| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
Reproductive Biology |
2Department of Biology, Sungshin Women's University, Seoul, Republic of Korea 136-742; 3Department of Biology, University of Massachusetts, Boston, Massachusetts 02125 USA
Received for publication June 13, 2002. Accepted for publication January 10, 2003.
 |
ABSTRACT |
|---|
 TOP ABSTRACT INTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONLITERATURE CITED |
|---|
Key Words: Costa Rica • flowering patterns • phylogeny • pollinators • sexual selection • succession • tropical forests
|
INTRODUCTION |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONLITERATURE CITED |
|---|
). Variables in flowering should also be strongly influenced by other reproductive characters such as pollen vectors and sexual systems that along with flowering time, frequency, and duration are selected to optimize reproductive success. Moreover, flowering patterns should also be influenced by successional status and habit, which themselves are correlated with sexual systems, pollination mechanism, and flower size (Darwin, 1877
; Stebbins, 1970
; Bawa, 1980a
; Primack, 1987
). However, there is no systematic treatment or discussion of how timing, frequency, and duration of flowering may be influenced by pollination mechanism, sexual systems, habit, or successional status. Here, using data for 302 species from the tropical wet evergreen forest at La Selva, Costa Rica, we examine how time, frequency, and duration of flowering correlate with sexual systems, pollination mode, successional status, and growth habit. We first propose hypotheses about the relationships between each major variable of flowering and other reproductive traits as well as growth habit and successional status. We then test predictions of these hypotheses with data from the wet, evergreen forest plants. The results not only conform to predictions, but also reveal new relationships not reported before.
Flowering time
Based on hypotheses that early successional species are r-selected and have evolved to rapidly use transient habitats (Baker, 1974
; Bazzaz, 1979
; Bazzaz and Pickett, 1980
), these species should be selected to flower and fruit any time of the year, unless constrained by pollinators or seed dispersal agents. Species pollinated by different vectors, on the other hand, should have evolved to flower at a time when these pollinators are most abundant. For example, in the dry forest, pollinating moths are abundant in the wet season (Haber and Frankie, 1989
), medium-sized bees in the dry season (Janzen, 1967
). Although the dry season in the wet, evergreen forests is not as pronounced as in the dry forest, flowering of species pollinated by different vectors may be seasonal, though not to the same extent as in the dry forest. Because there is an association between dioecy and generalist insects as pollen vectors (Bawa, 1980b
), flowering of dioecious species may also be seasonal.
In summary, in relation to timing we predict that early successional species should flower any time of the year, but species pollinated by various pollen vectors as well as dioecious species should flower nonrandomly across seasons.
Flowering frequency
In colonizing species, selection for rapid growth, early reproductive maturity, and relatively high reproductive effort (Baker, 1974
; Bazzaz, 1979
; Bazzaz and Pickett, 1980
) should make the evolution of supra-annual flowering unlikely. Supra-annual flowering is made possible through stored reserves (Bullock and Solís-Magallanes, 1990
; Oyama and Mendoza, 1990
; Bullock, 1992
). Such species flower massively over brief periods (Ashton et al., 1988
; de Jong et al., 1992
; Appanah, 1993
) to recruit pollinators. Understory species are less likely to accumulate reserves and should not be selected to have massive displays that function especially well for canopy species where long-distance visual cues may be important for pollinators. Thus, supra-annual flowering should be less common in understory than in canopy species. Finally, dioecious species should be selected to have an extended flowering period as will be argued later. By inference, supra-annual flowering should be rare among dioecious species.
Briefly then, supra-annual flowering should be less common in early successional as compared to late successional species, understory as compared to canopy species, and dioecious as compared to hermaphroditic species.
Flowering duration
Because early successional species have greater access to such resources as sunlight and nutrients (Pearcy, 1983
; Vitousek and Denslow, 1986
; Becker et al., 1988
), these species should flower over a longer period than late successional species. Understory species, too, should exhibit extended flowering, because they may not accumulate enough reserves, as mentioned earlier, for short episodes of massive blooming. Massive blooming might also be selected against because the raison d'être for massive blooming is the ability of such flowering to visually attract pollinators from long distances.
In dioecious species, a division of labor is associated with sexual reproduction (Darwin, 1877
). Individuals thus should have more resources for flowering and fruiting than do hermaphroditic species. Intra-sexual selection among males should also be more intense than for hermaphrodites because the males in dioecious species do not have to bear the cost of seed production. Competition among males may also select for longer breeding seasons. Indeed males flower for a longer period than the females in dioecious species (Bawa, 1980b
; Bullock et al., 1983
). Thus, on the basis of energetics and sexual selection, dioecious species should flower for a longer period than hermaphroditic species.
Reliability of pollination should also influence duration of flowering. Specialization in mutualistic interactions is assumed to result in reliability and predictability of benefits conferred by mutualists on each other, whereas generalist interactions may have the opposite effects (Waser et al., 1996
; Wilson and Thompson, 1996
; Johnson and Steiner, 2000
).
Thus, flowering duration should be longer in early successional as compared to late successional species, understory species as compared to canopy species, dioecious species as compared to hermaphroditic species, and species pollinated by generalist vectors as compared to specialized vectors.
We first describe the distribution of ecological characters such as successional status, habit, sexual systems, and pollen vectors among La Selva tree species. We then focus on the association between flowering and ecological characters. The relationship between flowering and ecological characters could be confounded by a taxonomic membership of species if certain specious taxa possessed a particular ecological character(s). Thus, we reexamine the relationship between flowering and ecological characters after excluding the effect of phylogeny acting at the level of family.
|
MATERIALS AND METHODS |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONLITERATURE CITED |
|---|
). More than 1000 individuals belonging to 302 species of 58 families were marked along the trails of the station. Sample size per species ranged from one to many (see Bawa et al., 2003
). Flowering time, frequency, and duration per flowering episode were observed weekly for tree species.
Because many species flowered in different months during the 3 yr of observations, flowering months were obtained by pooling the 3 yr of data for each species. Flowering time was then classified into four levels (first-wet, second-wet, both-wet, and dry), which were primarily distinguished by flowering in the first (May–July) or second (November–January) wet season, or the dry season (see Bawa et al., 2003
for description).
Flowering frequency was classified into four levels based on flowering frequency per year (annual flowering frequency): continual species that flower more or less continually throughout a year, episodic species that flower twice or more times a year, annual species that flower once a year, and supra-annual species that do not flower every year. For most episodically flowering species, the minimum flowering frequency per year (sub-annual flowering frequency) was also ascertained. The sub-annual flowering frequency is categorized into three levels (twice, three, and four or more times). Flowering duration per episode was treated as a continuous variable, ranging from less than 1 wk to up to 40 wk. Throughout this paper, if not specifically mentioned, flowering duration represents the duration per episode. For detailed descriptions for these flowering characters see Bawa et al. (2003)
.
Ecological information such as successional status and habit was obtained in situ. Data on sexual systems and pollen vectors were collected from previous studies (Bawa et al., 1985a
,b
; Kress and Beach, 1994
) at the same site. The successional status and habit of each tree species were classified into two categories, respectively, early vs. late successional status and understory vs. canopy species. Because the number of emergent species was few (N = 6), these species were combined with canopy species. Three sexual systems were distinguished, i.e., species with hermaphroditic flowers, monoecious flowers, and dioecious flowers. Thirteen different categories of pollen vectors were identified based on Bawa et al. (1985a)
and Kress and Beach (1994)
. Depending upon analyses, the pollen vector consists of five categories (the five most common vectors such as small bees, medium/large bees, beetles, moths, and small insects) or two categories (specialists including medium/large bees, beetles, hawk moths, bats, and hummingbirds and generalists including small bees, moths, small diverse insects, butterflies, wasps, thrips, wind, and flies).
We first examined the association between pairs of ecological and categorical flowering characters (flowering time and frequency). The association between these categorical characters was examined in two-way contingency table analyses on data pooled over all 58 families. The relationship between ecological characters and flowering duration was examined using ANOVAs with flowering duration as a dependent variable since flowering duration was treated as a continuous variable. We also examined the associations of phylogeny with ecological characters. Seven large families comprising 47.4% of the total 302 species were used as classes of the phylogeny variable in two-way contingency tables of phylogeny by each ecological character. In these contingency table analyses, we identified the categories that contributed significantly to the G value based on the adjusted standardized residuals. Because of limited sample size, the association between categorical variables could not be tested independently of phylogeny using a three-way table such as successional status x flowering time x phylogeny. Thus, the relationship between categorical characters after considering phylogeny was examined based on a subset data of species of 43 minor families with 
5 species each (number of species per family = 2.37 ± 1.24, mean ± 1 SE, N = 102). On the other hand, the relationships between ecological characters and flowering duration after considering phylogeny were directly tested with ANOVA incorporating phylogeny as a class variable. Only successional status among ecological characters was consistently associated with flowering frequency and duration. Thus, the relationship between flowering time, frequency, phylogeny, successional status, and duration was simultaneously examined in ANOVA with flowering duration as a dependent variable. Because of small and unbalanced sample sizes, only two-way interactions were considered in this ANOVA. When the effects of factors involving successional status were examined separately for each flowering frequency category, only episodic and annual flowering were considered since supra-annual flowering was found in only late successional species.
Data on ecological as well as on flowering characters of each species were not entirely complete. For example, the successional status was recorded for all of the species listed, but the type of pollen vector was known for about only half the number of species. Therefore, several two-way contingency tables, especially those involving the pollen vectors or phylogeny, created cells with low or zero count, violating the assumption of less than 20% of the cells with expected counts <5 (Marascuilo and Levin, 1983
). Contingency table analyses were still conducted to obtain a preliminary idea for the relationship between ecological and flowering characters. Flowering duration per episode was log transformed (base 10). The species flowering continually (N = 15) with no variation in flowering duration were deleted when conducting analyses of variance. Means of flowering duration are given with 1 SD throughout this paper. All analyses were conducted using the SAS (1999)
8.1 package.
|
RESULTS |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONLITERATURE CITED |
|---|
|
|
Relationship between flowering and ecological characters
Flowering time and individual ecological characters
the pattern before considering phylogeny—Early successional species flowered across all flowering seasons with peak flowering in the first-wet season and the lowest number of species flowering in the second-wet season. However, early and late successional species did not differ in the relative proportion of flowering species across seasons (Table 3A).
|
|
the pattern after considering phylogeny—Species of 43 minor families did not show different patterns of association between flowering time and ecological characters from species of all families (Table 3B).
Flowering frequency and individual ecological characters
the pattern before considering phylogeny—There was significant interaction between flowering frequency and several ecological characters such as successional status, habit, and sexual system (Table 4A). Species flowering continually and those flowering once in several years contributed largely to this interaction. Episodically and annually flowering species did not appear to be distinguished by a particular successional status, growth habit, and sexual system.
|
The relationship between flowering frequency and the five most common pollen vectors was statistically insignificant (Table 4A). However, the data indicated some, but not simple, associations between flowering frequency and pollen vectors. For example, only medium/large bees and small insects were recorded across all levels of flowering frequency. Small bees and wind were entirely associated with episodically or annually flowering species. Contrast also existed between bat- and hummingbird-visited species. The bat-visited Quararibea parvifolia (Bombacaceae) and one Calyptrogyne species (Palmae) flowered episodically, and Ceiba pentandra (Bombacaceae) flowered supra-annually. None of the bat-visited species bloomed continually. Three out of six hummingbird-visited species such as Psychotria elata, C. poeppigiana, and Hamelia patens were continually flowering species of the Rubiaceae, whereas this specialized vector visited none of the supra-annually flowering species.
The sub-annual flowering frequency of episodically flowering species was not particularly associated with any ecological character (two-way contingency table of sub-annual flowering frequency with successional status, G = 1.94, df = 2; with habit, G = 2.78, df = 2; with sexual system, G = 3.04, df = 4; with pollen vector, G = 10.87, df = 8, P > 0.05 for all analyses, N[total] = 72, 72, 70, and 28, respectively).
the pattern after considering phylogeny—Among species of minor families, only the association between flowering frequency and successional status remained significant (Table 4B); the absence of supra-annually flowering species in the early successional status was notable. Supra-annual flowering in understory species was not rare among species of the minor families (Table 4B), suggesting that the association of supra-annual flowering and canopy was largely accrued through the effect of phylogeny. The G score for the association of flowering frequency and sexual system was substantial: among supra-annually flowering species of the minor families, species with hermaphroditic flowers were six times more frequent than those with dioecious flowers.
Flowering duration and individual ecological characters
the pattern before considering phylogeny—Only the effect of the successional status on flowering duration was significant among ecological characters examined, accounting for only 8.9% of the total variance (Table 5). Early successional species flowered twice as long as late successional species.
|
Dioecious species flowered 1 to 2 wk longer than the monoecious and hermaphroditic species, but the effect of sexual system was not significant (Table 5).
Species pollinated by 12 different vectors differed by five-fold in mean flowering duration. Mean duration increased in the order of bat < wind < hawk moth, thrip < medium/large bee, beetle, wasp, moth < butterfly < small insect, small bee < hummingbird (Table 5). However, the effect of pollen vectors on flowering duration was statistically insignificant among all 12 vectors as well as between generalist vs. specialist vectors (Table 5). Notably, species visited by the two vertebrate vectors differed markedly in flowering duration. For example, bat-pollinated Ceiba pentandra flowered supra-annually for 3 wk and Quararibea parvifolia flowered episodically for less than 1 wk. Hummingbird-pollinated species bloomed for many weeks, for example, Symphonia globulifera (Guttiferae) for 8 wk per episode, Pentagonia donnell-smithii (Rubiaceae) for more than 20 wk, and Psychotria elata, P. poeppigiana, and Hamelia patens (all the Rubiaceae) continually throughout a year.
To avoid the confounding effect of flowering frequency on flowering duration, the relationship between flowering duration and successional status, habit, sexual systems, and pollen vectors was examined separately for each flowering frequency category. The longer flowering duration of early successional species relative to late successional was significant only in species flowering episodically, though this trend was also noticed in species flowering annually (Table 6, Fig. 2). Dioecious species as compared to those with hermaphroditic flowers—and species visited by generalist vectors as compared to specialist vectors—tended to flower longer, though statistically not significantly across flowering frequency categories (Table 6, Fig. 2). Understory and canopy species did not differ in mean flowering duration regardless of flowering frequency.
|
|
the pattern after considering phylogeny—In all ANOVAs with phylogeny and one of the ecological characters as main factors, neither two-way interaction terms nor phylogeny had significant effects (Table 7). Only successional status among ecological characters had a significant effect on flowering duration: early successional species exhibited 1.6 times longer flowering than late successional species (X = 9.48 ± 2.41 wk and 5.87 ± 1.84 wk, N = 13 and 56, respectively). Thus, the relationship between ecological characters and flowering duration did not change after considering phylogeny.
|
|
|
|
|
|
DISCUSSION |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONLITERATURE CITED |
|---|
Flowering time
We predicted that early successional species should flower any time of the year. This prediction could be realized in two ways. First, the early successional species could flower continually or for a long period. This prediction was realized (described later). Second, early successional species as a group could flower with less seasonality than the late successional species. Indeed early successional species as a group flower across all seasons, but the late successional species do also. Although there are seasonal peaks and troughs, the seasonal patterns are concordant between early and late successional species. This concordance explains continual flowering throughout the year at the community level at La Selva unlike the dry and cloud tropical forests (e.g., Frankie et al., 1974
; Opler et al., 1980
; Koptur et al., 1988
; Wright and Calderon, 1995
).
Aseasonal flowering has probably coevolved with the aseasonal availability of pollen vectors. Thus not surprisingly, our prediction that species pollinated by different vectors may flower in different seasons did not hold. This prediction was based on differential abundance of various pollen vectors in the dry and wet seasons in the dry tropical forests (Janzen, 1967
; Frankie et al., 1983
; Haber and Frankie, 1989
). Apparently, seasonality is not as pronounced in the wet tropical forests as we expected.
The prediction that dioecious and hermaphroditic species may differ in their flowering times was also based on the assumption of differential seasonal abundance of generalist pollen vectors that largely pollinate dioecious species (Bawa and Opler, 1975
; Bawa, 1980b
; Ibarra-Manríquez and Oyama, 1992
). Apparently, generalist and specialist vectors are equally abundant across seasons. The long flowering duration of dioecious species discussed later may also constrain seasonal segregation of flowering time in two groups of species. Our results may also have been compromised by two factors. First, we defined flowering time with only two categories, wet or dry season. Second, the small sample sizes for species pollinated by different vectors may have made seasonal patterns difficult to discern, if indeed such patterns exist.
Flowering frequency
We predicted that supra-annual flowering should be less common in early successional as compared to late successional species. In habitats occupied by early successional species, selection of rapid growth, early reproduction, and efficient dispersal (Baker, 1974
; Bazzaz, 1979
; Bazzaz and Pickett, 1980
) is unlikely to favor the evolution of supra-annual flowering, which is unpredictable and occurs once in several years. Early successional habitats are also ephemeral in nature. Because continual flowering accompanies continual fruiting and fruit dispersal, such species can enhance their opportunity to pre-empt transient habitats. Thus, in early successional species, continually flowering species are over-represented in contrast to supra-annually flowering species. Continual flowering, and subsequently continual fruiting, also require a more or less continual supply of photosynthetic resources as well as water. Open sites, such as clearings and treefall gaps, maintain a higher level of soil water and nutrients (Vitousek and Denslow, 1986
; Becker et al., 1988
).
We predicted that supra-annual flowering should be less common in understory than in canopy species. The massive floral displays over a brief period to recruit pollinators are characteristic of supra-annually flowering species (Ashton et al., 1988
; de Jong et al., 1992
; Appanah, 1993
) that are not likely to be as effective in the canopy, where large masses of flowers act as visual cues over long distances. Momose et al. (1998)
also explained the higher frequency of supra-annual flowering species in the canopy of Malaysian forests on the basis of floral displays. Massive display also depends on stored reserves (Bullock and Solís-Magallanes, 1990
; Oyama and Mendoza, 1990
; Bullock, 1992
). Understory species grow in light-poor environments and are less likely to accumulate reserves than canopy species (see also Momose et al., 1998
). Although supra-annual flowering is less common in understory than in canopy species, phylogeny seems to be important in the distribution of supra-annual flowering. More than one-third of canopy species in certain large families such as the Leguminosae and Lauraceae flower supra-annually.
We also predicted that supra-annually flowering species would be rare among dioecious species. The proportion of dioecious species in supra-annually flowering species is low compared to the overall proportion of species with dioecious flowers among La Selva species (9% vs. 23%). This result may reflect disadvantages in association between dioecy and supra-annually flowering. Supra-annual flowering is brief and massive perhaps to attract pollen vectors in species that flower unpredictably at irregular intervals. On the other hand, as argued earlier, sexual selection in dioecious species (Bawa, 1980a
, b
) may favor extended flowering, facilitated by division of labor (Darwin, 1877
). Such disparity in selection pressures for different life history traits is likely to decrease the frequency of dioecy in supra-annually flowering species. Because both sexual system and flowering frequency are constrained by phylogeny, the rarity of supra-annual flowering within confamilial dioecious species or vice versa should further support our argument. In the Moraceae, 88% of species have dioecious flowers, but none of them flower supra-annually. Thus, dioecious flowers may have been selected against in supra-annually flowering species during species divergence, at least in the Moraceae. However, the lack of association between dioecy and supra-annual flowering needs to be further examined because dioecy and flowering frequency may also be correlated with other unexamined ecological characters. Supra-annually flowering species produce seeds occasionally, but when they flower seed crops are huge. Mast fruiting is regarded as a strategy to satiate seed predators and is supposed to have evolved in response to pressure from seed predators (Janzen, 1974
; Curran and Webb, 2000
). In dioecious species, only female individuals produce seeds, limiting the ability of dioecious populations to satiate seed predators following mast fruiting.
Flowering duration
We predicted that flowering duration should be longer in early successional than in late successional species, in understory species than in canopy species, in dioecious species than in hermaphroditic species, and in species pollinated by specialized pollinators than by generalist pollinators. Our prediction held only for successional status, though we noted expected trends for other groups.
Contrary to our prediction that understory species should flower longer than canopy species, both understory and canopy species at La Selva flower on average for 5.6 wk. Interestingly, previous studies have shown a longer flowering in the understory than in the canopy (Frankie et al., 1974
; but see Stiles, 1978
) and similarly for shrubs as compared to trees (Opler et al., 1980
; Smith-Ramírez and Armesto, 1994
). Indirect support for our prediction comes from the relative rarity of supra-annually flowering species that flower briefly in the understory. Our sampling over time may not be adequate to quantify temporal intensity of flowering, thus making it difficult to discriminate massive vs. extended flowering more accurately than the average duration of flowering. On the other hand, at La Selva, photosynthetic responses to light levels, which might be the most important factors limiting growth, differ largely between early and late successional species, but not particularly between understory and canopy species (Fetcher et al., 1994
). Thus, understory species may not be particularly constrained by resources as canopy species may be.
Based on energetics and sexual selection, we predicted that species with dioecious flowers should have an extended flowering period. Species with dioecious flowers tended to bloom longer per episode and per year than did those with hermaphroditic flowers. However, longer flowering of dioecious flowers may be confounded or generated spuriously by a suite of ecological correlates of dioecy such as small, inconspicuous flower; small diverse insect vector; fleshy fruit; woody habit; and island habitat (Bawa, 1980b
; Renner and Ricklefs, 1995
). Sexual system was also constrained by phylogeny, which again was associated with flowering frequency. Thus, identification of the exclusive effect of dioecy on flowering duration should be examined by comparing related species with contrasting sexual systems.
Our prediction that species with generalist vectors would flower longer than those with specialist vectors did not hold, though there was such a trend. The association between flowering duration and pollen vectors is complex because of difficulties in characterizing pollen vectors as generalist and specialist. For example, thrips are small and quite unpredictable in their appearance, but they may be reliable pollinators for Shorea, which flower unpredictably at 2–10 yr intervals (Ashton et al., 1988
). Despite the wide usage of generalist vs. specialist vectors, distinguishing these two types of vectors is still very difficult without detailed studies (Johnson and Steiner, 2000
).
Several aspects of flowering phenology at La Selva also appear to be responsible for lack of a simple, dichotomous pattern of variation in flowering duration in relation to ecological characters. Flowering duration is constrained by flowering frequency and time (Bawa et al., 2003
). Selection for male competition in dioecious species or selection by specialized or generalist pollinators may not result in direct, concordant changes in flowering duration, if duration itself is influenced by timing and frequency. For example, Gentry (1974)
showed that species with different pollen vectors differed in flowering duration. However, he did not take flowering frequency into account even though species of the Bignoniaceae, the subject of Gentry's studies, vary greatly in flowering frequency.
The pattern of associations between flowering duration and ecological characters changed when flowering frequency was considered. Longer flowering in early successional species than in the late successional applies to only episodically flowering species. In species flowering episodically, early successional species extended flowering twofold longer than the late successional. Episodically flowering species may intrinsically be able to flower over a long period in resource-rich sites because episodic flowering is more or less analogous to continual flowering (Newstrom et al., 1994
). Piper arieianum, which flowers episodically, has been shown to prolong flowering under higher light levels (Marquis, 1988
). Our results suggest that species flowering episodically utilize various resources not only through multiple flowering episodes but also through the extended duration of these episodes. Thus, multiple episodes of flowering may reflect a very specialized feature that is employed by 40–50% of species at La Selva. Information on frequency and ecological diversity of episodically flowering species in other tropical areas is needed to validate the trends observed.
Concluding remarks
In this and the preceding paper, we shift the emphasis in phenology of flowering plants during the last four decades from the timing of flowering to other traits such as the duration and frequency of flowering. We show that the evolution of timing cannot be adequately understood without considering duration and frequency. Unlike previous analyses that demonstrate phylogenetic effects on timing (e.g., Kochmer and Handel, 1986
; Smith-Ramírez and Armesto, 1994
; Wright and Calderon, 1995
), we found timing to be independent of family membership. Furthermore, the duration is not constrained by phylogeny. Both duration and frequency seem to be strongly influenced by energetics (successional status). Sexual selection and pollinators may also influence various flowering parameters, and although we noticed trends in the expected direction, limited sample sizes precluded a thorough analysis. Other factors that we have not fully considered include seed predation (e.g., Janzen, 1974
), seed and fruit maturation time, and seed dispersal agents (e.g., Bawa, 1983
; Primack, 1987
). Thus, a number of selective forces act on flowering traits, but a preoccupation with timing of flowering has prevented rigorous analysis of the evolution of other variables. Studies of related species with contrasting flowering traits and ecological traits should help reveal the exact forces shaping the evolution and diversity of flowering traits within the context of the hypotheses proposed here.
|
FOOTNOTES |
|---|
4 Author for reprint requests (Tel: 617-287-6657; Fax: 617-287-6650; kamal.bawa{at}umb.edu
)
|
LITERATURE CITED |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONLITERATURE CITED |
|---|
Ashton P. S. T. J. Givnish S. Appanah 1988 Staggered flowering in the Dipterocarpaceae: new insights into floral induction and the evolution of mast fruiting in the aseasonal tropics. American Naturalist 132: 44-66[CrossRef][ISI]
Baker H. G. 1974 The evolution of weeds. Annual Review of Ecology and Systematics 5: 1-24
Bawa K. S. 1980a Mimicry of male by female flowers and intrasexual competition for pollinators in Jacaratia dolichaula (D. Smith) Woodson (Cariacaceae). Evolution 34: 467-474[CrossRef][ISI]
Bawa K. S. 1980b Evolution of dioecy in flowering plants. Annual Review of Ecology and Systematics 11: 15-40
Bawa K. S. 1983 Patterns of flowering in tropical plants. In C. E. Jones and R. J. Little [eds.], Handbook of experimental pollination biology, 395–410. Van Nostrand Reinhold, New York, New York, USA
Bawa K. S. S. H. Bullock D. R. Perry R. E. Coville M. H. Grayum 1985a Reproductive biology of tropical lowland rain forest trees. II. Pollination systems. American Journal of Botany 72: 346-356[CrossRef][ISI]
Bawa K. S. H. Kang M. H. Grayum 2003 Relationships among time, frequency, and duration of flowering in tropical rain forest trees. American Journal of Botany 90: 877-887
Bawa K. S. P. A. Opler 1975 Dioecism in tropical forest trees. Evolution 29: 167-179[CrossRef][ISI]
Bawa K. S. D. R. Perry J. H. Beach 1985b Reproductive biology of tropical lowland rain forest trees. I. Sexual systems and incompatibility mechanisms. American Journal of Botany 72: 331-345[CrossRef][ISI]
Bazzaz F. A. 1979 The physiological ecology of plant succession. Annual Review of Ecology and Systematics 10: 351-371
Bazzaz F. A. S. T. A. Pickett 1980 Physiological ecology of tropical succession: a comparative review. Annual Review of Ecology and Systematics 11: 287-310
Becker P. P. E. Rabenold J. R. Idol A. P. Smith 1988 Gap and slope gradients of soil and plant water potentials during the dry season in a tropical moist forest. Journal of Tropical Ecology 4: 173-184
Bullock S. H. 1992 Seasonal differences in nonstructural carbohydrate in two dioecious monsoon-climate trees. Biotropica 24: 140-145[CrossRef][ISI]
Bullock S. H. J. H. Beach K. S. Bawa 1983 Episodic flowering and sexual dimorphism in Guarea rhopalocarpa Radlk. (Meliaceae) in a Costa Rican rain forest. Ecology 64: 851-862[CrossRef][ISI]
Bullock S. H. J. A. Solís-Magallanes 1990 Phenology of canopy trees of a tropical deciduous forest in Mexico. Biotropica 22: 22-35[CrossRef][ISI]
Clark D. A. 1994 Plant demography. In L. A. McDade, K. S. Bawa, H. A. Hespenheide, and G. S. Hartshorn [eds.], La Selva: ecology and natural history of a neotropical rain forest, 90–105. University of Chicago Press, Chicago, Illinois, USA
Curran L. M. C. O. Webb 2000 Experimental tests of the spatiotemporal scale of seed predation in mast-fruiting Dipterocarpaceae. Ecological Monographs 70: 129-148
Darwin C. 1877 The different forms of flowers on plants of the same species. John Murray, London, UK
de Jong T. J. P. G. L. Klinkhamer M. J. van Staalduinen 1992 The consequences of pollination biology for selection of mass or extended blooming. Functional Ecology 6: 606-615[CrossRef][ISI]
Fetcher N. S. F. Oberbauer R. L. Chazdon 1994 Physiological ecology of plants. In L. A. McDade, K. S. Bawa, H. A. Hespenheide, and G. S. Hartshorn [eds.], La Selva: ecology and natural history of a neotropical rain forest, 128–141. University of Chicago Press, Chicago, Illinois, USA
Frankie G. W. H. G. Baker P. A. Opler 1974 Comparative phenological studies of trees in tropical wet and dry forests in the lowland of Costa Rica. Journal of Ecology 62: 881-919[CrossRef]
Frankie G. W. W. A. Haber P. A. Opler K. S. Bawa 1983 Characteristics and organization of the large bee pollination system in the Costa Rican dry forest. In C. E. Jones and R. J. Little [eds.], Handbook of experimental pollination biology, 411–447. Van Nostrand Reinhold, New York, New York, USA
Gentry A. H. 1974 Co-evolutionary patterns in Central American Bignoniaceae. Annals of the Missouri Botanical Garden 61: 728-759[CrossRef][ISI]
Haber W. A. G. W. Frankie 1989 A tropical hawkmoth community: Costa Rican dry forest Sphingidae. Biotropica 21: 155-172[CrossRef][ISI]
Ibarra-Manríquez G. K. Oyama 1992 Ecological correlates of reproductive traits of Mexican rain forest trees. American Journal of Botany 79: 383-394[CrossRef][ISI]
Janzen D. H. 1967 Synchronization of sexual reproduction of trees within the dry season in Central America. Evolution 21: 620-637[CrossRef][ISI]
Janzen D. H. 1974 Tropical blackwater rivers, animals, and mast fruiting by the Dipterocarpaceae. Biotropica 6: 69-103
Johnson S. D. K. E. Steiner 2000 Generalization versus specialization in plant pollination systems. Trend in Ecology and Evolution 15: 140-143[CrossRef]
Kochmer J. P. S. N. Handel 1986 Constraints and competition in the evolution of flowering phenology. Ecological Monographs 56: 303-325[CrossRef][ISI]
Koptur S. W. A. Haber G. W. Frankie H. G. Baker 1988 Phenological studies of shrub and treelet species in tropical cloud forests of Costa Rica. Journal of Tropical Ecology 4: 323-346
Kress W. J. J. H. Beach 1994 Flowering plant reproductive systems at La Selva Biological Station. In L. A. McDade, K. S. Bawa, G. S. Hartshorn, and H. A. Hespenheide [eds.], La Selva: ecology and natural history of a neotropical rain forest, 161–182. University of Chicago Press, Chicago, Illinois, USA
Marascuilo L. A. J. R. Levin 1983 Multivariate statistics in the social sciences. Brooks/Cole, Monterey, California, USA
Marquis R. J. 1988 Phenological variation in the neotropical understory shrub Piper arieianum: causes and consequences. Ecology 69: 1552-1565[CrossRef][ISI]
Momose K. R. Ishii S. Sakai T. Inoue 1998 Plant reproductive intervals and pollinators in the aseasonal tropics: a new model. Proceedings of the Royal Society, London B 265: 2333-2339[CrossRef]
Newstrom L. E. G. W. Frankie H. G. Baker R. K. Colwell 1994 Diversity of long-term flowering patterns. In L. A. McDade, K. S. Bawa, H. A. Hespenheide, and G. S. Hartshorn [eds.], La Selva: ecology and natural history of a neotropical rain forest, 142–160. University of Chicago Press, Chicago, Illinois, USA
Opler P. A. G. W. Frankie H. G. Baker 1980 Comparative phenological studies of treelet and shrub species in tropical wet and dry forests in the lowlands of Costa Rica. Journal of Ecology 68: 167-188[CrossRef][ISI]
Oyama K. A. Mendoza 1990 Effects of defoliation on growth, reproduction, and survival of a neotropical dioecious palm, Chamaedorea tepejilote. Biotropica 22: 119-123[CrossRef][ISI]
Pearcy R. W. 1983 The light environment and growth of C3 and C4 species in the understory of a Hawaiian forest. Oecologia 58: 26-32[CrossRef][ISI]
Primack R. B. 1987 Relationships among flowers, fruits, and seeds. Annual Review of Ecology and Systematics 18: 409-430
Renner S. S. R. E. Ricklefs 1995 Dioecy and its correlates in the flowering plants. American Journal of Botany 82: 596-606[CrossRef][ISI]
SAS. 1999 SAS system for Windows, release 8.1. SAS Institute, Cary, North Carolina, USA
Smith-Ramírez C. J. J. Armesto 1994 Flowering and fruiting patterns in the temperate rainforest of Chiloé, Chile ecologies and climatic constraints. Journal of Ecology 82: 353-365
Stebbins G. L. 1970 Adaptive radiation of reproductive characteristics in angiospersm. I. Pollination mechanisms. Annual Review of Ecology and Systematics 1: 307-326
Stiles F. G. 1978 Temporal organization of flowering among the hummingbird foodplants of a tropical wet forest. Biotropica 10: 194-210[CrossRef][ISI]
Vitousek P. M. J. S. Denslow 1986 Nitrogen and phosphorus availability in treefall gaps of a lowland tropical forest. Journal of Ecology 74: 1167-1178[CrossRef]
Waser N. M. L. Chittka M. V. Price N. M. Williams J. Ollerton 1996 Generalization in pollination systems, and why it matters. Ecology 77: 1043-1060[CrossRef]
Wilson P. J. D. Thompson 1996 How do flowers diverge?. In D. G. Lloyd and S. C. H. Barrett [eds.], Floral biology: studies on floral evolution in animal-pollinated plants, 88–111. Chapman & Hall, New York, New York, USA
Wright S. J. O. Calderon 1995 Phylogenetic patterns among tropical flowering phenologies. Journal of Ecology 83: 937-948[CrossRef]
This article has been cited by other articles:
|
|
 |
|
  K. S. Bawa, H. Kang, and M. H. Grayum Relationships among time, frequency, and duration of flowering in tropical rain forest trees Am. J. Botany, June 1, 2003; 90(6): 877 - 887. [Abstract] [Full Text] [PDF]
|
|
| ||||||||
