James D. Ackerman

Short- and long-term limitations to fruit production in a tropical orchid.

Fruit production in iteroparous flowering plants can be limited by a variety of conditions that need not be mutually exclusive nor immediately evident. We examined short- and long-term constraints to fruit production in a Puerto Rican population of the infrequently pollinated orchid Epidendrum ciliare. Average natural fruit maturation rate ranged from 5 to 15% over 4 yr. To evaluate limitations to fruit production, pollinations were augmented for a randomly chosen experimental portion of a population for two consecutive years. During this period, only 10 and 15% of flowers on control plants were naturally pollinated. Hand-pollinations of nearly all flowers produced by experimental plants increased fruit production to 33 and 49%, compared to 5 and 8% for controls. Thus, fruit production within seasons was partially limited by pollinations. Fruit predation also was heavy for both control (28 and 29% of fruits initiated) and pollination-augmented (20 and 32%) groups. No clear relationship existed between treatment and predation frequency. Furthermore, undamaged fruits often aborted (controls 11 and 37%, experimentals 15 and 28%), but frequencies were independent of self vs. outcrossed pollination and treatment. Seed crop mass declined as fruit set increased, suggesting that resources were limited. Long- term effects of heavy fruit loads were evident. Over 2 yr, the number of inflorescences, flowers, and flowers per inflorescence declined for experimentals relative to the controls. Moreover, plant size and vegetative proliferation decreased for experimentals relative to controls. Thus, through elevated fruit production the plant incurs future costs and lifetime fitness of E. ciliare might be more affected by resource constraints than by the other factors studied.

Diversity and host specificity of endophytic Rhizoctonia-like fungi from tropical orchids

All orchids have an obligate relationship with mycorrhizal symbionts. Most orchid mycorrhizal fungi are classified in the form-genus Rhizoctonia. This group includes anamorphs of Tulasnella, Ceratobasidium, and Thanatephorus. Rhizoctonia can be classified according to the number of nuclei in young cells (multi-, bi-, and uninucleate). From nine Puerto Rican orchids we isolated 108 Rhizoctonia-like fungi. Our isolates were either bi- or uninucleate, the first report of uninucleate Rhizoctonia-like fungi as orchid endophytes. We sequenced the internal transcribed spacer (ITS) region of nuclear ribosomal DNA from 26 isolates and identified four fungal lineages, all related to Ceratobasidium spp. from temperate regions. Most orchid species hosted more than one lineage, demonstrating considerable variation in mycorrhizal associations even among related orchid species. The uninucleate condition was not a good phylogenetic character in mycorrhizal fungi from Puerto Rico. All four lineages were represented by fungi from Tolumnia variegata, but only one lineage included fungi from Ionopsis utricularioides. Tropical epiphytic orchids appear to vary in degree of specificity in their mycorrhizal interactions more than previously thought.

Limitations to fruit production in Ionopsis utricularioides (Orchidaceae)

The proportionally low fruit set observed in many epiphytic tropical orchids usually has been attributed to insufficient pollination; for example, the tropical orchid Ionopsis utricularioides, a twig epiphyte, offers no floral rewards, is rarely visited, and has low fruit set. To determine what factors limit fruiting in this species, we assessed effects of resources, pollinator service, pollen parentage, and pollen quantity using greenhouse and field studies in which individual plants were followed over two flowering seasons. Plants were self-compatible, and pollination frequency substantially affected fruit set. Hand-pollination of field experimentals increased total inflorescence fruit sets from 6 to 19 percent, even though over one-third of such pollinations failed to yield fruit. For flowers pollinated on the same day, fruit failure was related to leaf length but not to position in the inflorescence, flower age, or pollen quantity (one vs two pollinia). High fruit set resulted in reduced growth and a higher probability of foregoing inflorescence production the following year. This last result suggests that resource availability may limit lifetime fruit production, even though pollination limitation occurs within a single season. We also explored some predicted consequences of resource-limited reproduction. First, we found no evidence for mate choice in the sense of selective abortion of fruits depending on selfed verses outcrossed pollen parentage. Second, our plants produced many more flowers than could mature fruit. Although fruit production on a per-flower basis decreased with inflorescence size, pollinarium removal increased with number of open flowers. This suggests that pollen donation by surplus flowers increases the fitness of those plants with larger inflorescences. However, most inflorescences are small, which suggests trade-offs between conflicting selective pressures. Third, we speculate that resource limitation of fruit production may have influenced the evolution of nectardeception pollination in many orchids. PLANTS OFTEN PRODUCE FAR fewer fruits than flowers. Resource shortages frequently limit fruit set (Stephenson 1981), but fecundity may also be reduced by adverse environmental conditions (Wyatt 1976, Coulter 1979, Lee & Bazzaz 1982), genetic incompatibility between maternal and paternal genomes (Schlising 1976, Wyatt & Hellwig 1979), paternal identity (Bertin 1982, Waser & Price 1983), seed predation (Boucher & Sork 1979, Arnold 1982, Haddock & Chaplin 1982), inadequate pollination (Bierzychudek 1981), or combinations of factors (Augspurger 1981, Heithaus et al. 1982, Hainsworth et ail. 1984). Fruit set in epiphytic tropical orchids often is low, probably in part because of infrequent pollinator visits (Darwin 1885, Dressler 1968, pers. obs.). Schemske (1980), Janzen et al. (1980), Ackerman and Montalvo (1983, pers. obs.), and Ackerman and Montero Oliver (1985) demonstrated that artificial pollination increased fruit set in natural populations of several such species. They suggested that fruit set may be pollinator limited, but also discussed complicating factors which might limit fruit production over the lifetime of these perennial plants. In Puerto Rico we examined factors limiting fruit production in an epiphytic, nectarless orchid, L. utricularioides (Sw.) Lindl. Our major questions were as follows. (1) Does pollination limit fruit set? We supplemented natural pollination on experimental plants in the field and compared subsequent fruit set of entire inflorescences to that of open-pollinated controls. An increase in fruit set indicates pollination limitation within a single season. (2) Do resources limit present or future fruit production? We followed individual plants for two flowering seasons to determine if fruit set in one year affected subsequent vegetative growth and inflorescence production. We also provided field-collected plants with water and nutrients in the greenhouse to determine effects on subsequent plant size and flower production. (3) Do pollen loads or paternal identity influence fruit set? We conducted self-compatibility experiments and measured the effects of two sizes of pollen loads on fruit set. These studies were supplemented with field observations of herbivory, disease, and natural pollination frequency. Proximate causes and ultimate consequences of limited fruit set are important in understanding many evolutionary questions (Stephenson 1981). We address the following implications of resource limitation. (1) Plants may be able to choose which ovules or fruits they mature according to I Received 28 May 1985, revision accepted 24 October 1985. 2 Order of authorship determined by a coin toss. 24 BIOTROPICA 19(1): 24-31 1987 This content downloaded from 157.55.39.248 on Thu, 28 Jul 2016 05:41:35 UTC All use subject to http://about.jstor.org/terms identity of pollen parents (Janzen 1977, Waser & Price 1983, Willson & Burley 1983), especially if resources are limiting. We tested for one level of female choice by measuring whether plants selectively mature fruits of outcrossed over selfed flowers. (2) If resources are limiting, then selection should favor a reduction in flower number unless surplus flowers contribute to fitness through pathways other than female fecundity. We explore some of the possible functions of surplus flowers. (3) Resource limitation may influence the evolution of nectar-deception, perhaps the most prevalent form of deceit pollination in orchids, in which flowers appear to contain nectar but, in fact, do not (Ackerman, 1985).

Seedling establishment in an epiphytic orchid: an experimental study of seed limitation

Will increased fruit and seed production in a severely pollination-limited orchid stimulate population growth? We tested whether safe sites for germination and seedling establishment are limiting for the twig epiphyte, Tolumnia variegata, by manipulating fruit set and monitoring subsequent seedling establishment for two seasons (1991–1992, 1992–1993). In the Cambalache Forest Reserve of Puerto Rico, we established 36 plots along a transect. Each plot consisted of nine trees. A center tree was designated as the site for attaching Tolumnia and manipulating fruit set. The other eight potential host trees were 1–3 and 3–5 m from the center tree in each of the cardinal directions. A 1-m length of stem 1 m from the ground was monitored for recruits on each of the nine trees of 24 fruit-enhanced plots and 12 controls (23 and 13, respectively for the 1992–1993 season). Fruit enhancement plots were divided among two treatments: one-fruit and five-fruit additions for the 1st year and one to five and more than five fruits for the 2nd year. Availability of suitable host species was not limiting. T. variegata showed little specificity for host tree species, good host trees and shrubs were common, and there was no evidence that the orchid had a preference for small branches, despite possessing the entire suite of characteristics thought to respresent “obligate” twig epiphytes. Fruit enhancement increased seed rain and seedling establishment consistently in only the high-fruit treatment plots. Most recruitment occurred near fruiting plants. Over the 2-year period, mortality was 18% for adults and 85.5% for the 1991–1992 cohort of recruits. Net recruitment was positive for both the treatment (average = 1.74) and control plots (average = 0.67). Seedling establishment at our study site was not microsite-limited. If selection for increased pollinator attraction occurs, then an increase in seed output should result in population growth.

Long-term ecology of euglossine orchid-bees (Apidae : Euglossini) in Panama

SummaryAbundance patterns during 6–7 years and orchid visitation were determined for 51 species of the 57 local euglossine bees. Male bees were counted at 3 chemical attractants presented in the same manner each month. Sites were separated by 75 km but included wet Atlantic forest at 500 m elevation, moist forest at 180 m near Barro Colorado Island, and cloud forest at 900 m near the Pacific ocean. 1. From 15 to 30 euglossine species of 4 genera were active in each month and site; monthly species number and general bee abundance were positively correlated. Many species had 3 annual abundance peaks (range 1–4) and were active throughout the year, but peak annual abundances rarely occurred during late wet or early dry seasons. In contrast, Eufriesea generally were present as adults only 1–2 months in a year. 2. Euglossine populations were exceptionally stable. Species at each site were more stable than any known insect population, and stability and abundance were positively associated. However, year-to-year population stability and the degree of seasonality were not correlated. Among the three sites, the more diverse (species rich) bee assemblages displayed lower stability; these were the wetter and upland sites. 3. The most abundant bees visited more orchid species. Eg. and El. each visited and average of 4 orchid species (range 0–13); Ex. and Ef. visited 0–3. Stable populations did not visit more or fewer orchid species than did unstable populations. 4. Less than 68% of species at each site visited orchid flowers; less than a few dozen of the 100–800 bees counted in a day carried orchid pollinaria. Over 20% of the euglossine species never were seen with pollinaria at any site and probably seldom visit orchids in central Panama. 5. Most bee species visited 1 or no fragrance orchids in a given habitat. Orchids tended to utilize common pollinators that seldom included more than 1 species, and they utilized stable or unstable, seasonal or aseasonal bees. However, the most stable and abundant bee, Eg. imperialis, rarely pollinated orchids; fewer than 10 of ca. 20000 bees carried pollinaria. 6. Orchids may interact primarily with discrete seasonal bee population peaks-probably the emerging adults. Although specialized orchid preferences are implicated for species that visit few or no local orchids but pollinate other species and carry pollinaria in other areas, euglossine bees do not need orchids to survive or reproduce.

Differences in mycorrhizal preferences between two tropical orchids

Orchids parasitize their mycorrhizal fungi and are dependent on them for seed germination. Controversy reigns over how specific the mycorrhizal association is in tropical species. Although there is little experimental evidence to support any viewpoint, some variation is known to exist. We compared mycorrhizal specificity and performance in two phylogenetically related epiphytic orchids from Puerto Rico, Tolumnia variegata and Ionopsis utricularioides (Oncidiinae) by integrating two techniques: phylogenetic analysis of mycorrhizal fungi based on nuclear ribosomal internal transcribed spacer (ITS) sequences, and symbiotic seed germination experiments. Most of the mycorrhizal isolates from T. variegata fell into four different clades of Ceratobasidium, while most of those from I. utricularioides were restricted to a single clade of the same genus. Seeds of T. variegata germinated equally well with fungi from both T. variegata and I. utricularioides, but seeds of I. utricularioides germinated significantly better with its own isolates. Seeds of I. utricularioides germinated and developed faster than those of T. variegata. Both the molecular phylogeny and the seed germination experiments showed that T. variegata is a generalist in its association with fungal symbionts. In contrast, I. utricularioides is more specialized and more effective at exploiting a specific fungal clade. Our data are consistent with the theoretical trade‐offs between specialized and generalized interactions.

Limitations to Sexual Reproduction in Encyclia krugii (Orchidaceae)

In the Guanica Forest Reserve, Puerto Rico, Encyclia krugii lacks pollinator rewards, is visited infrequently and matures few fruits. Hand-pollinations showed that flowers are self- incompatible. In 22 censuses, I monitored the number of flowers, fruits, vegetative shoots, and active inflorescences that were produced over two years. During the first 20 censuses, all open flowers of 43 randomly assigned, experimental plants were cross-pollinated. Fruiting of these plants was compared with 37 open pollinated controls to determine if plants were pollination limited. Fruit set estimates were 4% for controls and 8% for experimentals. Fruit production was limited by several factors. More than half the fruits initiated by both groups either aborted or were eaten. At least some of the abortions were probably caused by incompatible pollinations. Inflorescences that produced fruits subsequently produced fewer flowers than those that produced no fruits. This short- term adjustment of reproductive effort suggests that resources do limit the number of flowers and fruits produced. Moreover, experimentally augmented fruit production had some long-term effects on subsequent growth. Thus, sexual reproduction by E. krugii is limited by a combination of factors: incompatible pollinations, resource constraints, predation, and paucity of pollinations. In most orchids studied thus far, paucity of pollinations and resource availability are the dominant con-

Widespread mycorrhizal specificity correlates to mycorrhizal function in the neotropical, epiphytic orchid Ionopsis utricularioides (Orchidaceae)

Tropical orchids constitute the greater part of orchid diversity, but little is known about their obligate mycorrhizal relationships. The specificity of these interactions and associated fungal distributions could influence orchid distributions and diversity. We investigated the mycorrhizal specificity of the tropical epiphytic orchid Ionopsis utricularioides across an extensive geographical range. DNA ITS sequence variation was surveyed in both plants and mycorrhizal fungi. Phylogeographic relationships were estimated for the mycorrhizal fungi. Orchid functional outcomes were determined through in vitro seed germination and seedling growth with a broad phylogenetic representation of fungi. Most fungal isolates derived from one clade of Ceratobasidium (anamorphs assignable to Ceratorhiza), with 78% within a narrower phylogenetic group, clade B. No correlation was found between the distributions of orchid and fungal genotypes. All fungal isolates significantly enhanced seed germination, while fungi in clade B significantly enhanced seedling growth. These results show that I. utricularioides associates with a phylogenetically narrow, effective fungal clade over a broad distribution. This preference for a widespread mycorrhizae may partly explain the ample distribution and abundance of I. utricularioides and contrasts with local mycorrhizal diversification seen in some nonphotosynthetic orchids. Enhanced orchid function with a particular fungal subclade suggests mycorrhizal specificity can increase orchid fitness.

Geographic and Seasonal Variation in Fragrance Choices and Preferences of Male Euglossine Bees

Variation in choices and preferences for fragrances are examined for a large portion of the euglossine bee assemblages of Barro Colorado Island (BCI) and Cerro Campana, Panama. Census data were obtained from a year-long baiting program utilizing 16 chemical attractants. Interand intraspecific variation in fragrance choices occurred. Within sites, fragrance choices of species overlapped considerably, but each species was attracted to a unique set of baits. For some species, choices of infrequently visited chemicals differed among sites. The five baits preferred by most species were the same for both sites. Geographic variation in intraspecific fragrance preferences occurred for seven of eleven species that were abundant at both BCI and Cerro Campana. Seasonal variation in fragrance preferences occurred for 16 of 21 species on BCI. The lack of seasonal shifts was not related to the number of fragrances to which a species was attracted. Indirect evidence exists for age-dependent shifts in chemical preferences by bees. Substantial intraspecific variation in fragrance preferences occurs and is related to some combination of season, age, and geography. The degree to which natural fragrances are available and genetic differences among individuals affect variation in choices and preferences for fragrances remains unknown. THE NATURAL HISTORY OF THE NEOTROPICAL Euglossini (Apidae), commonly known as orchid bees, is unusual and male behavior is both perplexing and intriguing. Male bees utilize the same nectar sources as females (Ackerman 1985), and their foraging behavior at these flowers is similar (Ackerman et al. 1982). Males differ from females in that they have no role in nest construction or in cell provisioning. Furthermore, male bees collect fragrances from a variety of sources, including rotting fruits and wood, tree bark, exposed roots, and flowers (Ackerman

Orchid-phorophyte relationships in a forest watershed in Puerto Rico

Orchid diversity, distribution and host specificity were examined in a tropical water- shed in the Luquillo Experimental Forest of Puerto Rico. Eleven orchid species occur in the area. The low diversity is attributed to island isolation and large-scale hurricane disturbances. Pleurothlal- lis ruscifolia and Maxillaria coccinea were by far the most abundant species in the area and occurred on the largest number of host species and host zones. None of the orchids were host specific or host zone specialists although preferences for hosts and vertical host zones were encountered. Only 8.2% of the 426 trees and shrubs and 24.4% of the 45 species surveyed were orchid phorophytes (= hosts). Examination of host distribution by diameter at breast height (DBH) showed that 80.5% were greater than 16 cm DBH. Orchid species in the area tend to occur on rough bark hosts, but their preferences are not statistically significant. Guiarea guidonia (Meliaceae) and Dacryodes excelsa (Burseraceae) are the two most important orchid hosts in our study site com- prising 62.9% of all host trees. Careful management of these two tree species is suggested, since these species may be crucial to the maintenance of orchid abundance and diversity in the area. RESUMEN. Se examin6 la diversidad, distribuci6n y especificidad de especies epifiticas de