Skip to main content

Relative effects of climate and litter traits on decomposition change with time, climate and trait variability

Cite this dataset

Canessa, Rafaella et al. (2020). Relative effects of climate and litter traits on decomposition change with time, climate and trait variability [Dataset]. Dryad.


  1. Climate and litter quality drive litter decomposition, but there is currently little consensus on their relative importance, likely because studies differ in the duration, the climatic gradients, and variability in litter-trait values. Understanding these drivers is important because they determine the direct and indirect (via vegetation composition) effects of climate change on decomposition and thereby on carbon and nutrient cycling.
  2. We studied how microclimate (soil moisture and temperature) and litter traits interactively affect litter mass loss, by using a reciprocal litter translocation experiment along a large climatic gradient in Chile. We followed decomposition for two years and used 30 plant species with a wide spectrum of functional-trait values.
  3. Litter traits had a strong impact on litter decomposition across the gradient, while an increase in decomposition with soil moisture was observed only in the wettest climates. Overall, soil moisture increased considerably in importance, relative to trait effects, at later decomposition stages, from ca. 15% of the importance of traits after 3 and 6 months to ca. 110% after 24 months. Moreover, analyzing subsets of the 30 species showed that trait effects on litter decomposition gained in importance when including a greater variation in trait values.
  4. Synthesis. The relative effects of litter traits and climate on decomposition depend on the ranges in climate and litter traits considered and change with time. Our study emphasizes the critical role of representative ranges in climate and functional trait values for understanding the drivers of litter decomposition and for improving predictions of climate-change effects on this important ecosystem process.


This dataset belongs to the publication: Canessa, R., van den Brink, L., Saldaña, A., Rios, R. S., Hättenschwiller, S., Mueller, C. W., ... & Bader, M. Y. (2020). Relative effects of climate and litter traits on decomposition change with time, climate and trait variability. Journal of Ecology.

It contains litter decomposition (mass loss, %) data of 30 plant species along a climatic gradient in Chile (6 study sites, each with 3 plots), where a litter translocation experiment was set for 2 years, with retrievals after 3, 6, 9, 12 and 24 months. The data set also contains macroclimatic variables (annual precipitation, mean annual temperature) at the site level and microclimatic variables (mean soil moisture and mean soil temperature) at the plot level and per decomposition period. Additionally, the data set contains all plant functional traits used in this study, averaged at the species level. All methods described below belong to the mentioned publication.

Study sites and climatic data

We selected six sites with contrasting macroclimatic conditions: arid desert (Pampa Blanca - Pan de Azúcar National Park, henceforth, “AD” for Arid Dry), arid desert with fog influence (Las Lomitas - Pan de Azúcar National Park, “AF” for Arid Fog), semi-arid shrubland (Quebrada de Talca Private Reserve, “SA” for Semi-Arid), mediterranean forest (La Campana National Park, “ME” for Mediterranean), upland temperate rainforest (Nahuelbuta National Park, “TU” for Temperate Upland) and lowland temperate rainforest (Contulmo Natural Monument, “TL” for Temperate Lowland). During the study period, the climate was drier than usual in Central Chile (Garreaud et al., 2020), which was especially notable at the ME site, so that precipitation differed little and mean soil moisture did not differ at all between the ME and SA sites. The rainfall seasonality is similar at all sites, with rainfall occurring mainly during the austral winter (from May to August). AD and AF are located in the Atacama Desert, almost without rainfall. The coastal fog, however, is a relevant source of water at AF (Lehnert et al., 2018). Some fog-water input may also occur at AD on an irregular basis, but at much lower frequency and overall quantity compared to AF. More detailed information about the study sites is available in Bernhard et al. (2018) and Oeser et al. (2018). At each study site, three independent 10x10 m plots were selected.

Macroclimatic data (mean annual temperature -MAT- and annual precipitation -AP-) where retrieved from the following sources: 

MAT for AD and TU, and AP for TU: Ehlers, Blanckenburg, & Übernickel, 2019; data represent the experimental period of June 2016- May 2018.

MAT for AF: Laboratory for Climatology and Remote Sensing, University of Marburg, Germany, personal communication, April 2019.

MAT and AP for SA, ME and TL: INIA 2019. Stations Gabriela Mistral, La Cruz and La Isla were used for SA, ME and TL, respectively.

AP for AD: Thompson, Palma, Knowles, & Holbrook, 2003 (AP for AF is assumed to be the same as for AD).

We measured local soil moisture and temperature directly next to the litterbags (see following paragraphs) in each of the three plots per site for the duration of the experiment. We measured soil temperature at a depth of 2 cm using HOBO Micro Station dataloggers (H21-002) with two sensors (S-TMB-M002) and volumetric soil moisture at a depth of 14 cm using TMS-3 dataloggers (TOMST, Czech Republic). Based on the clay and sand content of our study sites (Bernhard et al., 2018), calibrations for sandy loam (AD and AF), loamy sand (SA and ME) and loamy soils (TH and TL) were used for the soil moisture measurements, as suggested by the provider (Wild et al. 2019). Sensors recorded data every 30 (temperature) or 15 minutes (moisture). We calculated mean soil temperature (MST) and soil volumetric water content (henceforth, mean soil moisture, MSM) for each decomposition period (0-3, 0-6, 0-9, 0-12 and 0-24 months, from June 2016 to June 2018). MST and MSM data were aggregated at the level of the plot (mean of two sensors).

Plant species and functional trait measurements

From the dominant plant species at each site, we selected five species per site (Table 1; at the AD site one lichen species was included) covering a wide spectrum of leaf traits expected to affect litter decomposition (Dias et al., 2017). For each species, we selected five individuals and measured specific leaf area (SLA, cm2 g-1) and force to punch (Fp, N cm-1) on 10 randomly-selected green leaves. The force required to punch a leaf (Fp, equivalent to leaf toughness) was measured for five healthy individuals located near the study plots, using ten leaves per individual. When it was impossible to obtain the leaves without heavily damaging the individual (Cistanthe sp, Cristaria integerrima, Festuca sp), we harvested as many leaves as possible and increased the number of individual plants until reaching 50 leaves per species in total, sampling from the spatially closest individuals. Upon harvest, leaves were wrapped in wet absorbing paper and placed in plastic bags, to avoid shrinkage and folding. Within the next eight hours, Fp was measured as maximum force to fracture, using a digital force gauge (IMADA, DS2-50N) attached to a 3 mm diameter cylindrical probe (IMADA, FR-EC-3J), mounted on a vertical manual test stand (IMADA, KV-50N), averaging two random punches per leaf, avoiding the midrib. For specific leaf area (SLA), ten additional leaves were collected from the same individuals, scanned and analyzed using the Image J software ( to estimate leaf area. These leaves were then dried at 70 °C for 72 h and weighed to get dry mass using an analytical scale. SLA was then calculated as the area (cm2) divided by the dry mass (g) of a leaf. Species-specific mean traits were calculated as average values across all individuals from a species for each site.

For three to five subsamples of leaf litter per species, obtained from leaf mixtures collected from at least 10 individuals (senescent litter used in the litter translocation experiment, see next section), we determined concentrations of lignin, carbohydrates, proteins, lipids, total phenolic compounds, tannins and the elements C, N, Al, Ca, Fe,  Mg, Mn, Na and P. Finally, we calculated the ratios C/N and Lignin/N. Each subsample (10 g) was milled to fine powder. For each subsample, carbon and nitrogen concentration (and thus C/N ratio) were determined at the University of Oldenburg, Germany, using a CHNS Analyzer (FLASHEA, 1112 Series; CE Elantech, Inc., Lakewood, USA). For each subsample, Al, Ca, Fe, K, Mg, Mn and P content were obtained by elemental analysis according to Jackson (1958) and Lim & Jackson (1982). Briefly, about 100 mg of milled samples were weighed in platinum beakers and treated with 1:1 mixture of 60% perchlorid acid (HClO4) and 65% nitric acid (HNO3). Samples were stepwise heated up to 150°C on a sand bath and left until the organic matter was oxidized. Subsequently, the temperature was increased to 300°C to vaporize the acid and 40% hydrofluoric acid (HF) was applied and left overnight. Afterwards, HClO4 was added, heated to 300°C to vaporize the acid again. Finally, 37% hydrochloric acid (HCl) was added and the sample was heated. The final sample was filtrated, and the total amount of the elements was analyzed by ICP-OES (Varian Vista-Pro). 

Total phenolics were measured colorimetrically using the Folin-Ciocalteau reagent following Marigo (1973) with gallic acid as a standard. From a 0.5 g sample of ground leaf material, total phenolics were extracted with 30 mL of a solution containing 50% methanol and 50% distilled water, shaken for 2 h and filtered (filter number 112, Durieux, Torcy, France). Tannins were measured according to Hagerman & Butler (1978) with the modified version of the protein precipitation method using the microplate assay (see Ann Hagerman’s “tannin handbook”, Tannins were extracted from a 0.1 g sample of ground leaf material with 1 mL of the same solvent used for the total phenolics extraction and left under ultrasound exposure for 30 min (samples within 15 mL Falcon tubes placed in a water bath). After extraction, the samples were centrifuged at 2000 rpm for 15 min and the supernatant kept for further tannin analyses. Microplates were prepared with 75 ml BSA (Bovine Serum Albumin) protein and buffer solutions (1:5) per well to which 25 ml of sample solution was added and mixed immediately with a microplate shaker for 10 min. After 15 min incubation, microplates were centrifuged for 1 h 15 min at 3700 rpm, and all supernatant was removed. 200 ml of SDS (sodium dodecyl sulfate) / TEA (triethanolamin) solution was added to each well and shaken for 10 min until precipitates were completely re-dissolved. 50 ml of ferric chloride was then added to each well, shaken, quickly centrifuged (500 rpm) for 1 min and then read with the microplate reader at 510 nm wavelength.

A combined sample of each species was subjected to 13C cross polarization magic angle spinning nuclear magnetic resonance (13C CPMAS-NMR) spectroscopy (Bruker AvanceIII200 spectrometer, Bruker, Billerica, MA, USA). Finely ground samples were spun in a zirconium oxide rotor at 6.8 kHz and analyzed with a recycle delay time of 1.0 s and 2500 scans for each sample – to obtain a sufficient signal to noise ratio some samples required more scans: Nolana crassulifolia, Nolana paradoxa, Tetragonia maritime, Porlieria chilensis (5000 each), and Frankenia chilensis (7500). The NMR spectra were integrated according to chemical shift regions: 0 – 45 alkyl C, 45 – 60 N-alkyl/methoxyl C, 60 – 95 O-alkyl C, 95 – 110 Di-O-Alkyl C, 110 – 145 aromatic C, 145 – 165 phenolic and 165 – 215 amide/carboxyl C (Nelson & Baldock 2005). Based on the NMR spectra, the relative content of carbohydrates (including cellulose, hemicellulose, muco-polysaccharides and smaller molecular weight saccharides), proteins, lignin and lipids were obtained using a molecular mixing model (Nelson & Baldock 2005). According to Bonanomi et al. (2013), a decomposition proxy based on the integrated NMR spectra was calculated as the ratio between 70-75 ppm (O-alkyl C in carbohydrates) and 52-57 ppm (methoxyl C in lignin).

For all analyses, traits were averaged at the species level per site.

Litter translocation experiment for decomposition data

We performed a full reciprocal litterbag translocation experiment, where litter from each species and site was incubated at each site (i.e., each climate zone). We harvested fresh senescent litter from a minimum of 10 individuals per species near the study plots during the late summer of 2016, either manually or with litter traps, depending on the height and deciduousness of the species. When used, litter traps were installed only under trees that allowed to obtain leaf litter of a single species to avoid potential contamination. For succulent species, green leaves were used. Litter was not washed to avoid the loss of leachable elements, and was oven-dried at 60° C for 72 h (or 96 h for succulent species) until constant weight. Subsamples of this litter material were used to determine nutrient contents (see previous section). We prepared 10x10 cm bags with a polyester mesh (1 mm). Bags were filled with 2 g of oven-dry single-species litter, recording the dry weight of each sample. For a few species with small leaf sizes, we used a second layer (same mesh size) to prevent losses. Litterbags were transported in individual paper bags and the initial weights corrected for any material left in these bags during transportation. One sample per incubation period, species and site was placed in each of the three plots per site. Considering 5 incubation periods, 30 species and 6 sites (climates), this triple replication added up to a total of 2700 litterbags.

The experiment was installed in early June 2016 (late autumn in the southern hemisphere). At each site, we carefully removed local soil litter and organic material, if present, and placed litterbags on top of the mineral soil. In study sites with a patchy vegetation cover, litterbags were placed between patches, but close (0.5 to 1 m) to shrubs. The experiment was protected against animals with poultry-wire mesh. In spite of this safeguard, some litterbags were damaged and could not be analyzed.

Groups of litterbags were harvested at five decomposition stages: 3, 6, 9, 12 and 24 months after installation, to observe both fast short-term changes and slower middle- and long-term changes (Zukswert & Prescott, 2017). At harvesting, litterbags were placed in individual paper bags, oven-dried at 60° C for 48 h and then litter samples were weighed. For each sample, the percentage of litter mass loss was calculated as M0-Mt/M0*100, where Mt is the final dry mass at decomposition stage t, M0 is the initial dry mass of a sample.


Bernhard, N., Moskwa, L. M., Schmidt, K., Oeser, R. A., Aburto, F., Bader, M. Y., ... & Brucker, E. (2018). Pedogenic and microbial interrelations to regional climate and local topography: New insights from a climate gradient (arid to humid) along the Coastal Cordillera of Chile. Catena, 170, 335-355. doi: 10.1016/j.catena.2018.06.018

Bonanomi, G.; Incerti, G.; Giannino, F.; Mingo, A.; Lanzotti, V.; Mazzoleni, S. (2013). Litter quality assessed by solid state C-13 NMR spectroscopy predicts decay rate better than C/N and Lignin/N ratios. Soil Biology & Biochemistry, 56, 40-48. doi: 10.1016/j.soilbio.2012.03.003

Dias, A. T. C., Cornelissen, J. H., & Berg, M. P. (2017). Litter for life: assessing the multifunctional legacy of plant traits. Journal of Ecology, 105(5), 1163-1168. doi: 10.1111/1365-2745.12763

Ehlers, T. A., von Blanckenburg, F., & Übernickel, K. (2019). EarthShape weather data collection. Retrieved from: Last accessed 2 December 2019.

Garreaud, R. D., Boisier, J. P., Rondanelli, R., Montecinos, A., Sepúlveda, H. H., & Veloso‐Aguila, D. (2020). The Central Chile Mega Drought (2010–2018): A climate dynamics perspective. International Journal of Climatology, 40(1), 421-439. doi: 10.1002/joc.6219

Hagerman A.E. & Butler L.G. (1978) Protein precipitation method for the quantitative determination of tannins. Journal of Agricultural and Food Chemistry, 26(4), 809-812. doi: 10.1021/jf60218a027

INIA. (2019) Red Agrometeorológica del Instituto Nacional de Investigación Agropecuaria, Chile. Retrieved from: Last accessed 11 March 2019.

Jackson, M.L. (1958). Soil chemical analysis. London. Constable & Co. Ltd.

Lehnert, L. W., Thies, B., Trachte, K., Achilles, S., Osses, P., Baumann, K., ... & Karsten, U. (2018). A case study on fog/low stratus occurrence at Las Lomitas, Atacama Desert (Chile) as a water source for biological soil crusts. Aerosol and Air Quality Research, 18(1), 254-269. doi: 10.4209/aaqr.2017.01.0021

Lim, C.H., Jackson, M.L. (1982). Dissolution for total elemental analysis, in: Page, A.L., R.H., Keeney, D.R. (Eds.), Methods of Soil Analysis. Part 2: Chemical and Microbiological Propereties. Madison. American Society of Agronomy, Inc.

Marigo, G. (1973). Sur une méthode de fractionnement et d’estimation des composées phénoliques chez les végéteaux. Analysis, 2, 106-110.

Nelson, P. N., Baldock, J. A. (2005) .Estimating the molecular composition of a diverse range of natural organic materials from solid-state C-13 NMR and elemental analyses. Biogeochemistry, 72, (1), 1-34. doi: 10.1007/s10533-004-0076-3

Oeser, R. A., Stroncik, N., Moskwa, L. M., Bernhard, N., Schaller, M., Canessa, R., ... & Fuentes, J. P. (2018). Chemistry and microbiology of the Critical Zone along a steep climate and vegetation gradient in the Chilean Coastal Cordillera. Catena, 170, 183-203. doi: 10.1016/j.catena.2018.06.002

Thompson, M. V., Palma, B., Knowles, J. T., & Holbrook, N. M. (2003). Multi-annual climate in Parque Nacional Pan de Azúcar, Atacama Desert, Chile. Revista Chilena de Historia Natural, 76(2), 235-254.

Wild, J., Kopecký, M., Macek, M., Šanda, M., Jankovec, J., & Haase, T. (2019). Climate at ecologically relevant scales: A new temperature and soil moisture logger for long-term microclimate measurement. Agricultural and forest meteorology, 268, 40-47. doi: 10.1016/j.agrformet.2018.12.018

Zukswert, J. M., & Prescott, C. E. (2017). Relationships among leaf functional traits, litter traits, and mass loss during early phases of leaf litter decomposition in 12 woody plant species. Oecologia, 185(2), 305-316. doi:  10.1007/s00442-017-3951-z

Usage notes


Species: Species name (generic name_specific epithet)

Origin_site: Study site to which a species belongs and where the litter was collected from. AD= Arid Dry, AF= Arid Fog, SA= Semi-Arid, ME= Mediterranean, TU= Temperate Upland, TL= Temperate Low.

Dest_site: Study site where the litter was decomposed. AD= Arid Dry, AF= Arid Fog, SA= Semi-Arid, ME= Mediterranean, TU= Temperate Upland, TL= Temperate Low.

MAT: Mean annual temperature in °C at the site level, corresponding to the study period (June 2016-May 2018).

AP: Annual precipitation (mm) at the site level, corresponding to the study period (June 2016-May 2018).

Dest_plot: Plot number within a destination site (from 1 to 3).

Plot_MST_period: Mean soil temperature (°C) for a plot, averaged for the corresponding decomposition period.

Plot_MSM_period: Mean soil moisture (m3/m3) for a plot, averaged for the corresponding decomposition period.

Mass_loss_p: Mass loss of a litterbag (%).

Mass_rem: Mass remaining in a litterbag (%).

Time_months: Time (in months) during which a litterbag was decomposed.

Fp: Force to punch (N cm-1) of a species. Data represent the average of 5 individuals.

SLA: Specific leaf area (cm2 g-1) of a species. Data represent the average of 5 individuals.

CN_2016: Carbon to Nitrogen ratio of a species. Data represent the average of subsamples of the litter used for the litterbag experiment for each species.

Carbohydrate: Carbohydrates concentration (%) for a species. Data represent the average of subsamples of the litter used for the litterbag experiment for each species.

Protein: Proteins concentration (%) for a species. Data represent the average of subsamples of the litter used for the litterbag experiment for each species.

Lignin: Lignin concentration (%) for a species. Data represent the average of subsamples of the litter used for the litterbag experiment for each species.

Lipid: Lipids concentration (%) for a species. Data represent the average of subsamples of the litter used for the litterbag experiment for each species.

Lignin_N: Lignin to N ratio for a species (Lignin/N, where both Lignin and N are in %). Data represent the average of subsamples of the litter used for the litterbag experiment for each species.

Al, Ca, Fe, K, Mg, Mn, Na and P: elemental concentrations (mg/g) for a species. Data represent the average of subsamples of the litter used for the litterbag experiment for each species.

Tannins: Tannins concentration (mg/g) for a species. Data represent the average of subsamples of the litter used for the litterbag experiment for each species.

TotPhenols: Total phenolics concentration (mg/g) for a species. Data represent the average of subsamples of the litter used for the litterbag experiment for each species.


Deutsche Forschungsgemeinschaft, Award: 3843/6-1

Deutsche Forschungsgemeinschaft, Award: 338/14-2

Deutsche Forschungsgemeinschaft, Award: 338/14-1

Deutsche Forschungsgemeinschaft, Award: 3021/6-1