter- 1984a, b, 1986). Beetles often congregate in sexual difference in floral longevity of nutmeg considerable numbers within such flowers or can be attributed to their different functions. inflorescences. Members of Palmae and Za- Once a male flower is depleted of pollen, its miaceae also have beetle pollination with many functional role is completed. The only other of the same characteristics (Norstog and Ste- possible function would be in contributing to venson, 1980; Henderson, 1986; Norstog, Ste- the overall attractiveness of the plant. If too venson, and Niklas, 1986; Barfod, Henderson, many pollen-depleted flowers are used as part and Balslev, 1987; Tang, 1987).

This content downloaded from 132.198.8.49 on Tue, 06 Sep 2016 16:15:01 UTC All use subject to http://about.jstor.org/terms 92 AMERICAN JOURNAL OF BOTANY [Vol. 76

The phytophagy of beetles involved in can- The successful cultivation of Myristica fra- tharophily is often called "destructive" be- grans in other tropical areas, including the neo- cause various floral parts are partly or wholly tropics, indicates that pollen-foraging beetles consumed. Many beetle-pollinated flowers are common to tropical forests and emphasizes show specializations for this type of pollination the generalist nature of this pollination syn- with some combination of numerous parts, drome. Given the prevalence of small, incon- fleshy perianth parts, fleshy anthers, stami- spicuous flowers in tropical forests, it is not nodes, or some other specialized structures to unreasonable to expect that Myristica-type provide the food reward (Endress, 1984a, b). beetle pollination will be discovered in other Such feeding is not necessarily destructive in species. the sense of detrimental. However, in some cases the distinction between beetles that func- Microcantharophily-The cantharophily of tion as floral pollinators and beetles that are Myristica, Drimys brasiliensis, and Diospyros floral predators may be difficult (Armstrong, pentamera is clearly distinctive, differing from 1979; Crowson, 1981). "typical" cantharophily in almost all respects The beetle pollination syndrome of Myris- (Table 3). To distinguish this beetle pollination tica is similar to only two other species, Drimys syndrome from typical cantharophily, we pro- brasiliensis (Gottsberger, 1977; Gottsberger et pose referring to the beetle pollination of My- al., 1980) and Diospyros pentamera (House, ristica, Drimys brasiliensis, and Diospyros 1985). The flowers of Myristica, Drimys bras- pentamera as "microcantharophily." The con- iliensis, and Diospyros pentamera are small, sistently small size of the pollen-foraging bee- inconspicuous, numerous, but not aggregated tles prompted the selection of the descriptive into large inflorescences, with few, exposed flo- prefix "micro-," although we recognize that ral parts, and little or no post-opening move- small beetles can be involved in typical can- ment of floral parts. The floral odors are ab- tharophily as well. We regard the suggestion solute rather than deceptive (sensu Faegri and that "it may soon be necessary to distinguish Pijl, 1971). The beetle pollinators do not feed between different syndromes of beetle polli- on floral parts, but forage specifically on pollen. nation" (Bernhardt and Thien, 1987) as very In Myristica male flowers alone provide a food prophetic. reward of pollen to floral visitors, while female The evolution of "deceptive" floral fra- flowers function by automimicry. However, we grances from compounds functioning in her- do not suggest that unisexual flowers or au- bivore deterrance, and an accompanying shift tomimicry necessarily are features to be as- from phytophagy to feeding/reproductive ac- sociated with this pollination syndrome. A fe- tivities resulting in high species constancy male reward of a stigmatic exudate was reported (Pellmyr, 1985; Pellmyr and Thien, 1986) is a in Drimys brasiliensis (Gottsberger et al., 1980), very attractive hypothesis to account for the and as we envision this syndrome, stigmatic origin of cantharophily. The evolution of mi- or glandular exudates would be the primary crocantharophily cannot be so easily attributed reward of female or female-phase flowers. to the same scenario because so many essential The beetles are very small, less than 3-4 mm elements of the beetle/plant interactions are long, and are pollen-foragers. Their visits to missing. It is equally difficult to derive micro- flowers are largely solitary, nondestructive (no cantharophily from cantharophily, which feeding on floral parts), and do not involve would involve a major shift in both floral func- mating or reproductive activities. In general, tion and beetle behavior, eliminating all ofthose these small beetles appeared to be active and aspects (reproductive activities, use of floral agile fliers. We interpret their relationship with parts as a brood substrate) that insure polli- the to be general rather than specific. nator constancy. Their brood substrates remain unknown, and The simpler beetle/plant interactions of mi- it is possible that each species reproductively crocantharophily were the basis for regarding specializes on some visited plant species, while the beetle pollination of Drimys brasiliensis as opportunistically foraging on other species in a primitive form of cantharophily (Gottsber- flower throughout the year. A similar situation ger, 1977; Gottsberger et al., 1980). The com- was reported where a beetle pollinator of Ac- plex beetle/plant interactions of cantharophily taea used a later-flowering species of Geum as have been interpreted as a derived condition its only brood substrate (Pellmyr, 1984). (Bemhardt and Thien, 1987), and most of the Members of Myristicaceae are found in the beetles involved in these complex interactions forests of all major tropical areas. Their similar are more modem groups, some of which have floral morphologies suggest the possibility of specializations for floral feeding and pollen pollination syndromes similar to Myristica's. digestion (Crowson, 1981). An exception to

This content downloaded from 132.198.8.49 on Tue, 06 Sep 2016 16:15:01 UTC All use subject to http://about.jstor.org/terms January 1 9 8 9] ARMSTRONG AND IRVINE-MYRISTICA FLORAL BIOLOGY 9 3

TABLE 3. Comparison offlower and beetle characteristics associated with cantharophily and microcantharophily

Cantharophily Microcantharophily

Plant characteristics Flower size Generally large (> 1 cm)a Generally small (< 1 cm) Flower number Fewa Numerous, not aggregated Floral appearance Conspicuous Inconspicuous Floral parts Numerous, movement, specialized struc- Few, no movement turesb Floral shape Trap or enclosure Enclosure or exposed Landing area Large Small Odor Deceptive Absolute Rewards Food: floral parts, pollen, ovules; protec- Food: pollen, stigmatic exudates, nectar (?) tion; breeding place; brood substrate Beetle characteristics Size Generally large Generally small (<4 mm) Flying ability Limited, clumsy Active, agile Visiting behavior Accumulate in numbers Largely solitary Visit duration Prolonged visits Brief visits Feeding behavior General floral phytophagyc Specialized pollen and/or nectar foraging Mating behavior Takes place within flowers Mating elsewhere Brood substrate May use floral parts Not known to use floral parts Relationship to plant Often specialized, high visitor constancy Generalist, diversity of visitors

a When small aggregated into large inflorescences that function similar to large, solitary flowers with numerous parts. b Broad, leaf-like stamens and staminodes. c Feeding on floral parts.

this is the association of the primitive beetle caceae) (Smith, 1937). This interpretation is family Nitidulidae with cantharophily and mi- consistent with the hypothesized phytogeo- crocantharophily (Table 2) (Crowson, 1981; graphic history of the nutmegs (Raven and Ax- Gazit, Galon, and Podoler, 1982). While mi- elrod, 1974; Thorne, 1974). The systematic crocantharophily is simpler than cantharophi- relationships and similarities between Annon- ly, involving fewer plant/beetle interactions, aceae, , and Myristicaceae (Thorne, the same type of difficulties exist in construct- 1974; Cronquist, 1981) are more evident when ing a scenario deriving cantharophily from mi- the differences in floral morphology are viewed crocantharophily. Thus, it is probably point- in the preceding perspective. less to use designations of "primitive" and Generalist entomophily, involving beetles, "specialized" for two beetle pollination syn- thrips, micropterigid moths, stoneflies, and dromes that do not necessarily have a phyletic possibly ancestors of Dipterans and Hyme- relationship. nopterans has been hypothesized for insect- Endress (1987) has proposed that among pollinated protoangiosperms and ancestral primitive angiosperms the unelaborated, Magnoliidae (Bernhardt and Thien, 1987). Mi- "open" construction of flowers may have al- crocantharophily might have arisen from such lowed the relatively easy transition between a general entomophily among protoangio- large, many-parted, spiral flowers and small, or early angiosperms. Perhaps, as we few-parted, cyclic flowers. This suggests that suggest, microcantharophily is best regarded the flowers typical of cantharophily and mi- as a distinctive, relatively archaic, perhaps re- crocantharophily can have a common origin, lictual, pollination syndrome. and, perhaps, can be found within the same taxon, e.g., Winteraceae. The flowers of My- ristica, which presently show no structural evi- LITERATURE CITED dence of a former bisexual or many-parted con- ARMSTRONG, J. A. 1979. Biotic pollination mechanisms dition (Wilson and Maculans, 1967; Armstrong in the Australian flora-a review. New Zealand J. Bot. and Tucker, 1986), and other Myristicaceae, 17: 467-508. may represent an archaic adaptation to micro- ARMSTRONG, J. E., AND B. A. DRUMMOND III. 1986. Flo- cantharophily. Interactions with the generalist ral biology of Myristica fragrans Houtt. (Myristica- ceae), the nutmeg of commerce. Biotropica 18: 32- beetle pollinators could have led to an early 38. evolution of dioecy in this family. Monoecy , AND A. K. IRVINE. 1989. Flowering, sex ratios, with simple 3-merous flowers is reported to pollen-ovule ratios, fruit set, and reproductive effort still exist in neotropical (Myristi- of a dioecious tree, Myristica insipida R. Br. (Myris-

This content downloaded from 132.198.8.49 on Tue, 06 Sep 2016 16:15:01 UTC All use subject to http://about.jstor.org/terms 94 AMERICAN JOURNAL OF BOTANY [Vol. 76

ticaceae), in two different rain forest communities. spatial pattern in some dioecious tropical rain-forest Amer. J. Bot. 76: 74-85. trees. Ph.D. dissertation, Australian National Uni- , AND S. C. TUCKER. 1986. Floral development in versity, Canberra. Myristica (Myristicaceae). Amer. J. Bot. 73: 1131- LITTLE, R. J. 1983. A review of floral food deception 1143. mimicries with comments on floral . In C. BARFOD, A., A. HENDERSON, AND H. BALSLEV. 1987. A E. Jones and R. J. Little [eds.], Handbook of exper- note on the pollination of Phytelephas microcarpa imental pollination biology, 294-309. Van Nostrand (Palmae). Biotropica 19: 191-192. Reinhold, New York. BAWA, K. S., S. H. BULLOCK, D. R. PERRY, R. E. COVILLE, LLOYD, D. G. 1982. Selection of combined versus sep- AND M. H. GRAYUM. 1985. Reproductive biology arate sexes in seed plants. Amer. Naturalist 120: 571- of tropical lowland rain forest trees. II. Pollination 585. systems. Amer. J. Bot. 72: 346-356. MCILROY, R. J. 1967. An introduction to tropical cash BEACH, J. H. 1982. Beetle pollination of Cyclanthus bi- crops. Ibadan University Press, Ibadan, Nigeria. partitus (Cyclanthaceae). Amer. J. Bot. 69: 1074-1081. NORSTOG, K. J., AND D. W. STEVENSON. 1980. Wind? BERNHARDT, P., AND L. B. THIEN. 1987. Self-isolation Or insects? The pollination of . Fairchild Trop. and insect pollination in the primitive angiosperms: Garden Bull. 35: 28-30. new evaluations of older hypotheses. P1. Syst. Evol. ,AND K. J. NIKLAS. 1986. The role of bee- 156: 159-176. tles in the pollination of Zamiafurfuracea L. fil. (Za- CARLQUIST, S. 1969. Toward acceptable evolutionary miaceae). Biotropica 18: 300-306. interpretations of floral anatomy. Phytomorphology PELLMYR, 0. 1984. The pollination ecology of Actaea 19: 332-362. spicata (Ranunuculaceae). Nord. J. Bot. 4: 443-456. COBLEY, L. S. 1976. Introduction to the botany oftropical 1985. Pollination adaptations in the Cimicifugae crops. 2d ed. Longman, London. and the evolutionary origin of pollinator-plant mu- CRONQUIST, A. 1981. An integrated system of classifi- tualism, 1-34. Acta Univ. Upsaliensis 2. Uppsala, cation of flowering plants. Columbia University Press, Sweden. New York. 5 AND L. B. THIEN. 1986. Insect reproduction and CROWSON, R. A. 1981. The biology of the Coleoptera. floral fragrances: keys to the evolution of the angio- Academic Press, New York. sperms? Taxon 35: 76-85. DEINUM, H. 1949. Nootmuskaat en foelie. Landbouw PRIMACK, R. B. 1985. Longevity of individual flowers. Ind. Arch. 2(b): 655-683. Annual Rev. Ecol. Syst. 16: 15-37. ENDRESS, P. K. 1984a. The flowering process in the Eu- PURSEGLOVE, J. W. 1968. Tropical crops: dicotyledons pomatiaceae (Magnoliales). Bot. Jahrb. Syst. 104: 297- 2. J. Wiley, New York. 319. QUELLER, D. C. 1983. Sexual selection in a hermaph- 1984b. The role of inner staminodes in the floral roditic plant. Nature 305: 706-707. display of some relic Magnoliales. P1. Syst. Evol. 146: RAVEN, P. H., AND D. I. AXELROD. 1974. Angiosperm 269-282. biogeography and past continental movement. Ann. * 1986. Reproductive structures and phylogenetic Missouri Bot. Gard. 61: 539-673. significance of extant primitive angiosperms. P1. Syst. SMITH, A. C. 1937 (1938). The American species of My- Evol. 152: 1-28. risticaceae. Brittonia 2: 393-510. . 1987. Floral phyllotaxis and floral evolution. TANG, W. 1987. Insect pollination in the Zamia Bot. Jahrb. Syst. 108: 417-438. pumila (Zamiaceae). Amer. J. Bot. 74: 90-99. FAEGRI, K., AND L. VAN DER PUJL. 1971. The principles THIEN, L. B. 1974. Floral Biology of Magnolia. Amer. of pollination ecology. Pergamon, New York. J. Bot. 61: 1037-1045. GAZIT, S., I. GALON, AND H. PODOLER. 1982. The role 1980. Patterns of pollination in the primitive of nitidulid beetles in natural pollination of Annona angiosperms. Biotropica 12: 1-13. in Israel. J. Amer. Soc. Hort. Sci. 107: 849-852. THORNE, R. F. 1974. A phylogenetic classification of the GOTTSBERGER, G. 1974. The structure and function of Annoniflorae. Aliso 8: 147-209. the primitive angiosperm flower-a discussion. Acta VOGEL, S. 1962. Duftdrusen im Dienst der Bestaubung Bot. Neerl. 23: 461-471. uber Bau und Funktion der Osmophoren. Akad. Wiss. 1977. Some aspects of beetle pollination in the Abh. Math.-Naturwiss. KI. 10: 598-763. evolution of flowering plants. P1. Syst. Evol. Suppl. 1: WILLSON, M. F., AND P. W. PRICE. 1977. The evolution 211-226. of inflorescence size in Asclepias (Asclepiadaceae). ,I. SILBERBAUER-GOTTSBERGER, AND F. EHRENDOR- Evolution 31: 495-511. FER. 1980. Reproductive biology in the primitive , AND B. J. RATHCKE. 1974. Adaptive design of relic angiosperm Drimys brasiliensis (Winteraceae). the floral display in Asclepias syriaca L. Amer. Midl. P1. Syst. Evol. 135: 11-39. Naturalist 92: 47-57. GRANT, V. 1950. The pollination of Calycanthus occi- WILSON, T. K., AND L. MACULANS. 1967. The mor- dentalis. Amer. J. Bot. 37: 294-297. phology of the Myristicaceae. I. Flowers of Myristica HENDERSON, A. 1986. A review of pollination studies in fragrans and M. malabarica. Amer. J. Bot. 54: 214- the Palmae. Bot. Rev. 52: 221-259. 220. HOUSE, S. M. 1985. Relationships between breeding and

This content downloaded from 132.198.8.49 on Tue, 06 Sep 2016 16:15:01 UTC All use subject to http://about.jstor.org/terms