Variations in seed size and number affect key ecological processes of colonization and establishment of plant species within and across different habitats (Silvertown & Bullock, 2003; Guariguata & Ostertag, 2001; Moles & Westoby, 2006). In fact, smaller seeds germinate faster (Dolan, 1984; Souza & Fagundes, 2014), contributing with a greater competitive advantage to the plant, especially in early suc-cessional stages (Baskin & Baskin, 1998). Con-trarily, larger seeds, while germinating more slowly, often have higher rates of germination than smaller seeds, being favoured in predictable habitats (Harper, 1977; Geritz, 1995; Ferreira & Borghetti, 2004). In addition, broad variations in seed size allow the species to be able to colonize different habitats and expand their geographic distribution ranges (Leishman, 2001; Souza & Fagundes, 2014). Therefore, the seed size can affect the distribution and abundance of plants in natural communities (Jako-bsson & Eriksson, 2000; Leishman, 2001).

The relationship seed number versus seed biomass is a recognized trade-off capable of shaping seed size (Westoby, Leishman, Lord, Poorter, & Schoen, 1996; Leishman, 2001; Moles & Westoby, 2006). There is strong empirical evidence for a negative relationship between seed mass and seed crop, both per individual per year and per square meter of canopy, or per gram of plant biomass, per year (Leish-man, 2001; Moles & Westoby, 2006). However, several ecological (e.g. resource availability for the plant and interactions with pollinators and seed predators) and evolutionary (e.g. plant life form and phenological patterns) factors can change the result of the trade-off size/ number of seeds (Westoby, Leishman, Lord, Poorter, & Schoen, 1996). These ecological and evolutionary interactions generate large variations in size and number of seeds produced by plants among and within species (Leishman, Westoby, & Jurado, 1995; Silvertown & Bullock, 2003; Moles & Westoby, 2006).

Trade-off size/ number of seeds occupy a central core in plant ecology because it reflects ecological and evolutionary aspects of plant species (Sadras, 2007; Souza & Fagundes, 2014). Studies that describe trade-offs size/ number of seeds are relatively common (see Moles & Westoby, 2006 and references). However, studies of this nature that focus on supra-annual fruiting species are still rare, especially when comparing the trade-off size/number of seeds between years of low and high reproductive investment. Supra-annual reproduction in plants is characterized by the alternating fruit production cycles in plant species, with years of intense fruit production and years of little or no fruiting (Kelly, 1994; Koenig & Knops, 2005; Souza & Fagundes, 2014). These cycles have an enormous potential to affect the trade off size/number of seeds. Theoretically, it is expected that in years of intense reproductive activity the resource available for seed production would have to be distributed to a large number of seeds (Silvertown, 1980). Therefore, supra-annual plants should produce larger fruits and seeds during years of low reproductive activity (Souza & Fagundes, 2014).

However, in these supra-annual species, the expected trade-off size/number of seed can be masked because resource available for plant reproduction can change among cycles of reproductive events (Koenig & Knops, 2005). Thus, the objective of this study was to test the hypothesis of trade-off size/number of seeds among and within years of high and low reproductive investment in Copaifera langsdorffii, an arboreal tropical species with supra-annual fruiting. Hence, two predictions for tradeoff size/number hypothesis were tested: (i) on a populational scale, seeds produced during years of greater reproductive investment showed reduced size and (ii) on individual scale (within same event reproductive) there is an inverse relationship between size and number of seeds produced by individual plant.

Materials and Methods

Studied system: Copaifera langsdorffii (Fabaceae: Caesalpiniaceae) is a tropical, deciduous, heliophytic, selective xerophytic tree species with 7-30 m height. The species have a wide distribution from Northeast Argentina to Southern Bolivia and can be found throughout Brazil (Fagundes, 2014). The plants show complete deciduousness in the dry season (July to September) and new leaves begin to appear immediately after the drop of leaves produced in the previous growing season (Fagundes, Maia, Queiroz, Fernandes, & Costa, 2013). C. langsdorffii exhibits supra-annual mass fruiting (Pedroni, Sanchez, & Santos 2002). Flowering is concentrated from November to January and fruit ripening occurs from August to October, with producing flowers and fruits in the terminal portion of the branch (Fagundes et al., 2013). Upon opening, each fruit exposes a single ellipsoid seed, which is black and shiny and is partially covered by a yellow-orange aril (Carvalho, 2003).

Study area: Fieldwork was conducted in a Cerrado (Savanna) area located in the municipality of Montes Claros (16°40'26" S - 43°48'44'' W), Northern Minas Gerais State, Brazil. This region is located between the domains of the Cerrado and Caatinga (Rizzini, 1997), with a semi-arid climate characterized by well-defined dry and rainy seasons. The soil of the study area is dystrophic with a developed herbaceous sub-shrub layer, generally affected by fire (Costa, Siqueira, Oliveira, & Fagundes, 2011; Fagundes, Camargos, & Costa, 2011). The historical average annual temperature is approximately 23 °C, and the average precipitation is about 1 100 mm/year, with the rainy season concentrated between November and February (Costa et al., 2011). While annual average temperature has not changed significantly during the study years, the rainfall ranged from 1 360, 1 139, 1 058 and 1 073 mm/ year during 2008, 2009, 2010 and 2011, respectively (INMET, 2014).

Data collection: In December 2007, a total of 102 C. langsdorffii plants were selected in the study area and properly identified with metal platelets. All trees were with a fully developed crown and in a good phytosani-tary state (e.g. without lianas, parasitic plants or evidence of pathogens attack). The plants selected for the study were monitored monthly from January to September during four consecutive years (2008 to 2011) to characterize occurrence of flowers and fruits. Additionally, during the month of April of each reproductive year (2008 and 2011), ten terminal branches (30 cm long) were collected from each plant in different points of the tree canopies to minimize the possible effects of microclimate variations on fruit development (Costa et al., 2011). These branches were taken to the Conservation Biology Laboratory of the State University of Montes Claros in order to determine average number of fruits per plant. All the marked trees produced fruits in 2008, none produced fruits in 2009 or 2010, and only 47 individuals (46 %) were fruiting in 2011. Plants that produced seeds in 2008 and 2011 (n= 65) were categorized in two groups: g1= plant fruiting in 2008 (n= 33) and g2 = plant fruiting in 2011, and not already selected in g1 (n= 30). The exclusion of plants belonging to treatment g1 from treatment g2 was necessary to avoid temporal pseudoreplication.

During September of 2008 and 2011 (period of fruit maturation), 100 fruits were collected from each plant belonging to group g1 e g2 in order to determinate seed size. All fruits were manually processed, eliminating malformed seeds and seeds exhibiting symptoms of attacks by predators or pathogens. The volume of the seeds was used as indicative of seed size. The seed volume was determined using the formula for cone volume [V= (n.r².l)/3], where "r" is the radius and "l" is seed length (Souza & Fagundes 2014).

The data were analyzed using R software (R Core Team, 2014). Regression tests were performed by constructing generalized linear models (GLM) to test hypotheses predictions proposed in this study. Variations in seed number and size between reproductive events were tested using mean of seed number or volume per plant as response variables and reproductive years (2008 and 2011) as explanatory variable. The analysis of trade-off size/number of seeds within reproductive event was tested using the number of seeds/branch and years as explanatory variable and the average size of seed produced by each plants as response variables. All minimal adequate models were created with the removal of non-significant terms and compared with null model. All models were built using the appropriate error distribution considering the nature of each response variable, following by model criticism via residual analysis (Crawley, 2007).

Results

The population of C. langsdorffii exhibited reproductive activity only during 2008 and 2011. It was also found that seed production per plant varied between the two reproductive years (x² = 16.82, p < 0.001, Fig. 1A). In fact, the mean number of seeds per branch was approximately 26.4 % greater in 2008, when compared to 2011 (Fig. 1A). In addition, it is worth stating that most plants belonging to studied population (all marked plants) produced fruits in 2008, while only few individuals did in 2011. Seed volume resulted greater in 2008, when plants produced higher number of fruits (F= 6.32, p < 0.05, Fig. 1B).

For both reproductive years (2008 and 2011), there was a negative relationship between seed volume and number of seeds produced per branch of C. langsdorffii, and the interaction term with the reproductive year was significant. The interaction suggests that the trade-off size/number of seeds was indeed stronger in 2011, when plants producing lower seed number (Table 1, Fig 2).

Discussion

The result of this study indicates seed crop produced by Copaifera langsdorffii population was significantly greater in 2008 comparatively to 2011. The data is especially relevant if we consider that most plants were fruiting in 2008 while only a few individuals did so in 2011. Moreover, plants produced bigger seeds in 2008, suggesting greater resource availability for seed development in this year of higher reproductive investment. Our data therefore do not support the first prediction of the trade-off seed size/ number hypothesis in populational scale.

In this scenario it is important to stress that C. langsdorffii have supra-annual phenology with fruiting intervals of at least two to three years. Moreover, years of intense fruiting can be followed by years where fruit production is reduced to less than 10 % (Pedroni et al., 2002; Fagundes, 2014; Souza & Fagundes, 2014). The main evolutionary advantages of supra-annual fruiting for plants are to increase in pollination efficiency and seed survival, due to seed predator satiation in years of mass production (Silvertown, 1980; Kelly & Sork, 2002; Schauber et al., 2002; Kolb, Ehrlen, & Eriksson, 2007). In addition, some authors have proposed that resource availability is a key factor for supra-annual species to display their reproductive activity and intensity of fruit production. Limited resource availability is probably a factor that inhibits fruit production in years of low reproductive activity (Sork, 1993; Koenig & Knops, 2005). While temperature has not significantly varied between years in the studied area, total rainfall was 21.2 % higher in 2008 comparatively to 2011. Thereby, the lower rainfall resource available in 2011 may explain the lower number of seeds, as well as the smaller seed volume produced by C. Langsdorffi during the year 2011.

Limited resource availability generates a stronger trade-off in the allocation of energy among the different plant drains, so an optimal investment in growth, storage and reproduction will not be generally satisfied simultaneously by plants (Chapin, 1991). Trade-offs can be also observed within each of these metabolic routes (Leishman, 2001), as for example, the strategy of producing few large seeds versus numerous small seeds, that represents a key trade-off in resource allocation for many plant species (Leishman, 2001; Moles & Westoby, 2006). On an individual scale, the results of this study support the prediction of the tradeoff seed size/number hypothesis, as it was observed a negative relationship between seed size and number of seeds produced per plant in both two reproductive years of C. langs-dorffii. The trade-off in seed size/number of seeds has a solid theoretical basis (Smith & Fretwell, 1974; Lloyd, 1987) and several studies have demonstrated this pattern for several plant species (Werner & Platt, 1976; Shipley & Dion, 1992; Turnbull, Rees, & Crawley, 1999; Jakobsson & Eriksson, 2000; Henry & Westoby, 2001; Leishman, 2001). However, it is important to emphasize that this relation was stronger in 2011, when plants produced fewer fruits. Again, it should be clear that the resource availability may play a central role to justify this variation. Therefore, resource (rainfall) was more limited during 2011, accentuating the trade-off size/number of seed observed in this study.

Finally, the study calls the attention to supra-annual fruiting pattern in C. langsdoffii, and suggests that the phenological patterns contribute to explain the wide variation in their seed size and geographical distribution. Indeed, variations in seed size represent an important attribute of plant life history (Smith & Fretwell, 1974; Lloyd, 1987) and influences important ecological processes of colonization and establishment of species across habitats (Guariguata & Ostertag, 2001). Larger seeds produce more vigorous seedlings (Dan, Mello, Wetzel, Popinigis, & Zonta, 1987) allowing greater competitive ability in predictable habitats (Geritz, 1995). In contrast, smaller seeds germinate faster and colonize transient habitats easily (Aniszewski, 2001). Thus, supra annual fruiting could be favouring the colonization of different habitats by C. langsdorffii, facilitating its extended geographic distribution (see also Souza & Fagundes, 2014). This study suggests that the resource availability affects the size/ number of seeds trade-off in in C. langsdorffii. However, predicting the effect of ecological (e.g. resource available, ecological interaction) and evolutionary forces on supra-annual reproduction in plants will be essential to understanding this phenological pattern, a subject that remains unsolved.

Acknowledgments

We thank all the collaborators of the Laboratorio de Biologia da Conservacao da UNIMONTES for logistical support in the fieldwork. Authors are indebted to Dr. Philip Donkersley, for his valuable revision of the English version. This study was carried out with financial support from Fundacao de Amparo a Pesquisa do Estado de Minas Gerais (FAPEMIG APQ 01926.11). The authors also acknowledge the CAPES and FAPEMIG by research grants.

References

Aniszewski, T. (2001). Seed number, seed size and seed diversity in washington lupin (Lupinus polyphyllus Lindl.). Annals of Botany 8777-82.

Baskin, C. C. & Baskin, J. M. (1998). Seeds - ecology, biogeography, and evolution of dormancy and germination. New York: Academic Press.

Carvalho, P. E. R. (2003). Espécies arbóreas brasileiras. Brasilia, Colombo - Embrapa Florestas.

Chapin, F. S. (1991). Integrated responses of plants to stress. BioScience 4129-36.

Costa, F. V., Siqueira, F. N., Oliveira, J. S., & Fagundes, M. (2011). Relationship between plant development, tannin concentration and insects associated with Copaifera langsdorffii (Fabaceae). Arthropod-Plant Interactions 59-18.

Crawley, M. J. (2007). The R Book. New York: John Wiley and Sons.

Dan, E. L., Mello, V. D. C., Wetzel, C. T., Popinigis, F., & Zonta, E. P. (1987). Transferencia de matèria seca como mètodo de avaliacào do vigor de sementes de soja. Revista Brasileira de Sementes 945-55.

Dolan, R. W. (1984). The effect of seed size and maternal source on individual size in a population of Ludwigia leptocarpa (Onagraceae). American Journal of Botany 711302-1307.

Fagundes, M. (2014). Community of galling insect associated with Copaifera langsdorffii (Fabaceae): the role of inter and intra-annual host plant phenology. En G. W. Fernandes, & J. C. Santos (Eds.), Neotropical Insect Galls (pp. 163-174). New York: Springer.

Fagundes, M., Camargos, M. G., & Costa, F. V. (2011). A qualidade do solo afeta a germinacào das sementes e o desenvolvimento das plántulas de Dimorphandra mollis Benth. (Leguminosae: Mimosoidae). Acta Botánica Brasílica 25908-915.

Fagundes, M., Maia, M. L. B., Queiroz, A. C. M., Fernandes, G. W., & Costa, F. V. (2013). Seed predation of Copaifera langsdorffii Desf. (Fabaceae: Caesalpinioi-deae) by Rhinochenus brevicollis Chevrolat (Coleoptera: Curculionidae) in a Cerrado fragment. Ecología Austral 23218-221.

Ferreira, A. G., & Borghetti, F. (2004). Germinacào: do básico ao aplicado. Porto Alegre: Editora Artmed.

Geritz, S. A. (1995). Evolutionarily stable seed polymorphism and small-scale spatial variation in seedling density. The American Naturalist 146685-707.

Guariguata, M. R. & Ostertag, R. (2001). Neotropical secondary forest succession: changes in structural and functional characteristics. Forest Ecology and Management 148185-206.

Harper, J. L. (1977). Population biology of plants.. New York: Academic Press

Henry, M., & Westoby, M. (2001). Seed mass and seed nutrient content as predictors of seed output variation between species. Oikos 92479-90.

INMET - Instituto Nacional de Meteorologia. (2014). Retrieved from http://www.inmet.gov.br/portalaccess:10/12/2014 ;.

Jakobsson, A., & Eriksson, O. (2000). Seed size and frequency patterns of understory plants in Swedish deciduous forests. Ecoscience 974-78.

Kelly, D. (1994). The evolutionary ecology of mast seeding. Trends in Ecology and Evolution 9465-470.

Kelly, D. & Sork, V. L. (2002). Mast seeding in perennial plants: why, how, where? Annual Review of Ecology, Evolution, and Systematics 33427-447.

Koenig, W. D., & Knops, J. M. H. (2005). The mystery of masting in plants. American Scientist 93340-347.

Kolb, A., Ehrlen, J., & Eriksson, O. (2007). Ecological and evolutionary consequences of spatial and temporal variation in pre-dispersal seed predation. Perspectives in Plant Ecology, Evolution and Systematics 979-100.

Leishman, M. R. (2001). Does the seed size/number tradeoff model determine plant community structure: An assessment of the model mechanisms and their generality. Oikos 93294-302.

Leishman, M. R., Westoby, M., & Jurado, E. (1995). Correlates of seed size variation: a comparison among five temperate floras. Journal of Ecology 83517-529.

Lloyd, D. G. (1987). Selection of offspring size at independence and other size-versus-number strategies. The American Naturalist 129800-817.

Moles, A. T., & Westoby, M. (2006). Seed size and plant strategy across the whole life cycle. Oikos 11391-105.

Pedroni, F., Sanchez, M., & Santos, A. M. (2002). Fenologia da copaiba (Copaifera langsdorffii Desf. -Leguminosae, Caesalpinioideae) em uma floresta semidecidua no sudeste do Brasil. Revista Brasileira de Botanica 25183-194.

R Development Core Team. (2014). R: A language and environment for statistical computing. R Foundation for Statistical Computing, Vienna, Austria. Retrieved from http://www.rproject.org.

Rizzini, C. T. (1997). Tratado de fitogeografia do Brasil: aspectos ecológicos, sociológicos e florísticos. Rio de Janeiro: Ambito Cultural.

Sadras, V. O. (2007). Evolutionary aspects of the trade-off between seed size and number in crops. Field Crops Research 100125-138.

Schauber, E. M., Kelly, D., Turchin, P., Simon, C., Lee, W. G., Allen, R. B., Payton, I. J., Wilson, P. R., Cowan, P. E., & Brockie, R. E. (2002). Masting by eighteen New Zealand plant species: the role of temperature as a synchronizing cue. Ecology 831214-1225.

Shipley, B., & Dion, J. (1992). The allometry of seed production in herbaceous angiosperms. The American Naturalist 139467-483.

Silvertown, J. W. (1980). The evolutionary ecology of mast seeding in trees. Biological Journal of the Linnean Society 14235-250.

Silvertown, J. W., & Bullock, J. M. (2003). Do seedlings in gaps interact? A field test of assumptions in ESS seed size models. Oikos 101499-504.

Smith, C. C., & Fretwell, S. D. (1974). The optimal balance between size and number of offspring. The American Naturalist 108499-506.

Souza, M. L., & Fagundes, M. (2014). Seed germination of Copaifera langsdorffii (Fabaceae): the role of seed size. American Journal Plant Sciences 52566-2573.

Sork, V. L. (1993). Evolutionary ecology of mast-seeding in temperate and tropical oaks (Quercus spp.). Vegetatio 107133-147.

Turnbull, L. A., Rees, M., & Crawley, M. J. (1999). Seed mass and the competition/ colonization trade-off: a sowing experiment. Journal of Ecology 87899-912.

Werner, P. A., & Platt, W. J. (1976). Ecological relationships of co-occurring goldenrods (Solidago: Com-positae). The American Naturalist 110959-971.

Westoby, M., Leishman, M., Lord, J., Poorter, H., & Schoen, D. J. (1996). Comparative ecology of seed size and dispersal. The Royal Society 351 , 1309-1318.