Plant Species Biology (2008) 23, 9–17 doi: 10.1111/j.1442-1984.2008.00202.x Spatial and temporal variation in the fruiting phenology of palms in isolated stands GREGORY H. ADLER* and THOMAS D. LAMBERT† *Department of Biology and Microbiology, University of Wisconsin–Oshkosh, Oshkosh, Wisconsin 54901, United States of America and †Department of Natural Sciences, University of Virginia’s College at Wise, Wise, Virginia 24293, United States of America Abstract Fruiting phenologies of two species of palms, Astrocaryum standleyanum L. H. Bailey and Attalea butyracea (Mutis ex L. f) Wess. Boer, isolated on eight small (1.7–3.7 ha) forested islands in the Panama Canal were studied over a 33-month period. Individual palms were permanently marked with numbered aluminum tags and censused each month for the presence of ripe fruits. The dataset consisted of 1106 monthly observations of palms with ripe fruits among the 634 marked individuals. Mean densities of palms of reproductive size varied widely among islands, ranging from a low of 0.3 ha-1 for A. standleyanum and 3.5 ha-1 for A. butyracea to a high of 44.9 ha-1 for A. standleyanum and 33.7 ha-1 for A. bu- tyracea. Both species showed distinctly seasonal periods of fruiting activity that varied in duration between the two species and among years. The timing of fruiting by A. stand- leyanum was highly synchronous among islands, whereas inter-island synchrony in A. butyracea was less pronounced. The percentages of marked individuals that fruited varied widely among islands and years. Results indicated that these palms responded to both spatially and temporally variable conditions that promoted fruit production. We suggest that pollinator abundances are a crucial factor affecting reproductive output. Conditions that favor successful reproduction and seed dispersal, such as pollinator activity and the attraction of dispersal agents, may be the ultimate factors that have influenced the reproductive phenologies of these two species of palms. Keywords: Astrocaryum standleyanum, Attalea butyracea, fruit production, palms, Panama, phenology. Received 28 February 2007; accepted 30 August 2007 Introduction (e.g. Henderson et al. 2000). However, when recruitment is dependent on animal seed dispersers, competition for In tropical forests that experience a seasonal climate, such dispersers may select for variable levels of fruiting many plants show distinct seasonal patterns in their synchrony (Thies & Kalko 2004). Indeed, in Veracruz, vegetative and reproductive activities (e.g. Janzen 1967; Mexico, 11 species demonstrated highly synchronous Daubenmire 1972; Frankie et al. 1974; Foster 1982; Croat flowering, with most species flowering during the dry 1978). Seasonal reproduction depends on both proximate season; however, species bore fruit almost year round cues that trigger flowering and ultimate factors that influ- (Ibarra-Manriquez 1992). Henderson et al. (2000) studied ence the evolution of a plant’s reproductive phenology the flowering phenology of 27 species of palms in the (i.e. timing of flowering and fruit maturation; Foster central Amazon and found relationships between flower- 1982). Pollination by certain groups of insects, such as ing synchrony, regularity and duration and the types of weevils, may select for highly synchronous flowering that insects that pollinated the flowers. The importance of takes advantage of periods of high pollinator abundance insect pollinators to fruit set may be seen in Scariot et al. (1991, 1995), who noted that individuals of Acrocomia Correspondence: Gregory H. Adler aculeata (Jacq.) Lodd ex Mart. flowered from August to Email: [email protected] mid December, but that those individuals that flowered in © 2008 The Authors Journal compilation © 2008 The Society for the Study of Species Biology 10 G. H. ADLER AND T. D. LAMBERT October had the greatest fruit set. Similarly, DeSteven et al. output, such as pollinator abundance, are likely to vary. (1987) noted that, although Bactris major Jacq. flowers all In contrast to the study by DeSteven et al. (1987), which year round, only flowers appearing in the latter part of the was conducted in one 15-km2 contiguous forest, we rainy season produce ripe fruit. Thus, flowering perfor- studied spatially isolated stands of palms. Therefore, we mance of palms does not necessarily equate with overall addressed among-stand rather than within-stand pheno- fruit production and reproductive success (Adler et al. logical differences and we examined fruiting rather than 1998), and individual reproductive success may vary con- flowering phenologies because fruiting is generally more siderably within populations (DeSteven et al. 1987). synchronous than flowering in these species (DeSteven Seasonal pollinator abundance is likely to be ultimately et al. 1987), and fruit production is a better indicator of determined by broad-scale patterns, such as day length reproductive success (DeSteven et al. 1987; Scariot et al. and rainfall. However, small-scale microclimatic patterns, 1995; Adler et al. 1998). Our aims were to: (i) describe the such as exposure to rainy-season storms and dry-season distributions and densities of the two species of palms; trade winds may also influence local insect abundance (ii) examine the timing and duration of fruiting in each and may lead to differences in reproductive output year; (iii) determine the degree of fruiting synchrony among spatially isolated stands. Thus, although stands among the isolated palm stands; and (iv) compare repro- within a region may experience the same ultimate repro- ductive output among those stands. If fruiting is con- ductive cues, proximate cues that influence fruit set may strained solely by broad-scale seasonal cues, such as vary and may have implications for stand dynamics and rainfall and day length, then the timing and length of the the maintenance of frugivore populations in fragmented fruiting seasons should be highly synchronous among forests. even isolated palm stands. However, if small-scale micro- Many palms reproduce seasonally. Of the 13 species of climatic and habitat differences lead to differences in the palms studied by DeSteven et al. (1987) in central Panamá, pollinator communities among isolated stands, then nine species reproduced seasonally. Astrocaryum standley- reproductive output should vary among both islands and anum L. H. Bailey and Attalea butyracea (Mutis ex L. F) years. We conclude by discussing the observed spatial Wess. Boer [Scheelea zonensis L. H. Bailey] are common and temporal variation within the context of proximate and widely distributed neotropical palms that are particu- cues and ultimate factors that may have shaped the larly abundant in second-growth forests (Leigh 1999). observed phenologies. Both species show extraordinarily high fruit production per unit area of crown (Hladik & Hladik 1969) and Materials and methods provide crucial resources for frugivores (e.g. Glanz et al. Study area 1982; Adler 2000). Astrocaryum standleyanum is a monoe- cious subcanopy tree that is 7–15 m tall at reproductive Over 200 islands occur within Gatun Lake (9°10′N, height (DeSteven et al. 1987) and ranges from south- 79°50′W), which is an artificial impoundment that was eastern Costa Rica to the Pacific slope of Colombia and created during the construction of the Panama Canal. We northern Ecuador (Henderson et al. 1995). Attalea butyra- include in this study eight small islands ranging in size cea is a monoecious canopy and subcanopy tree that is from 1.7 to 3.5 ha that are covered with tropical moist 6–15 m tall at reproductive height, with a maximum forest of varying ages and floristic compositions (desig- height of 30 m or more (DeSteven et al. 1987). This species nated as islands 4, 5, 8, 9, 12, 14, 51 and 52; described in ranges from southern Mexico southward into western detail by Adler 2000). These islands are located within a Amazonia (Henderson et al. 1995). Flowering and fruiting 40-km2 area and experience the same major climatic pat- by A. standleyanum are highly synchronous within popu- terns, such as rainfall seasonality and the occurrence of lations on Barro Colorado Island (BCI) in central Panamá, rainy-season storm systems and dry-season trade winds. whereas within-population reproduction by A. butyracea However, microclimatic patterns, such as exposure to is only moderately synchronous (DeSteven et al. 1987). storm systems and trade winds, vary among the islands However, among-population reproductive synchrony because some islands are better protected from such and differences in reproductive success have not been effects by neighboring land masses. The climate of the examined for these two species, despite their abundance study area is highly seasonal with respect to precipitation; and importance for frugivores. an 8-month rainy season is punctuated by a short but Our study focuses on the reproductive phenologies of severe dry season that typically lasts from January both palm species in central Panamá. The 33-month study through to April. Over 90% of the annual precipitation was conducted on eight small islands in Gatun Lake, falls during the rainy-season months, and dry-season where BCI is located centrally. Seasonal cues, such as day trade winds prevailing from the north-west have a pro- length and rainfall, were similar across the islands, but nounced drying effect. The mean annual rainfall on BCI is other factors that may have influenced reproductive 2612 mm (Windsor 1990). © 2008 The Authors Plant Species Biology 23, 9–17 Journal compilation