Effects of partial throughfall exclusion on the phenology of Coussarea racemosa (Rubiaceae) in an east-central Amazon rainforestAbstractIntroductionMethodsClimate and soilsTab1Treatment designThe study speciesPhenological observationsSystem water balance and PARStatistical analysesResultsStructures of Coussarea racemosa populationFig1Quantitative phenologyFig2Fig3Tab2Environmental correlates for leaf and flowering phenologyFig4Fig5DiscussionTiming of leaf flushing, flowering and fruitingFig6Fruit production and seed quality ConclusionAcknowledgmentsReferencesAbstract Severe droughts may alter the reproductivephenology of tropical tree species, but our under-standing of these effects has been hampered by con-founded variation in drought, light and other factorsduring natural drought events. We used a large-scaleexperimental reduction of throughfall in an eastern-central Amazon forest to study the phenological re-sponse to drought of an abundant subcanopy tree,Coussarea racemosa. We hypothesized that droughtwould alter the production and the timing of repro-duction, as well as the number of viable fruits. Thestudy system comprised two 1-ha plots in the TapajosNational Forest, Para, Brazil: a dry plot where 50% ofincoming precipitation (80% throughfall) was divertedfrom the soil during the six-month wet season begin-ning in January 2000, and a wet plot that receivednatural rainfall inputs. Fruit production of C. racemosawas quantified every 15 days using 100 litter traps(0.5 m2) in each plot. The production of new leavesand flowers was recorded monthly for C. racemosaindividuals. Soil water, pre-dawn leaf water potentialand solar radiation were measured to help interpretphenological patterns. Over the ~3.5-year period(April 2000 through December 2003), total fruit pro-duction remained similar between plots, declining by12%. In 2003, production was four times higher in bothplots than in previous years. In the dry plot, fruit fallshifted 40 and 60 days later into the dry season in 2002and 2003, respectively. Total fruit fall dry mass pro-duction was variable across the study period. Foliageand flower production coincided with peak irradianceearly in the dry season until delays in flowering ap-peared in the dry plot in 2002 and 2003. Plant waterstress, through its influence on leaf developmentalprocesses and, perhaps, inhibition of photosynthesis,appears to have altered both the timing of fruit fall andthe quality and number of seeds produced.Keywords Reproductive Æ Biology Æ Tropical ÆForest Æ DroughtIntroductionDroughts associated with Amazon deforestation (SilvaDias et al. 2002), biomass burning (Andreae et al.2004), higher temperatures and evapotranspirationassociated with global warming (Wang 2005), increas-ing El Nin˜o Southern Oscillation (ENSO) frequency(Timmermann et al. 1999), and other oceanic temper-ature disruptions (Rohter 2005), are predicted tochange forest productivity, carbon storage and vul-nerability to fire over extensive areas of the AmazonCommunicated by Jim Ehleringer.P. Brando (&) Æ D. Nepstad Æ G. CardinotInstituto de Pesquisa Ambiental da Amazoˆnia (IPAM),Av. Rui Barbosa, 136, Santare´m, PA, Brazile-mail: [email protected]. Ray Æ D. NepstadWoods Hole Research Center, 149 Woods Hole Road,Falmouth, MA 02543, USAL. M. CurranYale School of Forestry and Environmental Studies,370 Prospect St, New Haven, CT 06511, USAR. OliveiraLab. Ecologia Isoto´pica - CENA,Universidade de Sa˜o Paulo, Av. Centena´rio, 303,Piracicaba, SP 13.416-000, BrasilOecologia (2006) 150:181–189DOI 10.1007/s00442-006-0507-z123ECOPHYSIOLOGYEffects of partial throughfall exclusion on the phenologyof Coussarea racemosa (Rubiaceae) in an east-centralAmazon rainforestPaulo Brando Æ David Ray Æ Daniel Nepstad ÆGina Cardinot Æ Lisa M. Curran Æ Rafael OliveiraReceived: 13 February 2006 / Accepted: 5 July 2006 / Published online: 6 September 2006 Springer-Verlag 2006(Walther et al. 2002; Nepstad et al. 2002, 2004). Dryingmay also influence plant reproductive processes (Pen-uelas and Filella 2001; Ozanne et al. 2003), includingthe timing and amount of flower and fruit production,and thus affect future plant establishment (e.g., bychanging the optimal timing of pollination, seed dis-persal, seed germination and regeneration success;Walther et al. 2002). Such alterations in phenologicalpatterns can also influence plant–animal interactions,as they may disrupt the timing of fruit and seed avail-ability, seed dispersal patterns and seed predation(Curran and Leighton 2000; Curran and Webb 2000;Augspurger 1981, 1984).Among the many factors controlling the phenologyof tropical plants, subtle seasonal changes in day lengththat occur upon moving even short distances north orsouth of the equator have been proposed as the pri-mary trigger of flowering in aseasonal tropical forests(Rivera and Borchert 2000; Rivera et al. 2002; Borc-hert et al. 2005). However, in areas that experienceseasonal drought, plant water status is regarded as amore important mechanism influencing the variation inthe timing of flower initiation (Borchert 1991; Borchertet al. 1994; William 1997). For example, mast fruitingof trees in SE Asia and supra-annual fruit productionin the neotropics are associated with the droughts thatoccur during ENSO events (Curran et al. 1999; Wrightet al. 1999), although it is unclear whether the proximaltrigger for fruiting during ENSO is plant hydrationfollowing isolated rainfall events (Borchert et al. 2002),elevated radiation levels (Wright et al. 1999) or varia-tions in temperature preceding ENSO (Ashton et al.1988).Prolonged drought, however, may have a negativeeffective on plant reproduction because of the sub-stantial carbohydrate demands of both flowering andfruiting (Larcher 1995). Drought-induced reductions inleaf area and in quantum yield (Nepstad et al. 2002;Larcher 1995), or in stomatal conductance can inhibitphotosynthesis, constraining the amount of carbohy-drate available for investment in reproduction. Themagnitude of the changes that severe droughts mayimpose on the reproductive phenology of diversetropical forest tree species will depend upon theirability to cope with water stress through variousadaptations, including: (1) stomatal regulation of waterloss (Jones and Sutherland 1991), (2) internal adjust-ments in how carbohydrates are allocated (Chapinet al. 1990; Newell et al. 2002), (3) resistance todrought-induced cavitation (Cardiont, unpublisheddata), (4)