Canopy Trimming Experiment Litterfall Nutrients Data



Short name: 


Data set ID: 



Humid tropical forests have the highest rates of litterfall production globally, which fuels rapid nutrient recycling and high net ecosystem production. Severe storm events significantly alter patterns in litterfall mass and nutrient dynamics through a combination of canopy disturbance and litter deposition. In this study, we used a large-scale long-term manipulation experiment to explore the separate and combined effects of canopy trimming and litter deposition on litterfall rates and litter nutrient concentrations and content. The deposition of fine litter associated with the treatments was equivalent to more than two times the annual fine litterfall mass and nutrient content in control plots. Results showed that canopy trimming was the primary driver of changes in litterfall and associated nutrient cycling. Canopy trimming reduced litterfall mass by 14 Mg ha-1 over the 2.5 year post-trim period. Nutrient concentrations increased in some litter fractions following trimming, likely due to a combination of changes in the species and fractional composition of litterfall, and increased nutrient uptake from reduced competition for nutrients. Declines in litterfall mass, however, led to large reductions in litterfall nutrient content with a loss of 143 ± 22 kg N ha-1 and 7 ± 0.2 kg P ha-1 over the 2.5 year post-trim period. There were no significant effects of litter deposition on litterfall rates or nutrient content, contrary to results from some fertilizer experiments. Our results suggest that large pulsed inputs of nutrients associated with tropical storms are unlikely to increase litterfall production, and that canopy disturbance has large and lasting effects on carbon and nutrient cycling.

Date Range: 
2002-05-01 00:00:00 to 2007-11-20 00:00:00

Publication Date: 

2018-04-20 00:00:00





We used a complete randomized block design with three replicate blocks each containing four 30 x 30 m treatment plots separated by approximately 20 m buffers. Within each plot, a 20 x 20 m sampling area was defined and furthered divided into 16 subplots to minimize the effects of destructive sampling on long-term measurements. Treatments consisted of: (1) canopy trimming and litter deposition (trim + debris), (2) canopy trimming with the litter removed (trim + no debris), (3) intact canopy with litter added from the removal treatment (no trim + debris) and (4) no manipulation (no trim + no debris). Pre-treatment litterfall measurements began in November 2002. The manipulations spanned from late October 2004 to June 2005. Each treatment was completed within a given plot and block before the subsequent block was treated. Details of the treatments, plots, and timing are given in Shiels et al. (2010) and Richardson et al. (2010). Trimmed material was weighed using tarps and spring balances. Canopy trimming generated approximately 72 ± 2 Mg ha-1 of necromass. The necromass was not immediately distributed on the plots resulting in some loss of mass (Richardson et al., 2010; Shiels et al., 2010), and associated nutrient changes (Shiels and González, 2014). To determine the nutrient deposition from fine litter generated from trimming, we multiplied the dry mass (1.6 Mg plot-1) of leaves (which were pooled with fine twigs) by the mean annual nutrient concentrations in leaf litterfall in the same forest, using data generated from the unmanipulated plots (no trim + no debris) (Table 1). We used litterfall concentrations as opposed to values for fresh plant fractions because of potential nutrient loss prior to placement on the experimental plots (Shiels and González, 2014). This resulted in an estimated minimum nutrient deposition rate of 164 kg N ha-1, 5 kg P ha-1, 34 kg K ha-1, 157 kg Ca ha-1, and 40 kg Mg ha-1. These are minimum values because additional nutrients were added in other litter fractions (fine and coarse wood, fruits and flowers, and miscellaneous material) that were not quantified during the trimming events but were deposited on the plots.  Litterfall was collected every 14 days from 10 baskets (dimensions 43 x 43 cm) distributed in a stratified random fashion (to ensure plot coverage) inside each 20 x 20 m core area of each treatment plot. Baskets were leveled and fastened to poles at 1 m height. Baskets were removed during the trimming of respective treatment plots in order to prevent the baskets from getting broken by dropped branches, and replaced as soon as the canopy trimming was completed. We separately report the pretreatment (November 2002–October 2004) and post-treatment (July 2005–December 2007) data, recognizing that this excludes a small amount of data during the establishment of the experiment. Litterfall mass associated with the trimming events was estimated by weighing the litter generated during the canopy manipulation using tarps and spring balances. Following each litter collection, litter was dried at 40 oC for at least one week, and kept in a heated room until samples could be sorted into the following categories: leaves, wood, fruits and flowers (including seeds), miscellaneous (unidentifiable material >2 mm). We re-dried subsamples of litterfall at 65 oC and weighed them to establish a conversion to oven dry weight. Litterfall was pooled by fraction within each plot quarterly for chemical analyses. In this paper we report quarterly mass to compare with nutrient concentrations and nutrient content.  Litterfall samples were ground to pass through an 18 mesh sieve. Total C and N were determined using the macro dry combustion method on a LECO TruSpec CN Analyzer or LECO CNS-2000 Analyzer. The LECO CNS-2000 Analyzer was used to determine total C and N of litterfall in 2002–2004. The remaining total C and N analyses were determined utilizing the LECO TruSpec CN Analyzer. Blanks and reference materials were analyzed with each run at a rate of 1 per 10–20 samples to insure that the samples were directly comparable.  The ground litterfall samples were digested using a modification of the method recommended by Chao-Yong and Schulte (1985). This wet oxidation uses concentrated HNO3, 30% H2O2 and concentrated HCl and was achieved using a digestion block with automatic temperature control. The digests from 2002 to 2005 were analyzed on a Spectro Ciros ICP Emission Spectrometer, and those from 2006 to 2008 were analyzed in a Spectro Spectro-Blue ICP Emisson Spectrometer, all for Ca, K, P, Mg, Fe, Al, and Mn. The results are reported as mg g_1 on a dry basis at 105 _C. Blanks and National Institute of Standards certified reference material was analyzed with every run for quality assurance and quality control. The moisture factor correction at 105 _C was determined by the LECO Thermogravimetric Analyzer, model TGA 701 and applied

to all reported values. All laboratory procedures were conducted at the International Institute of Tropical Forestry in Puerto Rico. Carbon:N and N:P ratios were calculated using mass weighted values.



Time Period: