Showing papers in "Biotropica in 1985"

Breeding Systems in the Flora of a Tropical Deciduous Forest in Mexico

Abstract: Floral sexuality was characterized for most of the flora (708 species) of a forest reserve at Chamela, Jalisco, and compatibility was determined for a sample of hermaphroditic species. Nine types of floral sexuality were found; the frequency of these types was related to life form and systematics. Monostylous hermaphroditism was typical of epiphytes (96%) and herbs (82%) but less so among vines (68%), shrubs (62%), and trees (54%). Dioecy increased across these groups (4%, 1%, 10%, 14%, and 24%, respectively). Monoecy was concentrated in Euphorbiaceae and Cucurbitaceae; Boraginaceae and Rubiaceae predominated among the few distylous species. Contrasting vegetation types had similar frequencies of sexuality types. Compatibility tests showed reduced fruiting from selfcompared to cross-pollination in 94 percent of 33 species from 14 families, and 76 percent were considered self-incompatible. Intraspecific variation was not salient, but in some cases congeners were notably divergent.

A disturbing synergism between cattle ranch burning practices and selective tree harvesting in the eastern Amazon.

Abstract: The combination of pasture burning to control weeds and selective timber harvesting is resulting in the spread of wildfires in Amazonian rain forests near Paragominas, Para, Brazil. We visited 15 cattle ranchers who had permitted selective logging on their land and found that eight of the affected forests had burned shortly after logging. In selectively logged, high forests, fire-induced mortality was greatest for trees less than 10 cm DBH, whereas in low and second-growth forests, nearly all trees were killed. Fires affect regeneration mechanisms such that the more severely forests are burned, the more complex and prolonged will be the pattern of succession. OF ALL THE OPTIONS FOR ECONOMIC DEVELOPMENT in the Amazon region, the selective harvest of valuable timber species on a rotational basis is one of the most ecologically sound (Goodland 1980, Fearnside 1983). Selective tree harvesting, when done carefully, usually creates three to six canopy gaps per hectare which are akin to natural forest tree fall disturbances. Regeneration following natural tree falls is rapid (Brokaw 1984, Whitmore 1983). Pre-existing seedlings and saplings (advance regeneration) and fast-growing pioneer trees, usually originating from seeds buried in the soil, dominate the regrowth. Moreover, because only the nutrient-poor boles are removed (i.e., the relatively nutrient-rich twigs, leaves, and roots remain) and because regeneration is rapid, nutrient loss is probably inconsequential. Although regeneration occurs rapidly without further disturbance, selective tree harvesting leaves the forest in an open, fuel-rich, fire-prone state. We have observed that fires set to control weeds in degraded pastures in eastern Amazonia commonly spread into adjacent forests. Fire readily spreads through exploited (selectively logged) forests causing extensive damage, but fires reaching the edge of unexploited forests quickly die out. Thus, the effects of forest timber removal and pasture burning interact to produce more detrimental effects than either pro-

Tree Mortality in Mangrove Forests

Abstract: Twenty-eight worldwide reports of massive mangrove tree mortalities are reviewed. Massive mortality is defined as tree mortalities that occur in response to rapid environmental change and affect all size dasses. Massive mortality occurs in addition to normal tree mortality. Normal tree mortality was described using structural data from 114 mangrove stands. This mortality is density dependent, follows orderly time dependent patterns dictated by stand maturation (related to average tree diameter), and usually occurs in the smaller diameter size classes. Disease and other biotic factors do not appear to be primary causes of massive mangrove mortalities. Instead, these factors appear to attack forests weakened by changes in the physical environment. Mangrove environments are dynamic and cyclical and mangrove associations adapt to such environments by both growing and dying fast. Mangrove species' characteristics such as the capacity to produce large quantities of propagules that take advantage of dispersal agents, sharp species zonations, and even-aged populations contribute to the rapid growth-mortality cycles in mangroves. Humans may tilt the balance towards higher mortality rates by introducing chronic stressors that inhibit regeneration mechanisms. MANGROVE FORESTS ARE DYNAMIC ECOSYSTEMS. Factors such as sedimentation rates, soil subsidence, freshwater run-off, tidal forces, and changes in sea level influence mangrove growth and survival. In response, mangrove species arrange themselves in zones that reflect geomorphologic and hydrologic gradients (Lugo 1980). Consequently, significant changes in environmental conditions are generally followed by alterations in the vigor or zonation of vegetation. These alterations may include widespread tree mortality.

Spatial variation in the structure and litterfall of a Sumatran rain forest

Abstract: In this paper an explanation for the spatial variation in the structure of a Sumatran rain forest is suggested. The forest area consisted of a series of topographical units (terraces, mountain slopes) of increasing age and decreasing pH and fertility. With increasing soil age the production of leaf litter and fruit decreased whereas the proportion of mature-phase forest increased and gap cover decreased. There is variation in the growth strategies of trees depending on soil fertility; this results in higher tree longevity on poor soils, and hence in forest with fewer gaps and with more large trees, than on rich soils. Thus, Janzen's (1974) hypothesis of the effect of soil fertility on the prevalence of anti-herbivore defenses and fruiting periodicity is extended to include aspects of forest structure. THERE IS LONG-STANDING EVIDENCE of spatial variation in the floristic composition of the Malesian rain forests (van Steenis 1935; Richards 1952; Ashton 1964, 1982; Whitten 1982). More recently, much attention has been given to the temporal variation in floristic composition in both the Malesian (Poore 1968) and the neotropical (e.g., Hartshorn 1980) regions. The interest in forest dynamics has led to a number of studies in which structure and productivity are linked with secondary successional processes (e.g., Odum et al. 1970). For forests in the wet and moist tropical life zone it is established that leaf area index and leaf biomass are restored within a few years of the start of secondary succession (Golley 1978, Snedaker 1980) and that the rate of increase in total biomass and litter production slows down after some 20 years (Brown and Lugo 1982). In the mature phase, however, net primary production and perhaps litterfall tend to increase again (Whitmore 1975). There are no such dear patterns of spatial variation in structure and productivity of the forests for the wet and moist tropical life zone. The general concensus is that soil factors affect neither litter production (Tanner 1980, Leigh and Windsor 1982, Proctor et al. 1983b) nor structure and biomass (Proctor et al. 1983a). The aim of the present study is to compare the structure, dynamics, and litter production of a number of habitat types within one small area of tropical rain forest containing soils of various ages and fertilities. The results suggest that variation in soil fertility may be responsible for the observed spatial variation in structure and productivity. STUDY AREA The upper half of the course of the Alas river lies in the Semangko Rift zone, a graben running the full length of Sumatra and flanked by the Barisan mountain range (Verstappen 1973). The Ketambe field station (3041'N, 97?39'E) lies just below the confluence of the fast-flowing and rocky Alas and Ketambe rivers (Fig. 1) at an altitude of about 3 50 m. The graben is only a few km broad here and contains a number of accumulation terraces (van Beek 1982). The study area (about 200 ha) includes these terraces and the lower slopes of the mountain range (see fig. 16 of Rijksen (1978) for altitudes) which consists mainly of crystalline metamorphic schists, alternating in places with hard limestones of late Paleozoic origin (van Beek 1982). When the Toba volcano erupted some 75,000 years ago (Francis 1983) the exit from the rift valley was blocked by inflowing volcanic material. As a result the valley filled with ignimibrites and became a lake. The erosion base has since been lowered discontinuously by tectonic activity and, due to variation in the weathering resistance of the layered sediments, resulting in a series of terraces. It is estimated that the highest of these terraces became dry about 10,000 yr ago (van Beek, pers. comm.). The area is considered sacred by the local inhabitants of the Alas valley and, with the possible exception of one small part of the lowest terrace near the field station, has not been disturbed by human activities other than the collection of forest produce. Some people say that this part of the lower terrace was prepared for cultivation of rice around 1940 but was abandoned after an outbreak I Received 5 December 1983, revision accepted 14 June 1984. 196 BIOTROPICA 17(3): 196-205 1985 This content downloaded from 157.55.39.255 on Sat, 16 Apr 2016 04:28:25 UTC All use subject to http://about.jstor.org/terms

The palm community in a forest of central Amazonia, Brazil.

Abstract: The palm community in a forest of central Amazonia clearly shows three zones according to the hydromorphic condition of the soil: well-drained soils of the upland forest, poorly-drained soils of a transition zone, and water-logged soils of the seasonal swamp forest. The community is remarkable in three aspects: its size (with 2122 palms/ha, and the highest density on water-logged soils); its great diversity (32 species/1.2 ha), which also depends on the hydromorphic condition of the soil; and its variety of biological forms with characteristic acaulescent palms in the understory, monocaulous and multiple-stemmed palms in the upper understory, and arborescent palms only reaching the canopy in the seasonal swamp forest. PALMS ARE AN ABUNDANT and characteristic component of the forests of central Amazonia. They are found in all levels of the forest, from the understory to the canopy, on all soils and topographic sites, and exhibit a variety of growth forms. However, as Moore (1973a) lamented, our ignorance of the biology and ecology of palms is almost complete. Corner's marvelous book (1966) represents the cornerstone of our knowledge of the family in these ways. Several early works dealt specifically with the palm flora of Amazonia. In fact, since the early studies of this subject (Poiteau's (1822) history of palms in French Guiana; Wallace's famous book (1853); Spruce's works (1871, 1908); the numerous publications of Martius (18231853)3, Drude (1876-1908)3, Barbosa-Rodrigues (18751907)3, and Burret (1928-1956)3, only Macbride's work (1960) in Peru and Wessels-Boer's (1965) and (1971) in Suriname and Venezuela, respectively, have contributed significantly to the taxonomy of amazonian palms. Recently, Balick and de Silva (1982) established the state of systematic collections in regional herbaria and showed the necessity to intensify them. Works relating to the ecology of palms are even rarer (see Discussion). Thus, we initiated a series of ecological investigations on Amazonian palms in 1980 at the National Institute for Amazonian Research (INPA). This paper presents a community-level study of palms in central Amazonia and describes the distribution and abundance of the most important species in relation to topography and soils.

Soil Response to Forest Clearing in the United States and the Tropics: Geological and Biological Factors

Abstract: Twenty-six cross-sectional and time series studies of soil properties under natural forest and altered vegetation in the United States and ten tropical countries were examined to determine the changes associated with forest clearing. Organic C, total N, exchangeable Ca, Mg, and K, cation exchange capacity, available P, bulk density, and pH were considered. After forest clearing, only bulk density and available P tended to return to predearing levels. Losses of organic C, total N, and cation exchange capacity were 50 percent larger in soils developed on highly weathered parent materials in the tropics than in the same soils in the United States or in soils developed on young parent materials. Differences in soil response to clearing were related to local effects of temperature, rainfall, vegetation type, and soil acidity on organic matter decomposition. THIS STUDY TESTED THREE HYPOTHESES: 1) The removal of natural forest vegetation causes changes in soil properties; 2) changes in soil properties associated with changes in land use are long term; and 3) changes in soil properties associated with changed land use are more severe in the tropics or in soils developed on old parent materials. The study arose from the continued debate among soil scientists, ecologists, and land use planners over the adverse and beneficial effects of forest dearing, especially in developing countries in the tropics. On one side of this controversy are those who emphasize the drastic environmental changes associated with forest clearing in the tropics, frequently implying that they are irreversible (GomezPompa et al. 1972; Eckholm 1976, 1982; Myers 1976; Hecht 1982; N. J. H. Smith 1982). On the other side are those who have emphasized the variability of soils and therefore of the consequences of forest dearing (Jordan and Herrera 198 la, b), the potential for productive management of even very poor soils after deforestation (Sanchez et al. 1982), or the benefits from forest clearing, especially food production. At the heart of this debate is the question of whether soils in tropical climates respond more drastically to forest dearing than soils in temperate climates. This paper attempts to quantify the changes in soil properties that are associated with forest clearing and to identify factors associated with dramatic soil responses to clearing.

Relationships between soil and vegetation in a tropical rain forest in French Guiana

Abstract: We describe a new method to observe and map morphological and functional characteristics of soil. Drainage and surface hydromorphy were mapped in eleven plots including 16.8 hectares. In the same plots, all trees over 20 cm DBH were inventoried. Structural data of the vegetation are correlated to soils characteristics as well as the frequencies and densities of 32 taxa.

The influence of cyclones on the dry evergreen forest of Sri Lanka

Abstract: I assessed damage caused by a cyclone in November 1978 to a dry evergreen forest in Sri Lanka. Damage included defoliation, breakage of twigs, branches and trunks, tree fails and post-cyclone tree mortality. Within the 3 km2 study area there was a trend for increasing damage with forest height. Tree species in the upper forest layers had significantly more falls and post-cyclone mortality than trees in the more sheltered subcanopy and shrub layers. Mortality was significantly greater among trees which lost 40 percent or more of their branches and trunks than among trees with lesser crown damage. Defoliation and twig loss were extreme in the discontinuous upper layers and probably contributed to the greater tree mortality evident there. Overbrowsing by folivorous primates after the cydone may have contributed to the demise of some preferred feeding trees. Total tree loss was 46 percent from the upper forest layers, 29 percent from the subcanopy, or 40 percent from all tree layers. Five upper layer tree species were subject to tree losses of 80 to 100 percent representing a virtual elimination of 22 percent of species from the upper layers or 12 percent from the forest. Dry evergreen forest formerly covered 80 percent of the island's land area, and often has been described as "old secondary climax" in recognition of past disturbance. Extrapolations from meterological data indicate that 33 to 44 percent of the range of dry evergreen forest may be subject to cyclone damage per century. The species composition of dry evergreen forest on a local and wider geographical scale typically is fairly uniform in the subcanopy layer, but variable in the upper layers. I suggest that recurrent cyclone damage may be an important factor contributing to succession in dry evergreen forest and to the variation in species composition of the upper layers. TROPICAL CYCLONES OR HURRICANES are known to cause extensive damage to human life and property in the coastal regions of the Bay of Bengal in the Indian Ocean. Although natural plant and animal populations are also subject to cyclone damage, published reports addressing this phenomenon are rare and concern areas outside the Indian region (e.g., Webb 1958, Whitmore 1974). A recent cyclone which swept through the dry evergreen forest of Sri Lanka provided an opportunity to study the effects on phenology, forest structure and floristic composition. Quantitative data concerning these features were collected before the cyclone, allowing their comparison before and after the cyclone. The objectives of this paper are threefold: first, to assess the nature and extent of cyclone damage to the forest; second, to examine the causes of tree mortality in the 42 months following the cyclone; and third, to consider the effect of recurrent cyclones on the geographic variation in species composition of the dry evergreen forest of Sri Lanka. The long-term recovery of the forest from cyclone damage will not be examined in this report. Data concerning phenology are considered briefly, however, in order to show the immediate effects of the cyclone. NOMENCLATURE AND NATURAL HISTORY The forest types of Sri Lanka have been classified according to several schemes (e.g., Chapman 1947; Holmes 1956; Koelmeyer 1957, 1958; Gaussen et al. 1964; Fernando 1968). Mueller-Dombois (1968) offers an evaluation of these schemes and additional refinements have been made recently by Greller and Balasubramaniam (1980) and Greller et al. (1980). The forest type considered here is commonly known in Sri Lanka as the "dry-zone forest." It has also been referred to as "Tropical Dry Evergreen Forest" (Champion 1935, Chapman 1947, Holmes 1956), "Tropical Dry Mixed Evergreen Forest" (de Rosayro 1950; Koelmeyer 1957, 1958; Andrews 1961), "Semideciduous Forest" (Gaussen et al. 1964, 1965), "South Tropical Moist Deciduous Forest" (Chapman 1947, Koelmeyer 1957), "Tropical Lowland Seasonal Rain Forest" (Perera 1975) and "Semi-evergreen Forest" (Dittus 1977). The last label had been selected to conform to Walter's (1971) worldwide classification of tropical vegetation types, although Walter does not specifically refer to the dry-zone forest of Sri Lanka. In the literature the dry-zone forest is most frequently referred to as "Dry Evergreen Forest" and therefore I will follow this tradition here. Taxonomic nomenclature follows Abeywick-rama (1959) and Dassanayake and Fosberg (1981, 1983). I Received 2 August 1983, revised 11 May 1984, accepted 14 June 1984. 2 Field address: Smithsonian Institution Primate Project, 4/4 Galkanda Road, Anniewatte, Kandy, Sri Lanka. BIOTROPICA 17(1): 1-14 1985 1 This content downloaded from 207.46.13.127 on Wed, 12 Oct 2016 04:32:28 UTC All use subject to http://about.jstor.org/terms

The influence of fruit size and structure on composition of frugivore assemblages in New Guinea.

Abstract: In a mid-elevational rainforest in New Guinea, size and structure of fruits influenced feeding visits by various species of frugivorous birds: large (diam. > 12 mm) structurally unprotected fruits were taken mostly by fruit-pigeons and bowerbirds, structurallyprotected fruits were taken mostly by birds of paradise, and small, structurally-unprotected fruits were taken by nearly all species. Fruit-pigeons as compared with birds of paradise have short, thin bills; small feet; and fewer behaviors for reaching and handling fruits. Birds of paradise use complex food-handling techniques for removing fruits in capsules and other protective structures that are similar to techniques used to capture insects hidden in bark and dried foliage. Absence of large structurally-unprotected fruits in the diets of birds of paradise is not explained entirely as an effect of fruit size. Specialization for frugivory in the entirely frugivorous fruit-pigeons has resulted in exploitation of specific fruit types (large and small structurally-unprotected fruits) rather than all fruit types. Most fruiting plants were visited by a subset of frugivores, as influenced by fruit size and structure. Attracting specific feeding assemblages might be regarded as an adaptation of the plant for enhancing particular patterns of seed dispersal. However, other selective pressures also operate on the evolution of fruit size and structure. BIRD-DISSEMINATED PLANTS HAVE EVOLVED fruits with attributes attractive to birds, such as bright coloration, easy accessibility, convenient size, and nutritional reward (Ridley 1930, Snow 1971, van der Pijl 1972, Howe and Smallwood 1982, Denslow and Moermond 1982, Willson and Thompson 1982, Stiles 1982, Janson 1983). Similarities among fruits of bird-disseminated species have been explained as evolutionary convergence on the narrow range of adaptations competitively effective at attracting birds and dispersing seeds (Snow 1971, McKey 1975, Howe and Estabrook 1977, Stiles 1980). The low diversity of fruit types has itself been invoked as an explanation for the low diversity of frugivores compared with that of insectivores (Snow 1971, Snow and Snow 1971). The question of how fruit characteristics influence usage by frugivorous species has been examined through studies of frugivore autecology (e.g., Snow 1962, Snow 1970, Beehler 1983a, Wheelwright 1983), studies of feeding assemblages at one or a few trees (e.g., Leck 1971; Ricklefs 1977; Howe 1977, 1981; McDiarmid et al. 1977), and theoretical studies. For tropical bird-disseminated species, theoretical studies have emphasized two contrasting situations: large nutritious fruits consumed by large "specialized" frugivores which eat mostly fruit, and small fruits of poor nutritional quality taken by smaller "opportunistic" frugivores which also forage for insects or nectar or both (Snow 1971, McKey 1975, Howe and Estabrook 1977). Studies comparing avian assemblages at fruit-bearing plants are few (Terborgh and Diamond 1970, Ricklefs 1977, Skutch 1980, Howe 1981), and none sampled a broad spectrum of food plants at a single

Litter fall of Avicennia germinans L. in a one-year cycle in a mudflat at the Laguna de Mecoacan, Tabasco, Mexico

Abstract: The fall of four morphological components of Avicennia germinans L. was measured at monthly intervals along a transect in a monospecific mangrove near the inlet of the Laguna de MecoacAn, Tabasco, Mexico. Estimated mean total annual litter fall was 614.4 g.m-2 yr-1. Seasonal fluctuations of mean water level, insolation, temperature and evaporation were highly intercorrelated and were summarized in a principal component axis that explains 82 percent of their variability. The results suggest that litter fall in A. germinans responds to the environmental variables already mentioned but is statistically independent from local rainfall. Each litter component (leaves, flowers and reproductive parts, branches and woody parts, and propagules) shows a distinctive sequence, and significant correlations were obtained when comparing the principal component axis against leaf and propagule fall. Phenologic variation in litter fall as an adaptive response to environmental changes is discussed. DESCRIPTIONS OF THE PHENOLOGY OF MANGAL WOODY SPECIES have been done in Florida (Gill and Tomlinson 1971, Pool et al. 1975); Thailand (Christensen 1978, WiumAndersen and Christensen 1978); Malaysia (Gong et al. 1980); Veracruz, Mexico (Rico 1979); and northeastern Australia (Williams et al. 1981). Emphasis, however, has been given in most cases to the whole community, or to species of Rhizophora. Recent reports have made clear that there are different species responses to the mangal environment and dimate (e.g., Wium-Andersen 1981, Williams et al. 1981, Snedaker and Brown 1982). This paper describes the annual phenological variations in litter fall of Avicennia germinans L. in a mudflat environment and analyzes their relation to environmental parameters, such as temperature, insolation, water level, and rainfall.

Seed Dispersal, Seed Predation, and Juvenile Mortality of Aglaia sp. (Meliaceae) in Lowland Dipterocarp Rainforest

Abstract: Black hornbills (Anthracoceros malayanus) appear to be the principal long-distance seed dispersers of Aglaia sp. (Meliaceae) at Pasoh Forest Reserve in Peninsular Malaysia. The squirrel Callosciurus prevostii removed some of the large seeds at least as far as adjacent crowns and sometimes dropped them after consuming only the orange, oily sarcotesta. It chased other squirrel species, which are probably seed predators, and hornbills out of the fruiting crown. Seeds on the ground beneath the parent crown were removed more rapidly than those farther away by rodent and (possibly) phasianid seed predators. Sitophilus sp. (Curculionidae) was also an Aglaia seed predator or parasite. Aglaia juveniles grew slowly under closed canopy, and small ones (<50 cm in height) had only 9.7 percent mortality per year during a 4-year period. Seedling survival was positively size dependent. For unknown reasons, small seedlings near the parent had a higher death rate than those at a greater distance; this was apparently a long-term pattern, because large seedlings and saplings did not occur within 10 and 35 m, respectively, of the parent tree's base. For this species an advantage of seed dispersal is avoidance of disproportionate seed and seedling mortality near the parent. Although Aglaia grows much faster in gaps than under closed canopy, it may require several episodes of growth in successive gaps before becoming reproductively mature. IN THEIR REVIEW OF SEED DISPERSAL, Howe and Smallwood (1982) remarked that little is known about its ecological and evolutionary advantages to the plant. Here we combine observations on seed dispersal and predation and juvenile mortality and spatial distribution to show how dispersal helps the tropical tree Aglaia sp. (Meliaceae) avoid seed and seedling mortality that is greater near the parent than farther away. Webb et al. (1967), Janzen (1970), and Connell (1971) suggested that such a mortality pattern, generated by seed predators, herbivores, or allelopathy, would promote the maintenance of high tree species richness in tropical forests. More recently, Connell (1978, 1979) has withdrawn his support for this hypothesis because field studies have shown that seed and seedling mortality is not invariably higher near conspecific adults. However, Clark and Clark's (1984) review of 24 data sets on mostly neotropical, woody plants showed that most evidence indicates either densityor distance-dependence in progeny mortality, as originally predicted by Janzen and Connell. Hubbell (1979, 1980) challenged the Janzen-Connell hypothesis on empirical and theoretical grounds, daiming the latter to demonstrate that spacing between conspecific adults could contribute little to the maintenance of high species richness. His theoretical analysis has been questioned by Becker et al. (1985), who consider the issue unresolved for nonequilibrium

Leaf litter disappearance rates in Puerto Rican montane rain forest

Abstract: The time course of leaf litter disappearance of six rain forest tree species was monitored for 32 weeks at the El Verde lower montane rain forest study site in northeastern Puerto Rico. Dacryodes excelsa, Sloanea berteriana, and Drypetes glauca were chosen to represent primary climax species, while Cecropia peltata, Inga vera, and Ixora ferrea were selected to represent secondary successional species in this forest. The study was designed to elucidate differences in nutrient release associated with the successional position of different tree species to determine what properties of leaf structure influenced the observed rates of nutrient disappearance. The string, nonconfined, tethered leaf method was employed. Dry weights and concentrations of nitrogen, phosphorus, potassium, calcium, and magnesium were determined over a period of 32 weeks. The secondary canopy species exhibited slower decay rates than did the primary species. The same pattern emerged between the secondary and primary understory species. The interacting effects of the leaf structural characteristics such as percentage lignin and percentage fiber correlated most strongly with observed decay rates. Nutrients were considered important in establishing organic matter resource quality but did not appear to influence decay rates; instead, nutrient dynamics reflected intervals of element immobilization, mineralization, and importation. Secondary species exhibiting specific combinations of structural properties may act to conserve nutrients by their slower rates of leaf litter disappearance. THE BREAKDOWN AND MINERALIZATION OF DEAD ORGANIC MATTER through the actions of decomposer organisms and fluctuating environmental conditions results in the progressive disappearance of litter from the forest floor. Mineralization and the subsequent translocation of biologically important elements among the various ecosystem components insures maximal reutilization and minimal cycling loss. The study of this process, particularly as it pertains to evergreen tropical forests, contributes to the understanding of nutrient relations in these ecosystems. The present study determined the time course of leaf litter disappearance from six different tropical forest tree species and discusses properties of leaf structure that influence the observed rates. Prior research has centered on specific aspects of nutrient release and the factors affecting the overall rate of litter disappearance. Such contributing factors have included differences among species, season, and location (Wiegert & Murphy 1970), plant material (Jenny et al. 1959, Shanks & Olson 1961, Olson 1963, Cornforth 1970), environmental conditions (Bocock & Gilbert 1957, Nye 1961, Hopkins 1966, John 1973), or microflora and -fauna (Witkamp 1963, 1966, 1969; Witkamp & Crossley 1966; Edwards et al. 1970). Other efforts have considered nutrient release in terms of the structural properties of the decaying substrate. These studies have concentrated on the structural effects and modifications of organic materials during the disappearance process and on the influences that individual substrate constituents might have on decomposer communities and overall decay rates (King & Heath 1967, Bailey et al. 1968, Minderman 1968). Disappearance dynamics provide an insight into the mechanism of nutrient release. Rainfall and temperature in tropical rain forest ecosystems do not affect the decomposition rate of leaf litter (La Caro 1974). Constant rapid rates exist because the biological activity of decomposer organisms is never hindered or inhibited. The rate of disappearance is ultimately dependent on the nature of the organic matter resource. Each plant part, due to differences in chemical and physical properties, will yield different disappearance rates (Odum 1970). Litter resources can be further subdivided to distinguish between different species or forest types. Categorizations distinguishing species by successional life-support strategies may help identify physiological mechanisms in specific ecosystems. Identification of some specific nutrient cycling mechanisms may shed light on the dynamics of nutrient maintenance in tropical rain forest ecosystems. Secondary successional vegetation is characterized as having the ability to live in disturbed forest sites under widely ranging environmental conditions. Secondary species live comparatively short lives, are fast growers, good seed disp rsers, and generally shade tolerant. Secondary vegetation has also been reported to decay at a faster rate than primary because of the differing structural characteristics of their decaying materials (Ewel 1976). In contrast, climax species generally exhibit opposite properties (Richards 1952, Smith 1970). Leaf litter exhibiting a specific combination of structural characteristics that tends to imI Received 8 September 1983, revision accepted 20 August 1984. BIOTROPICA 17(4): 269-276 1985 269 This content downloaded from 207.46.13.149 on Mon, 03 Oct 2016 06:15:48 UTC All use subject to http://about.jstor.org/terms mobilize nutrients for longer periods of time in disturbed forest sites may provide a valuable nutrient conservation function.

Herbivory and defense in pioneer, gap and understory trees of tropical rain forest in French Guiana.

Abstract: Levels of leaf herbivory (percentage area loss) in primary forest understory, a pioneer plot, and vegetation in forest gaps were compared at St. Elie, French Guiana. Leaf laminas were analyzed for nitrogen, total phenol, and condensed tannin concentrations, as bulked vegetation samples irrespective of species. Mature leaves of the shaded forest understory were more heavily (greater than threefold) grazed than those on pioneer trees and small trees which grow up from the saplings in lighted gaps, while young leaves did not differ significantly in levels of defoliation between vegetation types. Defoliation levels averaged 5.5 percent on mature leaves. Herbivory on mature leaves was significantly negatively correlated with total leaf phenols. In the context of the optimal defense hypothesis, these chemicals may be deterring herbivory on pioneer and gap leaves, but a more likely primary cause is that total phenols are higher in plants in direct light than in those in shade conditions. FEENY (1976) AND RHOADES AND CATES (1976) have put forward an optimal defense hypothesis which predicts that persistent, slow-growing, and "apparent" K-selected plants will defend their tissues against herbivory by investing in compounds of high molecular weight such as polyphenolics. These "quantitative" defenses will reduce the digestibility of the plant tissue by forming tannins which decrease the availability of nitrogen to the predator. Tannins would be expected to deter specialists, although generalists would tolerate this low-quality diet. In contrast, ephemeral, fast-growing, "unapparent" r-species such as pioneers, for which energy for competition and reproduction is supposedly at a premium, are expected to coevolve less metabolically costly defenses as toxins of low molecular weight (e.g., alkaloids). Since these latter, "qualitative" defenses require special detoxifying enzymes, they effectively deter all but the few herbivores specialized to circumvent the defenses of a particular species. To this must be added a temporal dimension: ephemeral plants are often difficult to locate by specialists, and the actual incidence of predation in nature may be kept low simply by this passive mechanism of escape in time (Fox 1981). A model for the coevolution between predators and plant defenses has been put forward by Rhoades (1979). In the tropical rain forest two distinct physionomic strategies have been observed (Richards 1952, Whitmore 1975). The first type consists of tree species which germinate in the shade and are shade tolerant in the early stages of their sapling growth. Apart from tree species which complete their life cyde totally in the undercanopy, these shade-tolerant species will respond to light later and grow up into the main canopy. The second type consists of species which germinate in the light and are light demanding in their growth. These conditions occur when one or more trees open a gap in the upper canopy, allowing, especially in the larger gaps, a pioneer flora to establish. Pioneers are relatively short lived compared with the main forest trees, are fast growing, vigorously competitive, and have a high reproductive output of widely dispersed seeds (Whitmore 1975). If natural selection has operated to produce two such different growth strategies, coadapted traits (Stearns 1977) which act as defenses to predation are expected in accordance with the optimal defense hypothesis. This paper reports a test of this combined hypothesis in the rain forest of French Guiana, South America, by comparing four different vegetation types: shaded primary understory saplings, saplings of the understory recently exposed to light in small and, separately, large gaps created by tree fall, and a wholly pioneer flora establishing in a large area cleared by man. Large and small gaps differ in that large gaps have a number of pioneer species present.

Bat pollination of an old world Heliconia

Abstract: In the Solomon Islands Heliconia solomonensis with pendent, green, inconspicuous inflorescences is pollinated by the macroglossine bat Melonycteris woodfordi. The green and white, odorless flowers open in the evening and produce copious (over 200 pl/night), moderately concentrated (26-27% sucrose-equivalents), sucrose-dominant (s/(g + f) = 1.150-1.567) nectar. Self-pollinations by hand reveal that H. solomonensis is probably self-incompatible. Melonycteris is a solitary forager that visits inflorescences of Heliconia in a sequential fashion. The bat lands on an inflorescence and clings with feet and wing claws to the revolute margins of the cincinnal bracts. Nectar is taken from open flowers by the bat in such a fashion that pollen is deposited on the underside of its neck and chin. This report is the first documentation of flower visitation by the genus Melonycteris, the first account of bat pollination in the Solomon Islands, and the first description of bat visitation to the flowers of Heliconia, a primarily neotropical, hummingbird-pollinated genus.

Seasonality and diversity of mayfly adults (Ephemeroptera) in a «nonseasonal» tropical environment

Abstract: Fluctuations in abundance of adult Ephemeroptera were analyzed from a year of daily light-trap collections made at Miramar, a relatively nonseasonal locality on the Atlantic coast of northwest Panama. Most taxa occurred year-round and some of these showed within-year fluctuations in abundance that may be seasonal. The species in the genus Leptohyphes showed a pronounced lunar periodicity. A comparison of this sample with other samples of Ephemeroptera suggests that mayflies in the tropics may not be more diverse than in the temperate zone. SEASONAL FLUCTUATIONS IN THE ABUNDANCE of tropical insect populations have been systematically investigated only in recent years. The absence of a cold season allows many species to be present during much of the year, although the alternation of wet and dry seasons imposes its own constraints. Seasonal fluctuations in terrestrial tropical insect populations have been documented in a number of studies (Galindo et al. 1956; Owen 1969; Owen and Chanter 1970, 1972; Wolda 1979b, 1980a, b, 1982, 1983a; Denlinger 1980); for further references see Wolda (1978). It has been shown that the seasonal range--the length of time per year that a species is present as an active adult (Wolda 1979a)-is, on the average, longer in the tropics than in the temperate zone. A number of cases of insect population fluctuations have been documented in \"nonseasonal\" tropical habitats (W olda and Fisk 1981, Wolda and Galindo 1981)-that is, in habitats without a clear alternation of wet and dry seasons. Insect species in such areas with a less seasonal weather pattern tend to be less seasonal, but not necessarily aseasonal (Wolda 1983a). Studies of seasonal fluctuations in the abundance of tropical Ephemeroptera and other aquatic insects are scarce. Most have been done in African lakes (Hartland-Rowe 1958; Tj0nneland 1960, 1961; Chutter 1970; Corbet et al. 1974) and streams (Hynes 1975, Zwick 1976, Statzner 1976, Kopelke 1981), some in Asian streams (Bishop 1973, Bright 1982), and one in Panama (McE!ravy et al. 1982). Although studies of life histories and seasonal patterns of temperate Ephemeroptera are numerous (see Clifford 1982 for review), work on tropical members of this important part of the aquatic fauna has only recently 1 Received 3 April 1984, revision accepted 16 July 1984. 330 BIOTROPICA 17(4): 330-335 1985 begun. To date, seasonaliry studies have been published on the mayflies of the Ethiopian region (Tj0nneland 1960, 1961; Corbet et al. 1974; Kopelke 1981) and Malaya (Bishop 1973). All were conducted in areas with a distinct alternation of wet and dry seasons; this paper presents the first data on seasonaliry of mayflies in a tropical area which lacks a distinct dry season. The diversiry of aquatic tropical insects, relative to their temperate counterparts, is a subject of controversy (Stout and Vandermeer 1975, Stanford and Ward 1983). Tropical Trichoptera seem to be slightly more diverse than temperate ones (McElravy et al. 1982). The available information on Ephemeroptera will be discussed in this paper. STUDY SITE AND METHODS The study site was near the village of Miramar (9°N, 82°15'W) in the province of Bocas de! Toro, Republic of Panama. Miramar is on the Atlantic coast in the southwest corner of the Laguna de Chiriqui. Insects were collected with a light-trap, Pennsylvania rype, with a 15-W blacklight. The insects fell through a funnel in a collector with preservative (FAA, a mixture of alcohol, formaldehyde, and acetic acid). The trap was located in a row of trees among pastures, at the foot of a wooded slope. A small stream traversed the pastures about 50 m from the light-trap. The trap was operated nightly from 6 to 10 P.M. from 15 November 1978 to 13 November 1979. Local help was hired to collect the daily samples, which were then taken to the Smithsonian Tropical Research Institute where they were sorted to order. Due to logistic problems data are missing from 12-30 June, from 1225 July, and from 28-31 July 1979; therefore, in this TABLE 1. List of genera, number of individuals in each genus, and, among the males, number of species of mayflies caught in a light-trap during one year m 1978/ 79 in Miramar, Panama. Taxa Species Individuals

Germination response in a large-seeded Neotropical tree species, Gustavia superba (Lecythidaceae).

Abstract: This six-month study examined percent germination, days until germination, and early seedling survival of Gustavia superba (Lecythidaceae) through field experiments in a tropical moist forest in Panama. The study had three goals: (1) to examine whether germination is a factor determining density of seedlings at three forest sites with different ages; (2) to test whether burial of seeds affects germination; and (3) to test if microsite conditions associated with light gap, gap-edge, or forest understory influence germination. The experiment, which was designed to fit a three-way ANOVA model, used three study sites, three microsites, two burial treatments, four replicates and 25 seeds in each replicate (1800 seeds total). Although seedling densities of G. superba at the forest sites varied inversely with age of forest, there was no evidence that germination success contributes to this pattern. Results indicated that this species germinates readily under most conditions: the overall mean was 85.4 percent and only six of 72 plots had less than 70 percent germination. Germination percent was reduced and mean number of days until germination was increased in the largest light gap. However, percent germination and days until germination did not differ among sites when only seeds in the understory microsite were compared. Burial of seeds and microsite conditions had minor effects on germination. Seedling mortality occurring soon after germination was not large (5.7%). It is concluded that G. superba is not a tropical species whose seeds germinate in response to a light gap but its seeds do germinate successfully under a range of microsite conditions. ALTHOUGH PLANT POPULATION SIZES can be affected by mortality at many stages, the seed to seedling transition generally is considered to be one of the most vulnerable. Mortality agents include pathogens, insect and vertebrate predators of seeds, as well as abiotic factors affecting germination levels. All these factors are intertwined in affecting seedling establishment. As a result, controlled experiments are required to identify the separate contribution of each factor. The focus of this paper will be to discuss the role of germination per se in seedling establishment of Gustavia superba (Kunth) Berg (Lecythidaceae), a

Adjusting Foraging Effort to Resources in Adjacent Colonies of the Leaf- Cutter Ant, Atta colombica

Abstract: Two adjacent colonies of leaf-cutter ants Atta colombica were studied to examine the influence of resource availability on foraging. A colony located in older, secondary forest used 130 different resource species and had a dietary diversity equal to 30.7 equallyused species. This colony showed a cyclical annual pattern of gradually changing diet on a Bray-Curtis ordination. An adjacent colony in younger, secondary forest used 103 plant species and had a dietary diversity of only 11.8 equally-utilized species. Foraging by this colony was governed by the availability of a small number of high quality resource species. Tree species eliciting a strong recruitment of foragers from the colony were denser in the younger forest suggesting that the colony in the younger forest was specializing in a richer resource environment. Regression analyses indicate that the location of small patches of high quality resources determines the spatial distribution of foraging in the younger forest but much less so in the older forest. A weak correlation between maximum harvest rates and the synchrony of leaf production may indicate that the phenology of leaf production serves as a predator-avoidance mechanism. ONE OF THE CORNERSTONES OF FORAGING THEORY has been the premise that species adjust their foraging behavior to local conditions under the pressure of natural selection. Animals that identify productive patches of their ranges and concentrate their foraging effort on these small areas will forage most efficiently and be favored on an evolutionary time scale. For individual foragers, this is done with such behavioral mechanisms as changing turning rates and move distances (Pyke et ail. 1977, Krebs and Davies 1978); for social insects, recruitment behavior and trail systems play an important role (H6lldobler 1976, Holldobler and Lumsden 1980, Shepherd 1982). Animals appear to rank resources and concentrate their foraging effort on the best items, generally supporting the prediction (MacArthur and Pianka 1966; Schoener 1969, 1971) that foragers will specialize in times or areas of plenty and generalize in times or areas of scarcity (see Krebs 1979 for a recent review). A changing resource environment may also modify forager behavior. For example, Bernstein (1975) found that Veromessor pergandei switches from an individual to a group searching method during times of resource scarcity. Acromyrmex versicolor uses individual foraging on small dispersed items, but group foraging on large clumped items (Gamboa 1975). The interaction between a forager and the resource pool is thus very dynamic and most species will probably defy any simple categorization of their foraging behavior. The complexity of resource environments, as well as formidable practical difficulties, often prevents the kind of close examination of resources required to demonstrate fine-scale adjustment of a forager's behavior to changes in its resources. Leaf-cutter ants, however, seem ideal subjects for studying the interaction of foragers and resources. A colony forages as a single unit whose foraging effort is easily quantified by the numbers and distribution of workers out of the nest. Workers gather forage from large, stationary, identifiable resources (trees and lianas) and are easy to observe while they forage. For these reasons, leaf-cutter ants were used to answer the following questions: 1) How do differences in available resources affect long-term foraging behavior? 2) How does plant phenology affect exploitation by these generalist herbi-

How Fast and Why Do Germinating Guanacaste Seeds (Enterolobium cyclocarpum) Die Inside Cows and Horses

Abstract: Scarified and unscarified guanacaste seeds were subjected to in vitro fermentation using either rumen or caecal inoculum to test their germination response to the fermentation processes of cattle and horses. The fermentation was terminated at various times after inoculation, and the seeds planted under greenhouse conditions to evaluate viability. One day of exposure to either rumen or caecal in vitro fermentation killed 90 percent of the scarified seeds, and none survived longer than two days. No unscarified seeds germinated in the flasks. Scarification appears to be a critical process affecting the survivorship of seeds in the guts of large, mammalian herbivores. Because functional differences in the foregut between ruminants and nonruminants impose different chewing patterns and passage rates on these groups, we postulate that these factors are responsible for the lower proportion of scarified seeds in the feces of ruminants. In the large mammalian herbivores, nonruminants function as seed predators while ruminants are better disposal agents for guanacaste seeds. GUANACASTE (Enterolobium cyclocarpum: Leguminosae) is A LARGE TREE of the Pacific coastal lowlands of the drier

The role of nectar and oil in the pollination of Drymonia serrulata (Gesneriaceae) by Epicharis bees (Anthophoridae) in Panama

Abstract: The pollination biology of Drymonia serrulata was studied in central Panama. D. serrulata produces a small number of flowers daily for several months. Flowers are dichogamous with a one day male phase followed by a one day female phase. The average volume of nectar produced by a Drymonia flower over the two-day period was 264.4 Al (SD = 98.8, N = 4), a value much higher than has been reported for other bee-pollinated flowers. D. serrulata is completely self-compatible. Fruit set for between and within plant pollinations in greenhouse grown plants was 67 percent. Autogamy is precluded by the temporal separation of anther dehiscence and stigma receptivity. Glandular trichomes inside the corolla secrete an oil which is deposited on the anthers and transferred to the thorax of visiting bees. This oil facilitates pollen adhesion. Only Trigona pallens (Apidae) actively collects this oil, but it is ineffective as a pollinator. The main pollinators of Drymonia serrulata in Panama are Epicharis monozona nd Epicharis rustica (Anthophoridae). Visitors that remove nectar, but are ineffective pollinators include seven Euglossa spp. and three species of matinal butterflies (Hesperiidae). THE GESNERIACEAE IS A LARGE PREDOMINANTLY TROPICAL FAMILY consisting of about 120 genera and 2500 species of herbs, shrubs, lianas or epiphytes (Cronquist 1981). Although much is known about the floral morphology and presumed pollination of many of these species, few if any detailed pollination studies have been published. Wiehler (1978) estimated that 60 percent of Neotropical Gesneriaceae are pollinated by hummingbirds, 30 percent by nectar-seeking euglossine bees and about 10 percent by bats, butterflies, moths, flies or scent-collecting male euglossines. The actual number of field observations on which these estimates were based, however, was not presented. Dressler (1968) reported pollination of Gloxinia perennis and Drymonia tumrialvae by scent collecting Eulaema mefiana and Eulaema nignifacies males, respectively, while Wiehler (1977) reported his own observations of male and female Euglossa gibbosa visiting Drymonia ecuadorensis in Ecuador and observations by Dressler of Euglossa gorgoensis and Euglossa nigrosignata visiting Drymonia aciculata in Panama. Dressler (pers. comm.) has also observed Euglossa asarophora (female) visiting Drymonia dressleri in Panama and Eulaema speciosa (males) visiting an unidentified species of Dtymonia in Ecuador. The only non-euglossine flower visitors observed visiting Drymonia were Thygater sp. (Anthophoridae) to an undetermined Gesneriaceae species in Veracruz, Mexico (Dressler, pers. comm.). In Costa Rica, Feinsinger (pers. comm.) has found Drymonia rubra to be pollinated by hummingbirds. Recent studies have greatly increased our understanding of Neotropical bee pollinators through broad surveys (Heithaus 1979a, b, c; Roubik 1979) or detailed studies of specific taxa. These indude the Meliponini (Uohnson and Hubbell 1974, 1975; Roubik 1978, 1980, 1982a, b), Euglossini (Janzen 1971, 1981; Williams and Dodson 1972; Williams 1978; Schemske 1981; Dressler 1982; Ackerman 1983a, b; Ackerman et al. 1982) and some Anthophoridae (Frankie et al. 1976, 1979, 1980, 1983; Frankie and Haber 1983; Roubik et al. 1982). Very little has been published on the biology of Epicharis (Anthophoridae), a group of medium to large bees, except for Neff and Simpson's (1981) consideration of their oil collecting structures, ome information on nesting biology (Vessey-Fitzgerald 1939, Camargo et al. 1975, Roubik and Michener 1980) and floral visitation records from Costa Rica (Heithaus 1979a, c), and South America (Ducke 1901, Schrottky 1908, Vessey-Fitzgerald 1939, Vogel 1974, Roubik 1979). Female bees in several genera of Anthophoridae, induding Epichatis, visit members of the Malpighiaceae to collect pollen and floral oils (Vogel 1974; Simpson and Neff 1981, 1983; Steiner, pers. obs.). Both males and females, however, must visit nectar-producing flowers as well. Here I describe the interaction between several bee species and Drymonia serrulata (Jacq.) Mart., a nectar flower that deposits small quantities of oil on the dorsum of its pollinators and thus facilitates pollen transfer. MATERIALS, METHODS AND STUDY SITE This study was conducted on Barro Colorado Island (see Croat 1978 for climatic and floristic information) and nearby areas of the mainland of central Panama, primarily BIOTROPICA 17(3): 217-229 1985 217 I Received 1 December 1983, revision accepted 14 June 1984. 2 Present address: National Botanic Gardens of South Africa, Kirstenbosch, Private Bag X7, Claremont 7735, Republic of South Africa. FIGURE 1. Diagram of Drymonia serrulata in: A. male phase, and B. female phase. Scale lines are 1 cm. between 13 Sept and 19 Dec 1979. A few additional observations were made in Oct 1977 and in Aug 1981. Two principal study sites induded the Barro Colorado Island (BCI) shoreline just north of the dock area and a roadside about 2 km northwest ofGamboa. Minor sites were Frijoles tation, near the transisthmian railway line, and Pipeline Road, about 3.8 km northwest ofGamboa. These sites are referred tosubsequently as BCI, Gamboa, Frijoles and Pipeline Rd. All of the study sites were on the edge of or within 20 m of second growth forest sites capable of supporting tropical moist forest. Drymonia serrulata isa dimbing, often hemiepiphytic, deciduous vine that occurs in tropical forests from Mexico to Brazil and Bolivia (Croat 1978). In areas near the Panama Canal, it is common in and around second growth forest and often in more open disturbed areas as well (Steiner, pers. obs.). On BCI, it is common along the shoreline, but it is also found occasionally inthe forest canopy (Croat 1978). D. serrulata flowers during the rainy season from about mid-July tomid-December (Steiner, pers. obs.). Fruits mature from August o February. FLORAL MORPHOLOGY.-Drymonia serrulata has large, mildly fragrant symmetrical, tubular flowers (Figs. 1 and 2); they are cream-colored with violet o purple blotches on the corolla lobes and have similarly colored nectar guides on the waxy lower corolla surface. These guides converge on a groove or channel formed by the constriction of the lower sides of the corolla (Figs. 2a, b). This channel forces a bee to pass directly under the anthers or stigma. Access to the nectar chamber behind the ovary must be gained through one of two small openings (< 1 mm2) that flank the top of the ovary. At the bottom of the chamber is a large nectary that surrounds the dorsal base of the ovary. Since there are only two very small openings at the top of the chamber, nectar does not spill 218 Steiner out, even though the base of the flower is above the mouth of the corolla (Fig. 1). Floral development was followed from early bud stages to abscission of the corolla. Flowers were checked throughout he day and at night to determine the timing of anthesis and subsequent changes in the positioning of the sexual parts. Nectar was sampled from four bagged flowers from a single plant (two separate days) for the duration of the secretion period (35 h). Nectar was sampled only after the flowers had opened, even though nectar secretion begins earlier. To remove nectar, a small piece of tissue was removed from the nectar chamber; due to the thick waxy structure of the wall, this process did not result in premature senescence. Nectar was collected in calibrated micropipets and the concentration of each sample was measured in wt/total wt equivalents of sucrose (percent total dissolved solids) with a temperaturecompensated refractometer. The first sample was made at 0530 (local time) for two of the flowers and at 0630 for the other two. Nectar was removed hourly subsequent to the initial sample except during the night. The only night sample was taken at 2330. The amount present at that time and at 0530 was divided by the number of unsampled hours prior to those times to arrive at an average hourly secretion rate for the night period. Nectar was spotted onto Whatman #1 filter paper strips and later analyzed for sugar and amino acid composition and for the presence of lipids, proteins and phenolics using the methods of Baker and Baker (1977, 1982, 1983). Flower opening is a gradual process that can take several hours to complete and is often quite variable in time of initiation. However, virtually all flowers opened before dawn. Flowers usually last two days and are strongly protandrous. For the entire first day they are in the male phase (Figs. la, 2b, 3a and b). In this stage, the anthers occupy the center of the corolla tube forcing the bees to s ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~A ^ -s-P~W FIGURE 2. A. and B. are dose-ups of D. serrulata flower: A. Female phase showing nectar guides and constriction of lower corolla surface; B. Staminate phase. Scale lines are 0.5 cm. C. Early bud stage of D. serrulata flower with a portion of the corolla cut away. Scale line is 0.5 cm. D. Close-up of 2C. showing deposition of oil from glandular trichomes onto the anthers. Scale line is 1 mm. contact them when probing for nectar. The anthers are attached to the filaments in a manner which causes them to pivot and become inverted when pushed from below. When the anthers are inverted, pollen falls out of the pollen sacs and is usually deposited on the thoracic dorsum of the bee. Flowers remain open during the night between day one and day two, entering the female phase between 2330 and 0530. The filament bases become coiled and draw the empty anthers to the base of the ovary away from the area occupied by a visiting bee. The style elongates and moves the two-lobed stigma into the position previously occupied by the anthers (Figs. lb and 2a). In this position, the stigma lobes can contact the pollen-covered thorax of an incoming bee. Female phase flowers usually abscise late in the aftemoon of day two. Both male and female phase flowers are bo

Microbiology of the Larval Provisions of the Stingless Bee, Trigona hypogea, an Obligate Necrophage

Abstract: Five species of the genus Bacillus were the only microbes found in the larval provisions of Trigona hypogea, a bee for which dead animal tissue has replaced pollen as the sole protein source. Bacillus pumilus was the most frequent isolate followed by B. megaterium, B. subtilis, B. circulans, and B. licheniformis. These bacteria were metabolically active and produced many enzymes including proteases, amylases, and esterases. They may have a fundamental role in the metabolic conversion, fermentation, and preservation of the food of perennial colonial insects that rely on stored food in tropical environments. THE MELIPONINES ARE A PANTROPICAL GROUP of eusocial, stingless bees native to tropical forest. They live in perennial colonies and have a single, physogastric, flightless queen. Nests of stingless bees are usually subterranean or aboreal and are occasionally associated with ant or termite nests. The entrance is either a small opening just large enough to allow one bee at a time to pass or a more elongated tube-like entrance that may have a flared opening. Although colonies generally consist of fewer than several thousand adults, foraging bees visit a wide variety of resources, and large amounts of food are stored in the nests (Schwarz 1948, Wille 1983, Roubik 1983). Plant resins, other exudates, secreted wax, and mud used for nest construction, combined with the extreme selectivity of nesting sites, make the nest waterproof and highly resistant to attack from predators. Microbes associated with bees may be parasites, commensals, or mutualistic. Some can infect bees, and spoilage microbes can destroy pollen stores (Batra et a/. 1973). Also, the presence of microorganisms in the nests of perennial colonial insects that rely on stored food in humid, tropical environments may be fundamentally important in the metabollc conversion or preservation of the food. Stingless bees continuously forage for nectar and pollen in tropical forests, and colony food stores are kept in separate but grouped dusters of storage pots. The pots are built of secreted wax mixed with plant resins, the latter possibly serving as a biocidal agent (Michener 1974, Roubik 1983). Thus, these bees resemble some groups of bumble bees (Bombus spp.) with regard to food storage behavior but, unlike Bombus, construct or remove large numb rs of storage pots within the nest and do not rely on the availability of vacated brood cells for modification to food storage containers. Trigona workers provision cells with pollen and honey mixed with glandular secretions prior to oviposition by the queen. Unlike Apis honey bees, stingless bees use mass provisioning, and each larva is provided with the total amount of food required for adult development before the cell is sealed. Resources collected by foraging Trigona workers are much more diverse than those collected by Apis workers and include pollen, nectar, fungal spores, exudates from homopteran insects, plant resins containing terpenes and other sap exudates, mud, vertebrate feces, and carrion from vertebrate carcasses. True obligate necrophagy was first reported by Roubik (1982) for Trigona (Trigona) hypogea Silvestri. Dead animal tissue has replaced pollen as the sole protein source for these unique bees. Trigona hypogea bees lay down odor trails between the nest site and the animal carcass, which can be located within a matter of hours, and utilize massive recruitment to send hundreds of foragers to the site. Bees on vertebrate carcasses were observed to actively attack and drive off flies and wasps that had begun to feed. Thus, dipteran oviposition and larval competition for the carcass could not occur. Roubik (1982) has observed these bees feeding on the carcasses of toads, frogs, snakes, lizards, birds, fish, and monkeys. Trigona hypogea bees collect nectar and produce and store copious amounts of honey within large colonies (Roubik 1983). Unlike all other known Trigona species, they never collect pollen from flowers. The nests lack stored pollen in either the storage pots or in the provisions 28 BIOTROPICA 17(1): 28-31 1985 I Mention of a proprietary product or company name does not constitute an endorsement of this product by the U.S. Department of Agriculture. 2 Received 2 December 1983, accepted 24 February 1984. This content downloaded from 157.55.39.231 on Wed, 05 Oct 2016 04:23:37 UTC All use subject to http://about.jstor.org/terms in larval cells. Morphological adaptations of T. hypogea indude reduced corbiculae on the hind legs and strong mandibles with five enlarged teeth (Roubik 1982). Bees can be observed to bite off, masticate, and consume large pieces of musde and other tissue from the carcasses. They do not carry externally unmodified pieces of animal flesh back to the nest. The surface of the animal carcass often appears to glisten, and this is apparently due to salivary and other secretions produced by the feeding worker bees. We (Roubik and Buchmann unpublished) hypothesize that the younger worker bees further process the pre-digested material in their massive hypopharyngeal glands by a mechanism analogous to that of nurse bees of Apis mellifera which consume pollen and produce brood food or royal jelly via the hypopharyngeal glands. These younger Trigona bees then transfer the brown glandular secretion to large cerumen storage pots which are sealed when filled. When a section of brood comb has many empty larval cells, workers remove glandular material from the larger pots to provision individual brood cells. The queen then oviposits on the surface of the provisions, and the cells are capped. The brownish liquid secretion obtained from larval cell provisions and storage pots is highly viscous and has a pH of 3.0-4.0 which is very similar to the harsh, acidic Apis royal jelly. Other analyses (Buchmann, Schmidt, Roubik, and Law unpublished) have revealed a high protein concentration (-20%), large numbers of amino acids and hydrocarbons, and lipid and carbohydrate levels very similar to those found in royal jelly of Apis. Considering the rapidity with which dead animal flesh putrefies in the tropics due to microbial activity, the metabolic conversion of this rich protein foodstuff is the only viable strategy for long-term food storage by these bees. Many nests have been examined to date, and no evidence of rotten carrion has ever been found within a larval cell or large storage pot. Because of the distinctive fermentation odor, larval provisions of T. hypogea were examined for bacteria, yeasts, and molds to characterize microbes that might play a role in metabolic conversion or preservation of food. MATERIALS AND METHODS On February 18, 1982, brood provisions of T. hypogea were collected from a nest in Panama. The nest was opened by removing a side of a log, the batumen was carefully removed, the newest comb area and sealed storage pots were located and exposed, and the larval cells and intact storage pots were rapidly dissected indoors. The cells were opened with sterile forceps, and the contents of each cell were removed with a separate sterile capillary pipet. The pipets were sealed with Critoseal, placed in sterile test tubes, frozen immediately, and kept frozen during transport and until analyzed in Tucson. Two 20-30 ,ul samples of pooled larval provisions were inoculated onto duplicate plates of blood agar, nutrient agar (Difco), TYG agar (40 g Difco tryptic soy agar, 3 g yeast extract, 10 g glucose, and 1 liter distilled water), Czapek solution agar (Difco), and YM-1 agar (Wickerham 1951) to isolate bacteria, molds, and yeasts. TYG agar was developed by one of us (B.J.L.) and is an excellent growth medium for bacteria associated with bees and plants. To test for the presence of anaerobes, duplicate tubes of thioglycollate medium without indicator (Difco) were inoculated. One plate or tube of each medium was incubated at 250 and the other at 3 7C under aerobic conditions. During a 2-week incubation period, plates and tubes were periodically examined for microbial growth, and any colonies that developed were restreaked onto plates of TYG agar to test for pure cultures. These plates were incubated at 37?C. Cell suspensions from each microbial colony were stained by the Gram method, and all were found to belong to the genus Bacillus. This was confirmed by microscopic examination of the provisions. In this manner, the size, shape, and location of the spores within the sporangia were noted, and the morphology of the vegetative cells was determined. The organisms were maintained on TYG agar slants and were tested and identified according to Gordon et al. (1973) except that motility was determined in motility test medium (BBL) rather than microscopically. Additional information on enzymatic activity of each of the isolates was obtained by testing for 19 enzymes using the API ZYM system (Analytab Products) according to the manufacturer's directions.

Seed Weight as a Function of Life Form, Elevation and Life Zone in Neotropical Forests

Abstract: Seeds from over 365 species of eight plant families have been excised, and dry weights determined, from fruit previously collected from Costa Rica, Panama, and Peru, and deposited in the U.S. National Museum of Natural History. This study was undertaken to determine whether data from tropical plants were consistent with temperate zone work showing correlations of seed size with life form and certain aspects of the physical environment. In order to control for intraspecific variation, a seed weight class system (Baker 1972) was used. As seen in temperate zone studies, tropical species show significant differences in seed weight means among trees, shrubs, and herbs (largest to smallest). Tropical epiphytes, however, show a bimodal pattern; though seeds from some plant families (e.g., Orchidaceae) are tiny, many epiphyte seeds (primarily from Melastomataceae, Gesneriaceae and Bromeliaceae) were equivalent to, or larger than, those of shrubs. Seeds from tropical herbaceous vines were comparable in weight to those of herbs, while seeds of woody lianas (all from Bignoneaceae) were comparable to trees in weight. Seed weights analyzed by elevation and life zone showed no conclusive patterns. SEED WEIGHT VARIES AMONG SPECIES over ten orders of magnitude (Harper 1977), though it is usually conserved within a species at the expense of seed number. In order to begin identifying the major factors which determine the size of seeds, Baker (1972), Salisbury (1942), and Stevens (1932, 1957) carried out large scale surveys of seed weights relative to environmental conditions, successional stage, and plant life form. Baker (1972) and many others (Singh 1972, Cook 1972, Schimpf 1977, Seavey et al. 1977, Sorenson and Miles 1978, Yeaton 1978, Werner and Platt 1976, Hurka and Benneweg 1979, Schat 1981) found a tendency for species whose seedlings are exposed to drought to have heavier seeds. Other factors correlated with large seeds are shade (Salisbury 1942, Baker 1972, Carlquist 1966), intense competition (Harper and Clatworthy 1963, Harper et al. 1970, Stebbins 1976, Williams et al. 1968), and late successional stages (Salisbury 1942, Yeaton 1978, Opler et al. 1977). Baker reported that mean seed weight decreases with altitude in California. Baker (1972), Stevens (1932, 1957), Salisbury (1942), and Silvertown (1981) have also indicated that seed weight is associated with life form; mean seed weights increase from herbs to shrubs to trees in a given community. Some data on Central American legumes (Janzen 1969) and Pacific Island communities (Carlquist 1966) have been published. Studies of intraspecific seed size patterns indude those of Howe and Richter (1982) and Janzen (1978, 1982). Extensive community surveys by Herbert Baker, Gordon Frankie, and Paul Opler in Costa Rica have been published in only summary form (Opler et al. 1977, 1980). Many questions can be asked about seed weight in tropical communities. How will seed weights from important life forms such as vines, lianas, and epiphytes compare with those of trees, shrubs, and herbs? Is large seed size associated with plant biomass? Is long life span, as suggested by Silvertown (1981), an important factor in allowing a plant to develop large seeds? Studies of seeds from epiphytes and lianas, which are part of the canopy and have long life spans but relatively low biomass, would help answer this question. Is moisture availability as important a selective force in tropical communities as in temperate locations such as California, or is length of growing season more important? The tropical dry forest, for example, has a shorter growing season for most plants than the wet forest, yet seedlings and mature plants are exposed to drought for a portion of the year. The shorter growing season in the dry forest could select for smaller seeds, especially among annuals, but the fact that seedlings of both annuals and perennials are exposed to drought might select for larger seeds. Finally, can one expect a decrease in seed weight with altitude in the tropics as Baker found in California? There are some obvious and striking differences between tropical and temperate mountains. Janzen (1967), Leigh (1975), Grubb (1977), Holdridge et al. (1971), and Terborgh (1977) have reviewed changes in the plant community and in the physical environment with altitude in the tropics. Until one reaches extreme heights (over 3600 m) the growing season does not change with altitude, though average temperature does. Some workers (Janzen 1973, Janzen et al. 1976, Terborgh 1977) have postulated that net plant production might increase with elevation from sea level to approximately 1500 m. Janzen (1973) and Terborgh (1977) present a number of reasons why this might be so. Thereafter, net production should decrease I Received 3 October 1983, revised 12 March 1984, accepted 14 March 1984. 32 BIOTROPICA 17(1): 32-39 1985 This content downloaded from 157.55.39.180 on Tue, 27 Sep 2016 05:35:58 UTC All use subject to http://about.jstor.org/terms with altitude. Can average seed weight of a community or series of populations be correlated with net production? If so, mean seed weight would be expected to increase to 1500 m, followed by a decrease with altitude. The purpose of this study was to attempt to examine some of these questions by determining seed weights from a large number of species which induded as many life forms and life zones (Holdridge et al. 1971) as possible. In order to maximize the amount of data gathered, seeds were collected from herbarium specimens.

Site-Fidelity and Survival Rates of Some Montane Forest Birds in Malawi South-Central Africa

Abstract: From 1972 to 1982, 2030 forest birds of 33 species were banded in a study area on the Nyika Plateau, Malawi/Zambia, in south-central Africa. The movements and survival of these birds, particularly the starred robin Pogonocichla stellata, were monitored during three breeding seasons, 1979-1982. Overall, average movement recorded was 165 m (range 0-3125 m, N = 613). Five species of passerines moved from the largest forest into the smallest patches (about 2.5 km). However, most territorial adults were strictly sedentary, especially during the breeding season, when less than one-third of the movements recorded exceeded the diameter of a territory (N = 320). Some adults held the same territories for up to 10 years, with 39 percent known to be present for at least two consecutive seasons and 19 percent over three years or more (N = 1127). Most species maintained long-term pair bonds, but all female Pogonocichla left the study area in the nonbreeding season and returned regularly to the same territory each year (annual rates of 67-78%, N = 59). Intensive investigation of territories in forests around the main banding area showed that only 1.5 percent (N = 246) of the adult Pogonocichla sampled were banded individuals which had dispersed. Average annual mortality among territorial adults was between 13.2 and 22.2 percent for the four best-studied species (three robins and one monarchine flycatcher). Their expectancy of further life (4.0-7.1 years) agrees with the maximum longevities recorded: four passerine species had individuals which attained 9.7-10.8 years of age (the maximum possible in this study). THE NYIKA PLATEAU IS A HIGHLAND AREA immediately west of Lake Malawi, between 10 and 1 1?S (Fig. 1), with 1800 km2 above 1800 m altitude. Most of the plateau lies within Malawi, with a small segment in Zambia. More than 95 percent is montane grassland, but parts are forested; the evergreen fragments vary greatly in size, many being very small. The area is described in detail by Dowsett-Lemaire (1983). The smallest of the forests, many less than 1 ha in area, contain a large proportion of the bird species breeding on the Nyika. With often only one or two pairs of each species present, they provide ideal conditions for the frequent monitoring of their reproductive status, movements and population dynamics. This paper examines results from a study conducted on the southwestern Nyika concerning the mobility and mortality of forest birds, topics that have hitherto not been studied intensively in tropical Africa. These topics are of special relevance to the claims of Diamond and Hamilton (1980) that presentday distributions of forest birds on the continent may be the result of extensive movements between forests since the last glaciation. Diamond and Hamilton argue that forest vegetation and forest birds have spread in recent times from a few interpluvial refugia. In contrast, White (1983) suggests that the origins of the complex Afromontane forest flora are too diverse to indicate almost total disappearance during the phase of maximum glaciation, followed by rapid recolonization. Paleobotanical evidence is scanty and uncertain and cannot be used to prove the timing of past changes in forest distribution (White 1981). However, knowledge of the present mobility of forest birds gained in this study may help to test the hypothesis of long-distance intermontane dispersal. STUDY AREA AND METHODS The main study area on the Nyika Plateau consisted of a 25-ha arm of Chowo Forest (90 ha) and 12 isolated for st remnants 2.5 km distant in the Zovochipolo area, initially in the same valley and in sight of Chowo (Fig. 2). These patches ranged from 0.16 to 1.20 ha in size and totaled 7 ha. In 1973 I established a numbered grid of mistnet lines in Chowo. I banded birds there during 3-week visits in 1973, 1975 and 1977. Birds were also banded in three of the Zovochipolo patches in 1975 and 1977. During my residence on the Nyika from October 1979 to March 1982, birds were banded in Chowo and Zovochipolo. Mistnet lines were spaced evenly throughout the 25-ha arm of Chowo (as far as the topography of the forest allowed) and totaled 1900 m in length. Mistnetting sessions there between 1979 and 1982 each lasted three weeks; 300 m of net were erected daily and were moved to a fresh area every three days. The total number of meter-hours (one meter length of standard mistnet erected for one hour) varied from 13,830 to 39,870 per session (mean 26,038 ? 7405 m, N = 9). Mistnetting in Zovochipolo occupied a similar amount of time. The Chowo study area was surveyed and a detailed I Received 2 September 1983, revised 12 March 1984, accepted 14 March 1984. 2 Present address: FitzPatrick Institute, University of Cape Town, Rondebosch 7700, South Africa. BIOTROPICA 17(2): 145-154 1985 145 This content downloaded from 157.55.39.35 on Mon, 29 Aug 2016 05:22:35 UTC All use subject to http://about.jstor.org/terms

Some floral nectar-sugar compositions from Durango and Sinaloa, Mexico

Abstract: Determination des glucides du nectar de 34 especes et 18 familles de plantes tropicales et subtropicales. Discussion sur la relation avec les types de pollinisateur

Seasonal Predation by Merlins on Sooty Mustached Bats in Western Puerto Rico

Abstract: Predation by Merlins (Falco columbarius) on the insectivorous bat Pteronotus fulginosus was studied for two consecutive winters at a cave in Aguadilla, Puerto Rico. Falcons were present from October to April, reaching maximum densities in January and February. Predation efficiency of falcons was greater at the more exposed of two flyways used by bats. The falcon using the most easily observed perch captured an average of 5.1 bats/day. In the first season falcons took 1800 bats from the more exposed flyway or 1.5 percent of the standing crop as estimated photographically. By the end of the second year the exposed flyway was no longer used by bats.

Albumin Evolution in Tropical Poison Frogs (Dendrobatidae): A Preliminary Report

Abstract: The Dendrobatidae are a predominantly South American group with a minor center of diversity in southern Central America. A study of albumin evolution was initiated in order to help elucidate phylogenetic lineages within the family and to estimate their time of arrival in Central America. Using antisera to serum albumin from Phyllobates terribilis and Dendrobates auratus, the albumins of all five species of Phyllobates, 12 species of Dendrobates, and two species of Colostethus were compared by the quantitative immunological technique of micro-complement fixation. The results accord with the previous recognition of Phyllobates as a monophyletic group defined by the presence of batrachotoxins, unique skin alkaloids used by the Choc6 tribes of western Colombia as a potent dart poison. Speciation events leading to the living species of true dart-poison frogs (Phyllobates) appear to have occurred within the last five million years; two species in Costa Rica and western Panama were derived from a primitively striped ancestor that probably invaded Central America after uplift of the Isthmus of Panama about three million years ago. The species of Dendrobates (s. 1.) thus far studied are genetically much more variable than Phyllobates, which is consistent with accumulating evidence that Dendrobates is a polyphyletic assemblage. Several major lineages of dendrobatids seem to date back to the start of the Cenozoic, about 60 million years ago. NEOTROPICAL FROGS of the family Dendrobatidae have long attracted attention because of the brilliant coloration of many species and because of their unusual life-style, which includes diurnal habits, terrestrial eggs, and transport of tadpoles to water on the back of a nurse frog (reviewed by Myers and Daly 1983). It has been known since the early 19th century that Indians obtain a potent poison for blowgun darts from dendrobatid frogs in western Colombia. The few fragmentary investigations of this custom were among the reasons that led various writers to surmise that all brightly colored dendrobatids are toxic and that such frogs are used as a source of "arrow" poison by a diversity of Indian tribes outside of western Colombia. Myers et al. (1978) presented new ethnographic data and reviewed the relevant literature. They pointed out that three unusually toxic species of Phyllobates confined to western Colombia are the only frogs known with certainty to have been used for poisoning blowgun darts (not arrows), and that only the Embera Choco and Noanama Choco are positively known to have practiced this custom, which is now on the dedine. One still hears reports of frog poisons being used in other regions but these remain unconfirmed, and, in any case, the great majority of published statements are certainly fictional. The assumption that all brightly colored dendrobatids are "toxic" or at least noxious has proven to be essentially correct, although several relatively dull-colored species also secrete defensive