Litterfall dynamics along a successional gradient in a Brazilian tropical dry forest

Saimo R Souza,Mário M Espírito-Santo,Maria D M Veloso,Betânia G Souza E Brito,Jhonathan O Silva,Geraldo Wilson Fernandes,Arturo Sánchez-Azofeifa

doi:10.1186/s40663-019-0194-y

Abstract

BackgroundThis study aimed to determine the influence of soil traits, vegetation structure and climate on litterfall dynamics along a successional gradient in a tropical dry forest (TDF) in southeastern Brazil. We used a chronosequence design consisting of three successional stages (early, intermediate, and old-growth) defined based on forest age and vertical and horizontal structures.MethodsLitterfall was recorded monthly for three years in 12 plots of 50 × 20 m (four plots per stage) where vegetation parameters (species richness, basal area, density and height for trees with diameter at breast height ≥ 5 cm) and soil chemical and physical traits were previously obtained. We placed eight 0.5-m2 litter traps in each plot, totalling 96 traps. Samples were sorted into leaves, twigs, reproductive parts, and debris.ResultsLitterfall mass was composed mainly of leaves and varied slightly among years (4 to 4.5 Mg∙ha− 1), within the range observed for other TDFs. Annual litterfall mass was higher at the old-growth forest than at the early and intermediate forest stages and this successional pattern was driven by vegetation characteristics (forest structural parameters and plant functional groups) and soil traits related to water-holding capacity. Litter amount in the intermediate stage was lower than expected for its forest structure (and similar to the early stage), possibly because its higher soil clay content increased the water holding capacity and leaf retention during the dry season. Seasonal variations in monthly litterfall were strongly driven by forest deciduousness and affected by climatic factors related to water availability. This pattern was consistent across the successional gradient, although differences for each litterfall component were observed.ConclusionsOur results suggest that litter production in the studied TDF is influenced by multiple factors along succession, such as above-ground biomass and the degree of leaf retention mediated by soil water-holding capacity. Further studies on community phenological patterns can allow a better understanding of successional changes on litterfall and how fast this fundamental function recovers in secondary forests.

Highlights

This study aimed to determine the influence of soil traits, vegetation structure and climate on litterfall dynamics along a successional gradient in a tropical dry forest (TDF) in southeastern Brazil
The amount of litterfall is directly related to above-ground biomass accumulation (Lawrence 2005; Chave et al 2009; Feng et al 2019) and affected by soil traits involved in forest growth (Peña-Claros et al 2012; Yazaki et al 2016) and vegetative phenology (Holbrook et al 1995; Giraldo and Holbrook 2011)
Annual and successional variations in litterfall amount and composition In the studied TDF, litterfall was primarily composed by leaves (4.30 ± 0.01 Mg∙ha− 1), followed by twigs (0.94 ± 0.08 Mg∙ha− 1), reproductive parts (0.29 ± 0.02 Mg∙ha− 1), and debris (0.12 ± 0.01 Mg∙ha− 1)

Summary

Introduction

This study aimed to determine the influence of soil traits, vegetation structure and climate on litterfall dynamics along a successional gradient in a tropical dry forest (TDF) in southeastern Brazil. At coarser spatial scales (i.e., regional or global), above-ground biomass and decomposition rates in TDFs increase with mean annual precipitation (Martínez-Yrízar 1995; Becknell et al 2012; Gei and Powers 2014). The amount of litterfall is directly related to above-ground biomass accumulation (Lawrence 2005; Chave et al 2009; Feng et al 2019) and affected by soil traits involved in forest growth (Peña-Claros et al 2012; Yazaki et al 2016) and vegetative phenology (Holbrook et al 1995; Giraldo and Holbrook 2011). Litterfall is affected by the available water in the soil (Holbrook et al 1995; Jaramillo et al 2011) which depends, in part, on soil texture (Pastor and Post 1986)

Objectives

Methods

Results

Discussion

Conclusion