Skip to main content
NCBI home page
Annals of Botany logoLink to Annals of Botany
. 2017 May 26;120(1):147–158. doi: 10.1093/aob/mcx051

Life history traits influence the strength of distance- and density-dependence at different life stages of two Amazonian palms

Juanita Choo 1,*, Cecilia Carasco 2, Patricia Alvarez-Loayza 3, Beryl B Simpson 4, Evan P Economo 1
PMCID: PMC5737847  PMID: 28549080

Abstract

Background and Aims Natural enemies are known to be important in regulating plant populations and contributing to species coexistence (Janzen–Connell effects). The strength of Janzen–Connell effects (both distance- and density-effects) varies across species, but the life history traits that may mediate such a variation are not well understood. This study examined Janzen–Connell effects across the life stages (seed through adult stages) of two sympatric palm species with distinct phenologies and shade tolerances, two traits that may mediate the strength and timing of Janzen–Connell effects.

Methods Populations of two common palm species, Attalea phalerata and Astrocaryum murumuru, were studied in Manu National Park, Peru. Seed predation experiments were conducted to assess Janzen–Connell effects at the seed stage. In the post-seed stages, spatial point pattern analyses of the distributions of individuals and biomass were used to infer the strength of distance- and density-effects.

Key Results Seed predation was both negative distance- and density-dependent consistent with the Janzen–Connell effects. However, only seedling recruitment for asynchronously fruiting Attalea phalerata was depressed near adults while recruitment remained high for synchronously fruiting Astrocaryum murumuru, consistent with weak distance-effects. Negative density-effects were strong in the early stages for shade-intolerant Attalea phalerata but weak or absent in shade-tolerant Astrocaryum murumuru.

Conclusions Distance- and density-effects varied among the life stages of the two palm species in a manner that corresponded to their contrasting phenology and shade tolerance. Generalizing such connections across many species would provide a route to understanding how trait-mediated Janzen–Connell effects scale up to whole communities of species.

Keywords: Astrocaryum murumuru, Attalea phalerata, bruchid beetles, density-correlation function, density-effects, distance-effects, Janzen–Connell, pair-correlation function, phenology, seed predation, shade tolerance, spatial patterns

INTRODUCTION

Understanding the factors maintaining high levels of species diversity in tropical forests is a long-standing – and still central – goal of ecology (Wright, 2002). Over the years, numerous mechanisms have been proposed to explain species coexistence in hyperdiverse tropical forest communities (Coley and Barone, 1996; Hubbell, 2001; Wright, 2002). Janzen (1970) and Connell (1971) famously highlighted the role of natural enemies in promoting plant species coexistence (hereafter JC effects). They hypothesized that species-specific natural enemies regulate plant populations by attacking early-stage plants (seed, seedling and saplings) that are found near conspecific adults (distance-effects) or are found in high densities with other conspecifics (density-effects). The resulting distance- and/or density-dependent mortality leads to depressed recruitment near conspecific adults or in areas of high conspecific density, respectively, and facilitates the recruitment of other less abundant species (Janzen, 1970). At the community level, JC effects can promote stable coexistence and enhance community-level diversity if the natural enemies of a species suppress the density of conspecifics more than heterospecifics (Chesson, 2000; Gilbert, 2005).

The JC hypothesis has been extensively and continuously tested in field studies since it was first proposed (Comita et al., 2014). However, despite this significant body of research, there remain major gaps in our knowledge. Notably, while JC effects are commonly observed, their presence and strength vary considerably across species, and the life history traits that mediate that variation are not well understood (Comita et al., 2014). Furthermore, trait-mediated JC effects may manifest at different life stages as distance- and/or density-effects. Most studies have focused on a single life stage, and although distance- and density-effects are distinct processes, only a few studies have evaluated their importance simultaneously (Carson et al., 2008; Comita et al., 2014). Here we address the nexus of plant life history traits and the strength of distance- and density-effects at different life stages using a case study of two Amazonian palm species.

There are many plant traits that could potentially mediate JC effects, but here we focus on two key life history traits: fruiting phenology and shade tolerance. Fruiting phenology was proposed as an important mediator of JC effects in the seed stage because it influences the spatial and temporal availability of resources to seed predators (Hammond and Brown, 1998). Fruiting phenology in turn has important implications for the ability of plants to avoid predators. Episodic and synchronous production of large seed crops in plant populations, for instance, can promote the survival of plant propagules through predator satiation (Janzen, 1971). In addition to its impact on seed predation, fruiting phenology may indirectly influence the strength of JC effects in the seedling stages because seed predation affects the density and pattern of seedling recruitment (Schupp, 1990; Silman et al., 2003; Paine and Beck, 2007; Ferreira et al., 2011). Theoretically, weak density-effects due to predator satiation would lead to high seedling recruitment (Nathan and Casagrandi, 2004), although such high recruitment could consequently lead to strong density-effects in the later stages as conspecifics compete for resources (Comita et al., 2014). The differences among species fruiting phenologies could therefore explain reported variations in the strength of JC effects across life-stages and among species. Thus, investigations across life-stages are necessary to shed light on the impacts of phenology on the strength of JC effects over ontogeny.

In the post-seed stages, some studies suggest shade tolerance as an important life history trait influencing the strength of JC effects (Comita and Hubbell, 2009). Density-dependent mortality for tree seedlings (Kobe and Vriesendorp, 2011) and saplings (Velázquez et al., 2015) was found to be stronger for shade-intolerant than for shade-tolerant species. Such differences in the strength of JC effects between shade-intolerant and shade-tolerant species could be linked to the growth–survival trade-off hypothesis where fast growing and shade-intolerant species invest more in growth than defence and slow-growing shade-tolerant species invest more in defence than growth (Coley et al., 1985; Wright, 2002). However, shade tolerance and resistance against natural enemies can change over a plant’s lifetime (Herms and Mattson, 1992; Boege and Marquis, 2005). For example, young leaves experience greater herbivore damage compared to mature leaves in tropical woody tree species (Coley, 1980). Thus, assessments of JC effects encompassing multiple or all life stages (e.g. Zhu et al., 2015) are necessary to compare the strengths of JC effects when plant species have distinct shade tolerances.

In addition to linking how plant traits influence variation in the strength of JC effects among species, further work is also necessary to address the relative or joint importance of distance- vs. density-effects. Concurrent investigation of both distance- and density-effects is necessary to first untangle the potential confounding effects of distance and density due to the higher densities of offspring near parent trees (Gilbert et al., 1994; Terborgh, 2012). For instance, in a recent study, Zhu et al. (2015) compared density-effects for multiple plant species, but because distance-effects were not tested, it remains unclear whether distance to adults could have also contributed to the observed negative impacts of conspecific neighborhood density.

Many tropical tree species, however, are slow-growing and long-lived, posing a challenge to studying JC effects over their lifetimes. One way to overcome this problem is to infer JC effects from the spatial patterns and associations of extant palm populations, i.e. using space as a surrogate of ecological processes (McIntire and Fajardo, 2009). Janzen–Connell effects are particularly amenable to spatial analyses. For instance, distance-effects and decreasing positive associations between adults and recruits through ontogeny can be assessed using pair-correlation functions (e.g. Jacquemyn et al., 2010). Density-effects can be inferred from the changes in the spatial associations among conspecifics of increasing size classes using univariate pair-correlation functions (e.g. Johnson et al., 2012; Piao et al., 2013). Distance- and density-effects can also have a negative impact on plant growth or biomass (Peters, 2003), which can be detected using mark-correlation and density-correlation functions, respectively (e.g. Martínez et al., 2013; Fedriani et al., 2015).

In this study, we investigated how life history traits influence the strength of distance- and density-effects across the ontogeny of two Neotropical palm species. Although palms are abundant in humid tropical forests (Gentry, 1986; Kahn de Granville, 1992), JC effects have been understudied for this non-woody plant group compared to woody tree species (Carson et al., 2008). Attalea phalerata and Astrocaryum murumuru experience high seed predation by bruchid beetles (Terborgh et al., 1993) but exhibit contrasting phenologies and shade tolerances. We used a combination of seed predation experiments and spatial analytical approaches to address the following questions in our two palms. (1) Is there evidence that distance from conspecific reproductive palms (distance-effects) and conspecific seed density (density-effects) influence predation probability and intensity by bruchid beetles? If so, are seed predation impacts on seedling recruitment consistent with differences in the phenology of the two palm species? Although both distance- and density-dependent predation occur on palm seeds, we predict that seed predators would have a stronger impact on seedling recruitment for asynchronously fruiting Attalea phalerata compared to synchronously fruiting Astrocaryum murumuru because synchronous fruiting should lead to higher resource availability, which can lead to predator satiation. (2) Is there evidence of distance- and density-effects in the post-seed to adult stages in both palm species based on conspecific spatial associations and biomass? If so, do both palm species exhibit similar strength of JC effects given their different shade tolerance? We predicted stronger distance- and density-effects for the shade-intolerant Attalea phalerata than for the shade-tolerant Astrocaryum murumuru based on the findings of Kobe and Vriesendorp (2011) and Velázquez et al. (2015).

MATERIALS AND METHODS

Study species and site

This study was conducted in a 2·25-ha plot (150 × 150 m) at the Cocha Cashu Biology Field Station in Manu National Park, Peru. Our focal species, Attalea phalerata Mart. ex Spreng. and Astrocaryum murumuru Mart., belong to the palm family Arecaceae, the most abundant arborescent plant group in western Amazonia (Terborgh and Andresen, 1998). Both palms are among the five most common species at our study site (Gentry and Terborgh, 1990). Within a population, Attalea phalerata palms fruit asynchronously throughout the year, whereas Astrocaryum murumuru palms have a relatively synchronous fruiting season with all individuals fruiting between March and May (Cintra, 1997). The two species share a community of vertebrate and invertebrate seed predators and dispersers including bruchid beetles, rodents, capuchins and peccaries (Cintra, 1997; Choo et al., 2012). However, the predatory activities of bruchid beetle larvae belonging to three species – Pachymerus cardo, Caryoborus serripes and Speciomerus giganteus – are known to cause significant mortality (over 90 %) in the seeds of palms found under conspecific reproductive adults (Terborgh et al., 1993).

Seed predation patterns

To assess if seed predation probability and intensity was affected by seed density and distance from conspecific reproductive adults (Question 1), we used a plot-wide baiting experiment to assess bruchid activities at the palm population level. Since previous work has examined the predation of bruchids on Astrocaryum murumuru seeds at this study site (Terborgh et al., 1993; Cintra, 1997), we focused on the predation of Attalea phalerata seeds and compared our results with those from Astrocaryum murumuru. We established a total of 36 baiting stations spaced approximately 25 × 25 m apart in a 6 × 6 grid within our study plot. Seeds used for baiting bruchids were collected from two fruiting Attalea phalerata palms that were located outside the study plot. Capuchin monkeys had previously eaten the fruits from those palms and discarded the exposed seeds to the ground. We collected the seeds immediately after the monkeys left and checked that no bruchid eggs were present on all seed surfaces.

To examine how bruchids responded to different seed densities, we used seed clusters with six vs. two seeds to represent ‘high’ vs. ‘low’ seed densities, respectively. Due to the limited number of fresh seeds available, we were unable to include more seeds in the larger seed cluster. Bruchids are, however, able to discern minute differences in host volatile concentrations (Messina, 2002) and previous investigations have shown their ability to distinguish between seed clusters of four vs. eight seeds (Cintra, 1997). We hence assumed that the three-fold difference in seed number was probably sufficient to detect any differences in bruchid responses to ‘high’ vs. ‘low’ seed densities. At each station, we situated the ‘high’ and ‘low’ seed density clusters approx. 5 m apart. We revisited each seed station daily for 21 d during which palm seeds remain attractive to bruchids (Wilson and Janzen, 1972). Each day, we checked the seeds for the presence of bruchid eggs, which are visually distinct. Bruchid larvae generally emerged from the eggs after a few days, leaving a visible hole in the seed as a result of the larvae drilling into the seed endosperm. We used the presence of these holes and their numbers on each seed as evidence of predation probability and predation intensity, respectively.

We examined the effect of distance from Attalea phalerata adults (distance-effects) and density of conspecific seeds (density-effects) on the two dependent variables – predation probability (i.e. presence or absence of holes in seeds) and predation intensity (i.e. total number of holes on each seed). We assumed that predation probability followed a binomial distribution and predation intensity followed a Poisson distribution. Since the two response variables showed many more zero counts than the assumed distributions, fitting the model with an ordinary generalized linear model would lead to severe bias (Zuur et al., 2009). We thus we used the zero-inflated hurdle (ZIH) model instead.

The ZIH model is a two-part conditional model, which allows the zero and non-zero values to come from separate data-generating processes. The first part is a binomial model in which the data are divided into zero and non-zero components; the second part is the positive range of the data set. The ZIH allows the possibility that different mechanisms may be involved in determining the zero vs. non-zero components (Zuur et al., 2009). In our study, we modelled the first part of the model i.e. predation probability (presence or absence of holes in seeds) using a generalized linear mixed-effects model (GLMM) with binomial distribution and logit link (logistic regression) and the second part of the model, i.e. predation intensity (total number of holes on each seed), with GLMM with zero-truncated Poisson and log-link function (predation count data were not over-dispersed). In these two models, seed density was treated as a factor with two levels (large vs. small) and distance from reproductive palm adults was treated as a continuous variable. We included seed station as a random effect to account for correlation in the number of bruchid-made holes from the same station. Collinearity was not detected among variables. We fitted the ZIH model with the function glmmadmb in the glmmADMD package (Skaug et al., 2014) in the R statistical computing environment (R Core Team, 2017).

Palm census

To assess if distance- and density-effects in the post-seed to adult stages for Attalea phalerata and Astrocaryum murumuru affected palm spatial associations and biomass (Question 2), we first mapped and measured all individuals of the two species in the plot. We divided our populations into four distinct ontogenetic stages generally observed in palms (Tomlinson, 1990): (i) seedlings possess entire leaves and are stemless, (ii) juveniles have slitted-leaves and are stemless, (iii) non-reproductive adults have slitted-leaves and stems, and (iv) reproductive adults possess stems and produce fruits and flowers. We confirmed the reproductive status of each adult palm using evidence of flowers or fruits in the crown or remnants of flower or fruiting parts under its canopy, which are woody and persist on the forest floor for years. We mapped a total of 598 Attalea phalerata and 1598 Astrocaryum murumuru individuals in the study plot (Supplementary Data Fig. S1). For the Attalea phalerata population, we identified 230 seedlings, 300 juveniles, 35 non-reproductive adults and 30 reproductive adults, and for Astrocaryum murumuru, we identified 582 seedlings, 962 juveniles, 26 non-reproductive adults and 33 reproductive adults.

We measured the heights of seedlings and juveniles using the longest leaf length (from the petiole base to leaf apex). The heights of adult palms (from stem base to crown apex) were estimated using a clinometer (Suunto Tandem compass-clinometer, Vantaa, Finland). We estimated the above-ground biomass for non-adult palms using an allometric model based on leaf length (since palms in the pre-adult stages lack stems) and for adult palms, we applied an allometric model based on stem height (Goodman et al., 2013). Note that we did not use diameter-at-breast-height (dbh)-based models commonly used for woody tree species to estimate palm biomass. The dbh-based models are poor estimators of above-ground palm biomass because palms, as non-woody plants, do not possess stems until the adult stages, and stems grow in height but lack secondary or diameter growth (Goodman et al., 2013).

Spatial analyses

Distance- and density-effects inferred from palm patterns

We used two spatial summary statistics – the pair-correlation function and the nearest-neighbour distribution function – to detect spatial evidence of JC effects, namely the decrease in aggregation among conspecific palms (density-effects) and decrease in positive associations between reproductive and conspecific non-reproductive palms (distance-effects) through ontogeny (i.e. from seedlings to juveniles and non-reproductive adults).

The pair-correlation function (hereafter PCF), gij(r), is related to the derivative of the K-function, i.e. gij(r) = K′ij(r)/(2πr) (Illian et al., 2008) and represents the average density (density normalized) of palms of life stage j within an annulus r centred on the focal palms of life stage i, divided by the overall density λj of palms of life stage j (Wiegand and Moloney, 2004). We used the univariate function g11(r) to quantify changes in the spatial associations among conspecific palms over ontogeny to infer density-effects. To infer distance-effects, we used the bivariate function g12(r) to quantify changes in the spatial associations between non-reproductive stages (seedlings, juveniles, non-reproductive adults) and conspecific reproductive adults. Following Wiegand et al. (2007a), we compared observed values of g11(r) or g12(r) against null models based on heterogeneous Poisson processes with non-parametric intensity estimates using an Epanečnikov kernel. We used a bandwidth of 30 m to account for possible large-scale heterogeneity in the environment and as the maximal scale at which second-order effects are expected in tropical forests (Wiegand et al., 2007a; Getzin et al., 2014). The null models for univariate patterns were generated by randomizing the location of all palms (pattern 1) under a heterogeneous Poisson process. The null model for bivariate patterns was generated by keeping the locations of reproductive adult palms (pattern 1) fixed while randomizing the locations of conspecific non-reproductive palms (seedlings, juveniles or non-reproductive adults, pattern 2) using the heterogeneous Poisson process. We constructed Monte Carlo simulation envelopes using the 25th highest and lowest values of 999 simulations of the null model following Wiegand et al. (2007b). We tested the significance of our data against the null model for distance intervals r up to 10 m, where negative neighbourhood interactions are most likely to occur (Packer and Clay, 2000). Because testing several scales of r simultaneously can inflate type I errors, we additionally applied the goodness-of-fit (GOF) test to confirm the significance of these patterns (Loosmore and Ford, 2006).

When GOF returned a significant P-value, indicating departure from the null model, we examined the summary statistics of the PCF to determine the specific distances where these departures occurred (e.g. Fedriani et al., 2015). For univariate PCF, values of g11(r) falling above, within or below the simulation envelopes would indicate that conspecific palms (pattern 1) were aggregated, randomly distributed or regularly distributed, respectively. For the bivariate PCF, values of g12(r) falling above, within or below the simulation envelopes would indicate that seedlings, juveniles or non-reproductive adult palms (pattern 2) were positively associated, randomly distributed or negatively associated with conspecific reproductive adult palms (pattern 1), respectively. All analyses were conducted using the software Programita (Wiegand and Moloney, 2004, 2013). To facilitate statistical comparisons among the different palm groups (e.g. seedling vs. juvenile associations), we also compared the mean strength of associations based on values of g11(r) and g12(r) within the range 1–10 m using Tukey’s non-parametric multiple-comparison (R package ‘nparcomp’).

In addition to the PCF analyses, we analysed the nearest neighbour distribution function (NDF) (Diggle, 2003) among conspecifics (D11) and between reproductive and conspecific non-reproductive adults (seedling, juveniles and non-reproductive adults) (D12 and D21). Density and distance-effects are expected to increase the nearest neighbour distances among conspecifics and increase the distance between recruits and their nearest conspecific adults, respectively, through time (Jansen et al., 2008; Terborgh et al., 2008; Comita et al., 2014). We therefore compared the univariate NDF D11 among conspecifics across life stages to assess if there was a shift in the proportion of palms from their nearest conspecific neighbour within distance r over ontogeny (density-effects). We then compared the bivariate NDF D21 to assess if there was a shift over ontogeny in the proportion of non-reproductive palms (pattern 2) with the nearest conspecific reproductive adult (pattern 1) within distance r (following Jansen et al., 2008). In addition to D21, the D12 can additionally describe subtle details of local clustering (Wiegand and Moloney, 2013). For example, some reproductive adults (pattern 1) may be surrounded by fewer or more conspecific non-reproductive palms (pattern 2) within larger neighbourhoods. We compared D11 and D12 to a null model simulated under a heterogeneous Poisson process as described in the PCF analyses. For the bivariate D21, the null model was simulated by holding the location of reproductive palms (pattern 2) fixed while randomizing conspecific non-reproductive adult palms (pattern 1) under a heterogeneous Poisson process. For both univariate and bivariate NDF analyses, we used GOF to test the significance of these patterns as described in the PCF analyses.

Distance- and density-effects inferred from palm biomass

The spatial distribution of palm biomass may provide additional evidence for JC effects. Specifically, focal palms that are spatially close to conspecific reproductive adults (distance-effects) or close to other conspecifics in the same stage (density-effects) may have a lower biomass. We thus applied the r-mark correlation functions and the density-correlation function to detect such distance- and/or density-effects, respectively, on palm biomass.

The r-mark correlation function (Illian et al., 2008) km.(r) uses the test function t(mi, mj) = mi., where mi and mj represent the marks associated with points i and j (all subscripts of the r-mark correlation function follow those of Illian et al., 2008). For our study, the r-mark correlation calculates the mean biomass of non-reproductive palms (e.g. seedlings) mi that are r distance away from a conspecific reproductive adult palm j. This function is then normalized by its non-spatial mean μ taken over all pairs of non-reproductive palms. We compared the r-mark correlation function to a null model of independent marking, which shuffled the marks (i.e. palm biomass) randomly over non-reproductive palms i while keeping their locations fixed (Illian et al., 2008; Getzin et al., 2011). Values of km.(r) greater or lower than 1 indicate that non-reproductive palms at distance r away from a conspecific reproductive adult palm are larger or smaller than average, respectively.

The density-correlation function CmK(r) was originally developed by Fedriani et al. (2015) to assess the effect of conspecific density on plant reproduction. Here we adapted the density-correlation function to assess the effect of conspecific density on plant biomass. We used CmK(r) to estimate the Pearson correlation coefficient between the biomass mi of a palm i and the density of conspecific neighbours within distance r using the following test function t (r, mi, Ki) = [mi – μ][λΚi(r) – λΚ(r)], where mi is the biomass of a focal palm i, μ is the mean biomass of the palm population, λ is the total density of conspecific palms in the study area, λΚi(r) is the number of neighbours found within a conspecific focal palm i at distance r and λΚ(r) is the mean number of conspecific neighbours found within distance r for all palms. To normalize the density-correlation function, CmK(r) is divided by the product of the standard deviations of palm biomass mi and the individual K-functions, respectively, i.e. σmσK. We compared the density-correlation functions for conspecific palms using repeated simulations of a null model of independent marking. This involved randomizing the biomass mi over all conspecific palms.

For both r-mark and density-correlation functions, we performed 999 Monte Carlo simulations and used the 25th highest and lowest values of the test statistic as simulation envelopes (error rate of 0·05 for a single distance). We then applied the GOF tests described previously to determine departures from the null model. All analyses were carried out on the Programita software (Wiegand and Moloney, 2004, 2013).

RESULTS

The seed predation study for Attalea phalerata showed that seed density had a positive effect on predation probability and intensity, whereas distance of seed stations from reproductive Attalea phalerata adults had a negative effect on predation probability and intensity (Table 1).

Table 1.

Results of a two-part GLMM for palm seed predation; significant effects are in bold type

Effect Estimate s.e. z P
Probability of predation
Intercept 0·148 0·415 0·357 0·720
Seed density (low or high) 0·176 0·070 2·529 0·011
Distance from Attalea phalerata −0·043 0·014 −2·969 0·003
Intensity of predation
Intercept 0·169 0·300 0·565 0·572
Seed density (low or high) 0·133 0. 050 2·632 0·008
Distance from Attalea phalerata −0·024 0·009 −2·608 0·009
Open in a new tab

Density-effects

When we examined the spatial associations among conspecifics across life stages (seedlings, juveniles and adults), the univariate PCF and NDF for both palm species showed a decrease in aggregation among conspecifics from the seedling to adult stages. The univariate PCF for Attalea phalerata showed that for seedlings and juveniles, conspecifics were aggregated within 0–10 m (Fig. 1A, B), but the average strength of association among seedlings was stronger than those among juveniles (P < 0·01, Tukey’s non-parametric multiple comparisons). The spatial associations among conspecific adults (non-reproductive and reproductive) did not depart from the null model (Fig. 1C). The univariate NDF indicated a shift in the proportion of palms away from the nearest conspecifics from the seedling to adult stages (Fig. 1D–F). Approximately 80 % of seedlings and juveniles had a nearest conspecific neighbour within approx. 6 m whereas adults had a nearest conspecific neighbour within approx. 15 m, a two-fold increase when compared to the seedling and juvenile stages. The NDF also showed a significantly higher proportion of seedlings and juveniles that were closer to their nearest conspecific neighbour in the same stage when contrasted with the null model (Fig. 1D, E), whereas spatial associations among adults did not differ from the null model (Fig. 1F).

Fig. 1.

Fig. 1.

Open in a new tab

The univariate point pattern analysis of the different stages – seedlings, juveniles and adults for Attalea phalerata (A–F) and Astrocaryum murumuru (G–L). The univariate pair-correlation function g11(r) and nearest neighbour distribution function D11(r) over scale r of the data (black dots) is contrasted with the expected g11/D11 function under the heterogeneous Poisson null model (black line) and the simulation envelopes (dashed lines) being the 25th highest and lowest values of 999 Monte Carlo simulations of the null model. The null model used an Epanečnikov kernel estimate of the intensity of the pattern with bandwidth R=30 m. The P-value of the goodness-of-fit test indicates whether observed values depart significantly from the null model over the distance interval of 1–10 m.

The univariate PCF for Astrocaryum murumuru showed similarly aggregated patterns in the seedling and juvenile stages (Fig. 1G, H). The strength of associations in both stages was not significantly different (P > 0·7, Tukey’s non-parametric multiple comparisons). There was, however, a decrease in aggregation among conspecifics from the juvenile to adult stages (Fig. 1H–I). The spatial associations among conspecific adults did not differ from the null model (Fig. 1I). The univariate NDF for Astrocaryum murumuru indicated a shift in the proportion of palms away from the nearest conspecifics from the seedling to adult stages (Fig. 1J–L). Approximately 80 % of seedlings and juveniles had a nearest conspecific neighbour within ∼5 m, whereas adults had a nearest conspecific neighbour within ∼16 m, which is more than a three-fold increase when compared to the seedling and juvenile stages. When compared to the null model, the NDF also showed a significantly higher proportion of seedlings and juveniles that were closer to their nearest conspecific neighbour (Fig. 1J, K), while adults did not differ from the null model (Fig. 1L).

Distance-effects

In both palm species, the bivariate PCF and NDF indicated a decrease in positive associations between reproductive adults and conspecific non-reproductive stages through ontogeny (i.e. from seedlings to juveniles, and non-reproductive adults). The bivariate PCF indicated that Attalea phalerata seedlings were positively associated with conspecific reproductive adults with peak recruitment occurring at ∼4–10 m (Fig. 2A). Juveniles showed peak recruitment at ∼8 m (Fig. 2B), but their associations with conspecific reproductive adults did not depart from the null model. Non-reproductive adult associations with conspecific reproductive adults also did not differ from the null model (Fig. 2C). The strength of associations between reproductive adults and conspecific seedlings was significantly greater than reproductive adult associations with either conspecific juveniles or non-reproductive adults (P<0·005, Tukey’s non-parametric multiple comparisons). The bivariate NDF for Attalea phalerata indicated a shift in the proportion of non-reproductive palms away from the nearest conspecific reproductive adult from the seedling to non-reproductive adult stages (Fig. 2D–F). Approximately 80 % of seedlings, juveniles and non-reproductive adults had a nearest conspecific reproductive adult neighbour within approx. 17, 20 and 22 m, respectively. Compared to the null model, the bivariate NDF showed a significantly higher proportion of seedlings closer to their nearest conspecific reproductive adult neighbour (Fig. 2D), while juveniles and adults (non-reproductive and reproductive) did not differ from the null model (Fig. 2E, F). The bivariate NDF D12 indicated some adults were surrounded by fewer conspecific seedlings than predicted by the null model (Supplementary Data Fig. S2a).

Fig. 2.

Fig. 2.

Open in a new tab

The bivariate point pattern analysis between reproductive adults and conspecific non-reproductive stages for Attalea phalerata (A–F) and Astrocaryum murumuru (G–L). The bivariate pair-correlation function g12(r) and nearest neighbour distribution function D21 (r) over scale r of the observed data (black dots) is contrasted with the bivariate g12/D21 function under the heterogeneous Poisson null model (black line) and simulation envelopes (dashed lines) being the 25th highest and lowest values of 999 Monte Carlo simulations of the null model. For g12 the null model used an Epanečnikov kernel estimate of the intensity of the non-reproductive palm (pattern 2) with bandwidth R=30 m, while the locations of conspecific reproductive adults (pattern 1) remained fixed. For D21 the null model used an Epanečnikov kernel estimate of the intensity of the non-reproductive palm (pattern 1) with bandwidth R=30 m, while the locations of conspecific reproductive adults (pattern 2) remained fixed.

In the case of Astrocaryum murumuru, the bivariate PCF indicated seedling recruitment was concave-shaped and seedlings were positively associated with conspecific reproductive palms with peak recruitment occurring at 1–7 m (Fig. 2G). Juveniles showed a small peak in recruitment at ∼ 8^m (Fig. 2H), but both juvenile and non-reproductive adult associations with conspecific reproductive adults did not depart from the null model (Fig. 2H, I). The strength of associations between reproductive adults and conspecific seedlings was significantly greater than reproductive adult associations with either conspecific juveniles or non-reproductive adults (P<0·005, Tukey’s non-parametric multiple comparisons). The bivariate NDF for Astrocaryum murumuru indicated a shift in the proportion of palms away from the nearest conspecific reproductive adult from the seedling to non-reproductive adult stages (Fig. 2J–L). Approximately 80 % of seedling, juveniles and adults had a nearest conspecific reproductive neighbour within approx. 17, 18 and 19 m, respectively. Compared to the null model, the bivariate NDF also showed a significantly higher proportion of seedlings closer to their nearest conspecific reproductive adult neighbour (Fig. 2J), while juveniles and adults (non-reproductive and reproductive) did not differ from the null model (Fig. 2K, L). The bivariate NDF D12 indicated some adults were surrounded by fewer conspecific juveniles than predicted by the null model (Supplementary Data Fig. S2e).

Density-effects on palm biomass

Density-correlation functions indicated that the biomass of the seedling and juvenile population for Attalea phalerata was negatively correlated with the density of conspecifics (Fig. 3A). The Attalea phalerata adult population (non-reproductive and reproductive) also showed density-effects in their biomass (Fig. 3B). For Astrocaryum murumuru, on the other hand, we found no spatial structure in their biomass for the seedling and juvenile, as well as for the adult population (Fig. 3C, D).

Fig. 3.

Fig. 3.

Open in a new tab

Effect of conspecific palm density on palm biomass. The density-correlation function CmK(r) is the correlation between the number of conspecific neighbours within distance r and the biomass of Attalea phalerata (A, seedlings and juveniles; B, non-reproductive and reproductive adults) and Astrocaryum murumuru (C, seedlings and juveniles; D, non-reproductive and reproductive adults). Other conventions are as in Fig. 1.

Distance-effects on palm biomass

The r-mark correlation function indicated that the biomass of Attalea phalerata seedlings was significantly lower near conspecific reproductive adults (Fig. 4A). In addition, the biomass of Attalea phalerata juveniles was significantly lower near conspecific reproductive adults (Fig. 4B). There were no marked differences between the two results. These results contrast with the biomass of Attalea phalerata non-reproductive adults near conspecific reproductive adults, which did not depart from the null model (Fig. 4C). For Astrocaryum murumuru, there was no significant spatial structure in the biomass of Astrocaryum murumuru seedlings, juveniles and non-reproductive adults found near conspecific reproductive adults (Fig. 4D–F).

Fig. 4.

Fig. 4.

Open in a new tab

Effect of reproductive adult distance on conspecific non-reproductive palm biomass. The r-mark correlation function km.(r) shows the correlation in the distance r from a reproductive adult and mean biomass of a conspecific non-reproductive palm at for Attalea phalerata (A, seedling; B, juvenile; C, non-reproductive adult) and Astrocaryum murumuru (D, seedling; E, juvenile; F, non-reproductive adult). Other conventions are as in Fig. 1

DISCUSSION

Our study of two Neotropical palms highlights the role of natural enemies in regulating plant populations across life stages, as well as the traits that mediate the strengths of those effects. We found that Attalea phalerata and Astrocaryum murumuru exhibited patterns of density- and distance-effects across their life stages, although the significance and strength of each effect differed across the life stages of each species and corresponded with predictions arising from their respective phenology and shade tolerance.

Distance- and density-effects in the seed stage

Bruchid predation on Attalea phalerata seeds was both negative distance- and density-dependent, consistent with the Janzen–Connell hypothesis prediction that distance from conspecific reproductive palms (distance-effects) and conspecific seed density (density-effects) affect predation probability and intensity by bruchid beetles. These predation results for Attalea phalerata were consistent with results of the predation patterns for Astrocaryum murumuru seeds (Terborgh et al., 1993; Cintra, 1997). Despite the similar predation patterns for both palm species, our results supported our prediction that seed predators would have a stronger impact on seedling recruitment for asynchronously fruiting Attalea phalerata compared to synchronously fruiting Astrocaryum murumuru. The hump-shaped recruitment for Attalea phalerata (Fig. 2A) suggests seed predation was able to decrease recruitment beneath the stems of conspecific reproductive adults. The influence of bruchid seed predators in generating distance-dependent seedling recruitment has similarly been reported for another palm species, Iriartea deltoidea (Wyatt and Silman, 2004). It is possible, however, that in addition to seed predation, mortality in the seed–seedling transition for Attalea phalerata may have also contributed to the hump-shaped seedling recruitment.

In contrast to the pattern for Attalea phalerata, seedling recruitment for Astrocaryum murumuru was highest beneath the stems of conspecific reproductive adults, indicating seed predation did not have a strong impact on seedling recruitment. This concave-shaped recruitment pattern (Fig. 2G) suggests that many Astrocaryum murumuru seeds escaped predation and survived into the seedling stage near conspecific reproductive adults. The high recruitment near conspecific adults probably occurred as a consequence of Astrocaryum murumuru palms fruiting synchronously and satiating predators. Seed predator satiation has similarly been reported for other synchronously fruiting species (e.g. Augspurger, 1981; Curran and Leighton, 2000). The concave-shaped recruitment pattern observed for Astrocaryum murumuru under predator satiation has also been predicted by theoretical models (Nathan and Casagrandi, 2004). In addition to seed predator satiation, the fruiting season for Astrocaryum murumuru palms also occurs during a period of high fruit abundance (Terborgh and Andresen, 1998). Hence, it is possible that fruits from other species may have been available to the bruchid seed predators to additionally contribute to predator satiation. Overall our results support our predictions that the JC effect in the seed stage had a weaker impact on seedling recruitment for synchronously fruiting Astrocaryum murumuru than asynchronously fruiting Attalea phalerata.

Density-effects in the post-seed stages

Our spatial analyses of palm patterns and palm biomass indicated distinct density-effects on the two palm species that corresponded with their shade tolerance. We detected strong density-effects in the early stages (seedling–juvenile transition) for shade-intolerant Attalea phalerata. In contrast, density-effects were weak or absent in the early stages for shade-tolerant Astrocaryum murumuru. These findings support our prediction that in the early stages, shade-tolerant Astrocaryum murumuru is more resilient than shade-intolerant Attalea phalerata against density-dependent enemies and is consistent with the growth–survival trade-off hypothesis (Coley et al., 1985). The strong density-effects that we detected in the early stages for Attalea phalerata were similarly found for several dicot tree species (Zhu et al., 2015). In addition to natural enemies contributing to density-effects in shade-intolerant Attalea phalerata, we cannot rule out the possibility that conspecific competition may have also contributed to the density-dependent mortality in Attalea phalerata. Further studies will be necessary to assess the relative importance of natural enemies vs. conspecific competition in contributing to the density-dependent mortality in the early stages of Attalea phalerata.

In the shade-tolerant Astrocaryum murumuru, the weak density-effects in the seedling and juvenile stages could have been due to their better defended leaves compared to Attalea phalerata. The spines and thicker leaves of Astrocaryum murumuru may afford better protection against herbivores in the early stages although this still needs to be confirmed with further studies. In the later stages (juvenile to adult), decreasing aggregation among conspecific Astrocaryum murumuru (cf. Fig. 1H and Fig. 1I) may have been due to competitive interactions rather than natural enemies. Competition for resources such as light intensifies as a plant grows older and this can increase competition among conspecific and/or heterospecific neighbours and density-dependent mortality in the later life stages (Uriarte et al., 2004; Stoll and Newbery, 2005; Velázquez et al., 2015; Zhu et al., 2015). Arborescent palms in particular undergo a vertical growth phase in the juvenile to adult stages when palm trunks emerge from the ground and require high light conditions (Kahn and de Granville, 1992). Such growth phases may exacerbate competition for light among neighbouring plants and contribute to density-dependent mortality in the later stages.

Distance-effects in the post-seed stages

In contrast to density-effects, our analyses indicated that both shade-intolerant Attalea phalerata and shade-tolerant Astrocaryum murumuru experienced strong distance-effects in the early stages (seedling to juvenile, Fig. 2). Previous investigations have similarly found distance-effects in the seedlings of Astrocaryum murumuru (Wyatt and Silman, 2004). The strong distance-effects observed in our Astrocaryum murumuru population suggest that shade-tolerance was not associated with greater defence against the primary natural enemies that were attacking these palms. Such a scenario is possible when the natural enemies involved are mammal herbivores, which are not influenced by a plant’s shade tolerance, compared to enemies such as soil pathogens that have a greater impact on plants that are more shade-intolerant (McCarthy-Neumann and Ibáñez, 2013). Peccaries have previously been reported as an important agent of mortality of Astrocaryum murumuru seedlings near conspecific adult palms as a result of their trampling and herbivory activities (Wyatt and Silman, 2004; Beck, 2007). We thus hypothesize that peccaries, rather than soil pathogens, were also the primary contributors of distance-effects for our Astrocaryum murumuru population. This hypothesis is further supported by the fact that we did not find palm biomass to be suppressed near conspecific reproductive adults (Fig. 4D–F), which would have occurred if soil pathogens were involved (e.g. Peters, 2003).

CONCLUSIONS

Our study illuminates the potential interactions between JC effects and life history traits as well as life stages (Comita et al., 2014). We found distance- and density-effects varied among the life stages and in a manner that corresponded with differences in the phenology and shade tolerance of each palm species. If such connections can be generalized across many species, it provides a route to understanding how trait-mediated JC effects scale up to whole communities of species, a necessary step for a trait-based community ecology in tropical forests (McGill et al., 2006; Kraft and Ackerly, 2010).

SUPPLEMENTARY INFORMATION

Supplementary data are available online at www.aob.oxfordjournals.org and consist of the following. Figure S1. Attalea phalerata and Astrocaryum murumuru individuals in the study plot. Figure S2. The bivariate nearest neighbour distribution function D12 for Attalea phalerata and Astrocaryum murumuru.

Supplementary Material

Supplementary Data
Click here for additional data file. (607.9KB, zip)

ACKNOWLEDGEMENTS

We thank Thorsten Wiegand and José Fedriani for their advice on aspects of the spatial analyses, and reviewers for their comments that improved the manuscript. We also thank M. Foster, V. Swamy and J. Terborgh for their assistance at the Cocha Cashu study site, and Instituto Nacional de Recursos Naturales, Peru, for permission to conduct the field research in Manu National Park. This work was supported by the Lindbergh Foundation and the Francis J. Bossuyt Fellowship. J.C. and E.P.E. were supported by subsidy funding to the Okinawa Institute of Science and Technology.

LITERATURE CITED

  1. Augspurger CK. 1981. Reproductive synchrony of a tropical shrub: experimental studies on effects of pollinators and seed predators in Hybanthus prunifolius (Violaceae). Ecology 62: 775–788. [Google Scholar]
  2. Beck H. 2007. Synergistic impacts of ungulates and falling palm fronds on saplings in the Amazon. Journal of Tropical Ecology 23: 599–602. [Google Scholar]
  3. Boege K, Marquis RJ.. 2005. Facing herbivory as you grow up: the ontogeny of resistance in plants. Trends in Ecology & Evolution 20: 441–448. [DOI] [PubMed] [Google Scholar]
  4. Carson WP, Anderson JT, Schnitzer SA, Leigh EG.. 2008. Challenges associated with testing and falsifying the Janzen–Connell hypothesis: a review and critique In: Carson WP, Schnitzer SA, eds. Tropical forest community ecology. Oxford: Blackwell, 210–241. [Google Scholar]
  5. Chesson P. 2000. Mechanisms of maintenance of species diversity. Annual Review of Ecology and Systematics 31: 343–366. [Google Scholar]
  6. Choo J, Juenger T, Simpson BB.. 2012. Consequences of frugivore‐mediated seed dispersal for the spatial and genetic structures of a neotropical palm. Molecular Ecology 21: 1019–1031. [DOI] [PubMed] [Google Scholar]
  7. Cintra R. 1997. A test of the Janzen–Connell model with two common tree species in Amazonian forest. Journal of Tropical Ecology 13: 641–658. [Google Scholar]
  8. Coley PD. 1980. Effects of leaf age and plant life history patterns on herbivory. Nature 284: 545–546. [Google Scholar]
  9. Coley PD, Barone JA.. 1996. Herbivory and plant defenses in tropical forests. Annual Review of Ecology and Systematics 27: 305–335. [Google Scholar]
  10. Coley PD, Bryant JP, Chapin FS.. 1985. Resource availability and plant antiherbivore defense. Science 230: 895–899. [DOI] [PubMed] [Google Scholar]
  11. Comita LS, Hubbell SP.. 2009. Local neighborhood and species’ shade tolerance influence survival in a diverse seedling bank. Ecology 90: 328–334. [DOI] [PubMed] [Google Scholar]
  12. Comita LS, Queenborough SA, Murphy SJ, et al. 2014. Testing predictions of the Janzen–Connell hypothesis: a meta‐analysis of experimental evidence for distance‐and density‐dependent seed and seedling survival. Journal of Ecology 102: 845–856. [DOI] [PMC free article] [PubMed] [Google Scholar]
  13. Connell JH. 1971. On the role of natural enemies in preventing competitive exclusion in some marine animals and in rain forest trees In: den Boer PJ, Gradwell GR, eds. Dynamics of populations. Wageningen: Pudoc, 298–312. [Google Scholar]
  14. Curran LM, Leighton M.. 2000. Vertebrate responses to spatiotemporal variation in seed production of mast-fruiting Dipterocarpaceae. Ecological Monographs 70: 101–128. [Google Scholar]
  15. Diggle PJ. 2003. Statistical analysis of spatial point patterns. London: Arnold Publishers. [Google Scholar]
  16. Fedriani JM, Wiegand T, Calvo G, et al. 2015. Unravelling conflicting density and distance dependent effects on plant reproduction using a spatially explicit approach. Journal of Ecology 103: 1344–1353. [Google Scholar]
  17. Ferreira A, Bruna EM, Vasconcelos HL.. 2011. Seed predators limit plant recruitment in Neotropical savannas. Oikos 120: 1013–1022. [Google Scholar]
  18. Gentry AH. 1986. Species richness and floristic composition of the Choco region plant communities. Caldasia 15: 71–75. [Google Scholar]
  19. Gentry AG, Terborgh J.. 1990. Composition and dynamics of the Cocha Cashu ‘mature’ floodplain forest In: Gentry AG, ed. Four neotropical rainforests. New Haven, CT: Yale University Press, 542–564. [Google Scholar]
  20. Getzin S, Worbes M, Wiegand T, Wiegand K.. 2011. Size dominance regulates tree spacing more than competition within height classes in tropical Cameroon. Journal of Tropical Ecology 27: 93–102. [Google Scholar]
  21. Getzin S, Wiegand T, Hubbell SP.. 2014. Stochastically driven adult–recruit associations of tree species on Barro Colorado Island. Proceedings of the Royal Society of London B: Biological Sciences 281: 20140922. [DOI] [PMC free article] [PubMed] [Google Scholar]
  22. Gilbert GS. 2005. Dimensions of plant disease in tropical forests In: Gilbert GS, Burslem DFRP, Pinard MA, Hartley SE eds. Biotic interactions in the tropics: their role in the maintenance of species diversity. Cambridge: Cambridge University Press, 141–164. [Google Scholar]
  23. Gilbert GS, Foster RB, Hubbell SP.. 1994. Density and distance-to-adult effects of a canker disease of trees in a moist tropical forest. Oecologia 98: 100–108. [DOI] [PubMed] [Google Scholar]
  24. Goodman RC, Phillips OL, Torres DDC, et al. 2013. Amazon palm biomass and allometry. Forest Ecology and Management 310: 994–1004. [Google Scholar]
  25. Hammond DS, Brown VK.. 1998. Disturbance, phenology and life-history characteristics: factors influencing distance/density-dependent attack on tropical seeds and seedlings In: Newbery DM, Prins HHT, Brown ND, eds. Dynamics of tropical communities: the 37th symposium of the British Ecological Society, Cambridge University, 1996. Cambridge: Blackwell Science Ltd, 51–78. [Google Scholar]
  26. Herms DA, Mattson WJ.. 1992. The dilemma of plants: to grow or defend. Quarterly Review of Biology 67: 283–335. [Google Scholar]
  27. Hubbell SP. 2001. The unified neutral theory of biodiversity and biogeography. Princeton, NJ: Princeton University Press. [Google Scholar]
  28. Illian J, Penttinen A, Stoyan H, Stoyan D.. 2008. Statistical analysis and modelling of spatial point patterns. Chichester: Wiley-Interscience. [Google Scholar]
  29. Jacquemyn H, Endels P, Honnay O, Wiegand T.. 2010. Evaluating management interventions in small populations of a perennial herb Primula vulgaris using spatio temporal analyses of point patterns. Journal of Applied Ecology 47: 431–440. [Google Scholar]
  30. Jansen PA, Bongers F, Van Der Meer PJ.. 2008. Is farther seed dispersal better? Spatial patterns of offspring mortality in three rainforest tree species with different dispersal abilities. Ecography 31: 43–52. [Google Scholar]
  31. Janzen DH. 1970. Herbivores and the number of tree species in tropical forests. American Naturalist 104: 501–528. [Google Scholar]
  32. Janzen DH. 1971. Seed predation by animals. Annual Review of Ecology and Systematics 2: 465–492. [Google Scholar]
  33. Johnson DJ, Beaulieu WT, Bever JD, Clay K.. 2012. Conspecific negative density dependence and forest diversity. Science 336: 904–907. [DOI] [PubMed] [Google Scholar]
  34. Kahn F, de Granville JJ.. 1992. Palms in forest ecosystems of Amazonia. Heidelberg: Springer-Verlag. [Google Scholar]
  35. Kobe RK, Vriesendorp CF.. 2011. Conspecific density dependence in seedlings varies with species shade tolerance in a wet tropical forest. Ecology Letters 14: 503–510. [DOI] [PubMed] [Google Scholar]
  36. Kraft NJB, Ackerly DD.. 2010. Functional trait and phylogenetic tests of community assembly across spatial scales in an Amazonian forest. Ecological Monographs 80: 401–422. [Google Scholar]
  37. Loosmore NB, Ford ED.. 2006. Statistical inference using the G or K point pattern spatial statistics. Ecology 87: 1925–1931. [DOI] [PubMed] [Google Scholar]
  38. Martínez I, Taboada FG, Wiegand T, Obeso JR.. 2013. Spatial patterns of seedling-adult associations in a temperate forest community. Forest Ecology and Management 296: 74–80. [Google Scholar]
  39. McCarthy-Neumann S, Ibáñez I.. 2013. Plant-soil feedback links negative distance dependence and light gradient partitioning during seedling establishment. Ecology 94: 780–786. [Google Scholar]
  40. McGill BJ, Enquist BJ, Weiher E, Westoby M.. 2006. Rebuilding community ecology from functional traits. Trends in Ecology & Evolution 21: 178–185. [DOI] [PubMed] [Google Scholar]
  41. McIntire EJB, Fajardo A.. 2009. Beyond description: the active and effective way to infer processes from spatial patterns. Ecology 90: 46–56. [DOI] [PubMed] [Google Scholar]
  42. Messina FJ. 2002. Host discrimination by seed parasites In: Edwin EL, Campbell JF, Sukdheo MVK, eds. The behavioural ecology of parasites. New York: CAB International, 65–87. [Google Scholar]
  43. Nathan R, Casagrandi R.. 2004. A simple mechanistic model of seed dispersal, predation and plant establishment: Janzen‐Connell and beyond. Journal of Ecology 92: 733–746. [Google Scholar]
  44. Packer A, Clay K.. 2000. Soil pathogens and spatial patterns of seedling mortality in a temperate tree. Nature 404: 278–281. [DOI] [PubMed] [Google Scholar]
  45. Paine CET, Beck H.. 2007. Seed predation by neotropical rain forest mammals increases diversity in seedling recruitment. Ecology 88: 3076–3087. [DOI] [PubMed] [Google Scholar]
  46. Peters HA. 2003. Neighbour-regulated mortality: the influence of positive and negative density dependence on tree populations in species-rich tropical forests. Ecology Letters 6: 757–765. [Google Scholar]
  47. Piao T, Comita LS, Jin G, Kim JH.. 2013. Density dependence across multiple life stages in a temperate old-growth forest of northeast China. Oecologia 172: 207–217. [DOI] [PMC free article] [PubMed] [Google Scholar]
  48. R Core Team. 2017. R: A language and environment for statistical computing. R Foundation for Statistical Computing, Vienna, Austria; https://www.R-project.org/. [Google Scholar]
  49. Schupp EW. 1990. Annual variation in seedfall, postdispersal predation, and recruitment of a neotropical tree. Ecology 71: 504–515. [Google Scholar]
  50. Silman MR, Terborgh JW, Kiltie RA.. 2003. Population regulation of a dominant rain forest tree by a major seed predator. Ecology 84: 431–438. [Google Scholar]
  51. Skaug H, Fournier D, Nielsen A, Magnusson A, Bolker B.. 2014. glmmADMB: generalized linear mixed models using AD Model Builder. R package version 0.8.0 http://glmmadmb.r-forge.r-project.org/.
  52. Stoll P, Newbery DM.. 2005. Evidence of species-specific neighborhood effects in the Dipterocarpaceae of a Bornean rain forest. Ecology 86: 3048–3062. [DOI] [PubMed] [Google Scholar]
  53. Terborgh J. 2012. Enemies maintain hyperdiverse tropical forests. The American Naturalist 179: 303–314. [DOI] [PubMed] [Google Scholar]
  54. Terborgh J, Andresen E.. 1998. The composition of Amazonian forests: patterns at local and regional scales. Journal of Tropical Ecology 14: 645–664. [Google Scholar]
  55. Terborgh J, Losos E, Riley MP, Riley MB.. 1993. Predation by vertebrates and invertebrates on the seeds of five canopy tree species of an Amazonian forest. Plant Ecology 107: 375–386. [Google Scholar]
  56. Terborgh J, Nuñez-Iturri G, Pitman NCA, et al. 2008. Tree recruitment in an empty forest. Ecology 89: 1757–1768. [DOI] [PubMed] [Google Scholar]
  57. Tomlinson PB. 1990. The structural biology of palms. Oxford: Oxford University Press. [Google Scholar]
  58. Uriarte M, Condit R, Canham CD, Hubbell SP.. 2004. A spatially explicit model of sapling growth in a tropical forest: does the identity of neighbours matter? Journal of Ecology 92: 348–360. [Google Scholar]
  59. Velázquez E, Kazmierczak M, Wiegand T.. 2015. Spatial patterns of sapling mortality in a moist tropical forest: consistency with total density dependent effects. Oikos 125: 872–882. [Google Scholar]
  60. Wiegand T, Moloney KA.. 2004. Rings, circles, and null-models for point pattern analysis in ecology. Oikos 104: 209–229. [Google Scholar]
  61. Wiegand T, Moloney KA.. 2013. A handbook of spatial point pattern analysis in ecology. Boca Raton, FL: Chapman and Hall/CRC press. [Google Scholar]
  62. Wiegand T, Gunatilleke S, Gunatilleke N.. 2007a. Species associations in a heterogeneous Sri Lankan dipterocarp forest. The American Naturalist 170: 77–95. [DOI] [PubMed] [Google Scholar]
  63. Wiegand T, Gunatilleke S, Gunatilleke N, Okuda T.. 2007b. Analyzing the spatial structure of a Sri Lankan tree species with multiple scales of clustering. Ecology 88: 3088–3102. [DOI] [PubMed] [Google Scholar]
  64. Wilson DE, Janzen DH.. 1972. Predation on Scheelea palm seeds by bruchid beetles: seed density and distance from the parent palm. Ecology 53: 954–959. [Google Scholar]
  65. Wright JS. 2002. Plant diversity in tropical forests: a review of mechanisms of species coexistence. Oecologia 130: 1–14. [DOI] [PubMed] [Google Scholar]
  66. Wyatt JL, Silman MR.. 2004. Distance-dependence in two Amazonian palms: effects of spatial and temporal variation in seed predator communities. Oecologia 140: 26–35. [DOI] [PubMed] [Google Scholar]
  67. Zhu Y, Comita LS, Hubbell SP, Ma K.. 2015. Conspecific and phylogenetic density dependent survival differs across life stages in a tropical forest. Journal of Ecology 103: 957–966. [Google Scholar]
  68. Zuur A, Ieno EN, Walker N, Saveliev AA, Smith GM.. 2009. Mixed effects models and extensions in ecology with R. New York: Springer. [Google Scholar]

Associated Data

This section collects any data citations, data availability statements, or supplementary materials included in this article.

Supplementary Materials

Supplementary Data
Click here for additional data file. (607.9KB, zip)

Articles from Annals of Botany are provided here courtesy of Oxford University Press

RESOURCES