Patterns of Diametric Growth in Stem-Analyzed Laurel Trees (Cordia alliodora) in a Panamanian Forest

Based on cross-dated increment cores, yearly diameters of trees were reconstructed for 21 laurels (Cordia alliodora) growing in a natural secondary forest on Gigante Peninsula, Panama. From this sample of dominant-codominant trees, ages were 14–35 years with an average of 25 years. Growth typically slowed at 7 years old, indicating effects of competition from closure of gaps in the canopy. Rate of growth in diameter was modeled using the Bertalanffy-Richards growth function. Four patterns of growth were detected, of which 57% were sigmoidal, 19% were convex, 10% were monomolecular, and 14% were inverse sigmoid. RESUMEN—Con base de testigos fechados de madera, diámetros anuales se reconstructaron para 21 árboles de laurel (Cordia alliodora) creciendo en un bosque natural secundario en la Peninsula Gigante, Panama. De esta muestra de arboles dominantes-codominantes, los edades varı́an de 14 a 35 años con un promedio de 25 años. Tı́picamente el crecimiento de los árboles disminuó a siete años de edad, señalando los efectos de competencia debido de la cierre de los claros en el dosel. La tasa de crecimiento diamétrico de los árboles se modeló usando la function de crecimiento Bertalanffy-Richards. Cuatro patrones de crecimiento se encontraron; 57 por ciento de los patrones eran sigmoideos, 19 por ciento eran convexos, 10 por ciento eran monomoleculares, y 14 por ciento eran sigmoideos inversos. Growth of an individual tree is subject to complex interacting influences. In the early 20th century, there was a great deal of debate in ecology about aggregation versus nominalism (Mitman, 1992). Aggregation refers to grouping entities together on some basis and considering the assemblage as a whole. Nominalism rejects universals; it holds that the various objects labeled by the same term have nothing in common but their name. The aggregationists argued that nominalism is logically inconsistent: were there nothing in common among individuals, they could not be considered as members of the same species and combined under one name. Individuals share many common features and often exhibit aggregate behavior. It makes sense to use the average to characterize a population. Nominalists believe that models that deal with a population as a whole (aggregated large-scale models) are based on unrealistic assumptions, gloss-over ecological mechanisms and individual variability, and ignore reality. Aggregation violates the biological principle that each individual is different. Under the banner of nominalism, the future of ecology belongs to the individual-based modeling approach. Today, it is understood that the modeling paradigm depends on the objective for the model; both approaches are valid (Kaufmann and Ryan, 1986). In our study, we examined patterns of growth of individual laurels (Cordia alliodora), an important tropical tree, to better understand dynamics of the population. Laurels are tropical hardwoods with an extensive natural range extending from Mexico to Argentina, 258N to 258S. This tree is a pioneer species, invading clearings such as pastures, burned sites, and new gaps in forests. Laurels have a narrow, self-pruning crown allowing a diversity of plants to grow beneath it (Hummel, 2000b). As an important species for timber, the laurel frequently is used in agroforestry as shade for cacao, coffee, other agricultural crops, and farm animals. The commonly used commercial name for C. alliodora is salmwood. Laurels reach sexual maturity in 5–10 years (Liegel and Stead, 1990). Under optimum conditions, laurels can grow to 25–45 m in height and 60–100 cm in diameter. Typically, however, they grow to ca. 20 m in height and 46 cm in diameter (Somarriba and Beer, 1987; Liegel and Whitmore, 1991). In seasonally dry lowland tropical forests, most species of trees lose their leaves during the dry season, but in Panama, laurels become deciduous for several months after the rains begin (Foster, 1985). Laurels are evergreen as a seedling, semideciduous (in the dry season) as a sapling, and deciduous (in the wet season) as a mature tree (Menalled et al., 1998). To fully exploit the forestry and agroforestry potentials of this species, a better understanding of dynamics of its growth is needed. Objectives of our analysis were to discover patterns of growth of laurels on Gigante Peninsula in Panama, draw inferences about growth of laurels in this secondary natural forest, and make comparisons with growth of laurels at other sites. MATERIALS AND METHODS—We selected trees from a secondary natural forest on Gigante Peninsula, Panama (9810N, 79851W). The terrain had many hills with sedimentary rock containing volcanic debris. The soil was rich in clay, making permeability poor, particularly below the fine-root zone, which included the top 10–20 cm of the generally 50-cm depth of soil (Dietrich et al., 1996). The site was directly across Gatun Lake from Barro Colorado Island. Gigante Peninsula, as well as Barro Colorado Island, was formed during 1910–1914 when the Chagres River was dammed. Today, Gigante Peninsula is part of the Barro Colorado Island Nature Monument, which was established in 1977. The area is protected from disturbances such as shifting agriculture and selective cutting. The last major disturbance occurred in 1937 with banana blight. Gatun Lake is part of the Panama Canal and the area is considered an intermediate moistclimate zone, one of three possible climatic regions in Panama. The three forested regions in Panama are dry, moist, and wet (Foster, 1985). Average annual temperature in a clearing at Barro Colorado Island was 278C with a diurnal range of 98C. Potential evapotranspiration was ca. 12 cm/month (Dietrich et al., 1996). The rainy season was April–November, with November being the wettest month averaging 450 mm of rain (Foster, 1985). Of the 2,600 mm of annual precipitation, 90% was during the rainy season (Dietrich et al., 1996). Most rainfall occurred during short-term storms in early afternoon. During the dry season, rain is light and sporadic (Rand and Rand, 1996) because convective storms are turned away by trade winds (Dietrich et al., 1996). During August 1991, we sampled 21 laurel trees growing on Gigante Peninsula. Using an increment borer, we collected 2 cores/tree. Only dominant or codominant trees were chosen. Cores were collected at about breast height (1.3 m). Diameter (inside bark) at each year for each tree was determined by cumulatively summing widths of rings of the 2 cores/tree. The current year was not measured because that growth was incomplete. If a core missed the pith, we estimated number of years and distance to pith. These data were collected originally for use in a dendroecological study. However, this type of stemanalyzed data permits analysis of rate of growth in diameter and total growth in diameter. For a description of how cores were prepared and cross-dated, see Devall et al. (1996). There were 457 diameter-by-age data points available for fitting the Bertalanffy-Richards function. The Bertalanffy-Richards growth model is a mathematical expression of the essential causes of the growth phenomenon, with parameters that possess a gross physiological or biological interpretation. Pienaar and Turnbull (1973) posed this model and its corresponding principles as a general theory of growth of broad and general validity. While some researchers question legitimacy of the theoretical foundations (e.g., Ricker, 1979), this model has enjoyed widespread use. This equation is valued for its accuracy and is used more than any other function in studies of growth of trees and stands (Zeide, 1993). Schnute (1981), working with populations of fish, has contributed greatly to our understanding of models of growth and Lei et al. (1997) have shown equivalence between Bertalanffy-Richards and