Skip to main content

Litter decomposition rates across tropical montane and lowland forests are controlled foremost by climate

Cite this dataset

Ostertag, Rebecca et al. (2022). Litter decomposition rates across tropical montane and lowland forests are controlled foremost by climate [Dataset]. Dryad.


The “hierarchy of factors” hypothesis states that decomposition rates are controlled primarily by climatic, followed by biological and soil variables. Tropical montane forests (TMF) are globally important ecosystems, yet there have been limited efforts to provide a biome-scale characterization of litter decomposition. We designed a common litter decomposition experiment replicated in 23 tropical montane sites across the Americas, Asia, and Africa and combined these results with a previous study of 23 sites in tropical lowland forests (TLF).  Specifically, we investigated (1) spatial heterogeneity in decomposition, (2) the relative importance of biological factors that affect leaf and wood decomposition in TMF and, (3) the role of climate in determining leaf litter decomposition rates within and across the TMF and TLF biomes. Litterbags of two mesh sizes containing Laurus nobilis leaves or birchwood popsicle sticks were spatially dispersed and incubated in TMF sites, for 3 and 7 months on the soil surface and at 10-15 cm depth. The within-site replication demonstrated spatial variability in mass loss. Within TMF, litter type was the predominant biological factor influencing decomposition (leaves > wood), with mesh and burial effects playing a minor role. When comparing across TMF and TLF, climate was the predominant control over decomposition, but the Yasso07 global model (based on mean annual temperature and precipitation) only modestly predicted decomposition rate. Differences in controlling factors between biomes suggest that TMF, with their high rates of carbon storage, must be explicitly considered when developing theory and models to elucidate carbon cycling rates in the tropics.


Study sites

We conducted a large-scale field experiment across 23 TMF sites (column B in data file, rows 2-24) located in 14 countries in the Americas, Asia, Africa, and several islands in the Caribbean and the Pacific . These sites spanned a wide range of elevations (600-3202 m), mean annual temperatures (MAT, 3.0-23.0 ⁰C), mean annual precipitations (MAP, 335-5010 mm), and dry season length (0-6 months). We augmented our study by including a previously published TLF dataset (Powers et al. 2009) that comprised of 23 tropical sites in 14 countries in Asia, Latin America, and Oceania, spanning a wide range of elevations (5-1200 m), mean annual temperatures (MAT, 18.5-26.5 ⁰C), mean annual precipitations (MAP, 760-5797 mm), and dry season length (0-10 months) (column B in data file, rows 25-45). 

To examine climatic control over decomposition, we used three homogenized global datasets (1-km resolution) to characterize each site based on 21 variables (Table S2).  WorldClim (Hijmans et al. 2005; provided 19 bioclimatic variables describing annual trends, seasonality, and extreme conditions and CGIAR CSI provided potential evapotranspiration (Zomer et al. 2007, Zomer et al. 2008;  Due to the importance of clouds in TMF, we used mean annual cloud frequency (MODCF mean annual; % cloudy days) derived from MODIS (Wilson & Jetz 2016;

Experimental design

The CloudNet study followed the general design of the Powers et al. (2009) study to allow a combined analysis, with two exceptions: litter type and sampling intervals. Following the TLF study we used bay leaves, but instead of using raffia leaves we used wood as a second substrate (Powers et al. 2009). In contrast to Powers et al. (2009) that incubated the litterbags for 1, 3, 5, 7, and 9 months, we only incubated the bags for 3 and 7 months, times selected based on the observed decomposition rate for bay leaves in the Powers et al. (2009) dataset. Bay leaf (Laurus nobilis) (Frontier Co-op, was the common leaf material in the TMF and TLF studies, and white birch (Betula papyrifera) wooden popsicle sticks (1 cm x 11.5 cm; Ben Franklin, Hilo, HI, USA) the common wood material for the TMF sites. To determine initial mass of the litter types for the litterbags, the bay leaves (with absent or very short petioles) were oven dried at 70° C for at least 48 hours and subsequently weighed and placed into mesh litter bags (Figure 1b); the wood was initially dried at 120 °C to release volatiles following the standardizing wood pre-treatment protocol (Cheesman et al. 2018), and then dried in the oven at 70°C for 48 hours.

Mesh bags (10 x 15 cm) contained either 1.00 ± 0.01 g (mean ± 1 SD) dry mass of whole bay leaves or one popsicle stick (1.34 ± 0.12 g). Two nylon mesh sizes were used (52 μm-mesh and 1920 μm-mesh; Component Supply Company, Fort Meade, FL) to selectively exclude mesofauna (defined as 100 μm to 2 mm) and fine roots. Litterbags were assembled in Hawai‘i and mailed to all participants.

The experiment started at each of the 23 TMF sites between May and Dec 2017 and concluded between Dec 2017 and July 2018. At each site, we incorporated within-site replication by having litterbags deployed along a 50 m transect, subdivided into six stations 10 m apart. At each station, we deployed 16 bags: 2 litter types, 2 burial depths, 2 mesh size treatments, and 2 collection times for a total of 2208 bags (Figure 1b). The burial depths were on top of the litter layer at the soil surface and 10–15 cm depth. We used a knife or machete to cut into the soil at a 45o angle, inserted the bag and filled the hole. At each station, bags were anchored to a central stake using fishing line to facilitate retrieval.  The initial bag placement occurred right before or at the beginning of the wet season (May-Dec 2017) for all sites with the exception of the Ngel Nyaki site, where the start of the experiment occurred in Nov 2017 at the beginning of the November-March dry season. After retrieval at each harvesting time, the bags were cut open, the contents removed, and gently washed with tap water. Any soil or root material attached to the litter type was removed using tweezers. Samples were dried at 70 oC and weighed to ± 0.01 g. 

To compare decomposition rates between tropical montane and lowland forest, we subsampled the TLF dataset to include only bay leaves and match the 3- and 7- months sampling times.  In the TMF dataset, the 3- and 7-month harvests occurred after 95 ± 8 days (mean ± 1 SD) and 226 ± 23 days, respectively; in TLF, those harvest times were 106 ± 22 days and 218 ± 15 days. This resulted in a subset of 483 bags (3 bags x 2 burial depths x 2 mesh treatments x 2 times) in the TLF dataset.

​​​​​​​Data analysis

We conducted extensive exploratory data analyses to clean the dataset, identify outliers and investigate the source of their distinctiveness, and examine collinearity among variables to be used in the main analyses (Zuur & Ieno 2016). We received information for 2200 out of the 2208 TMF bags that were deployed out in the field. After careful examination of the data, we retained 2162 for the analysis. For quality control, we subtracted the final weights of the litterbags from the initial weights with the expectation that differences should be positive; if the reported final weight of the litterbag was 10% greater than the initial weight, the point was discarded.  In some instances, we contacted the relevant co-author to clarify problematic data points that could not be resolved by the lead authors (e.g., mislabeled samples). Thus, 11 bags were discarded due to larger final versus original weights, and another 27 were discarded due to other reasons (e.g., lost, ruptured, unearthed bags). We used R version 4.0.3 (2020-10-10) to perform all subsequent analyses (R Core Team 2018).

Usage notes

There is a ReadMe file to explain the variables in the Datafile. 


National Science Foundation, Award: DEB-1146446

National Science Foundation, Award: (DEB)‐1146446