Data and code from: Good days, bad days: Weather patterns driving bryophyte carbon-exchange dynamics in tropical rainforest canopies

Published Jul 24, 2026 on Dryad. https://doi.org/10.5061/dryad.2fqz6130f

Abstract

These files were used to test the unfavourable timing hypothesis, that attributes the limited biomass accumulation of tropical lowland bryophytes to limiting photosynthetic and respiratory activity on many days in tropical lowland areas. The CSV files serve as inputs to run the PoiCarb (Version 2.0) on three epiphytic canopy species. The RMD scripts perform different tasks: i.) deriving species-specific physiological profiles and the corresponding model parameters to run PoiCarb 1.1, ii.) implementing PoiCarb 1.1, and iii.) visualizing the inputs and outputs of this model, included in the manuscript "Good days, bad days: weather as a driver of bryophyte carbon-exchange dynamics in tropical rainforest canopies"; to facilitate replicability, RData files are also provided. The HTML document contains the fine-resolution output of the model PoiCarb (Version 1.1), used to simulate bryophyte water content and carbon exchange under different weather patterns occurring in a tropical lowland forest with a temporal resolution of 10 minutes for 290 days.

This dataset contains four groups of files: INPUT, MODEL, OUTPUT, and Visualization.

Here, the user can find the description and, if applicable, indications about each group content.

Metadata for each group are presented in a separate file, METADATA.txt.

Description of data and files

Input

Input files include HTML, RData, RMD, and CSV files.

Inputs.HTML displays the required input files and shows the code to process them so a suitable input (Inputs.RData) for PoiCarb 1.1 is created.

The user can run the code Inputs.Rmd with the provided input tables (Input_*.CSV files) or with their own data of physiological observations and climate records. It is recommended to run this code as an R project, after saving both the code and the input file in the same directory.

See INPUT in METADATA.txt.

Inputs.RData is provided so the user can run the model even without having to run Input_RMD. Here, a description of the objects contained in the Inputs.RData file, in the order they appear in the PoiCarb1.1.RMD:

install.load.package: R user function to by Erik Kusch ([https://www.erikkusch.com/post/krigr-mats/krigrworkshop/ (https://www.erikkusch.com/post/krigr-mats/krigrworkshop/)) used to install and/or load required packages to run this RMD script.

WCD_pars: R list containing a matrix with the parameters of the modeled desiccation curve for the target species, where the model parameters (maximum water content, WCmax, and the desiccation coefficient, k) are in the rows and their statistical estimate (Estimate), standard error of the estimate (Std. Error), statistics (t value) and significance (Pr(>|t|) are in the columns. This desiccation curve is modeled from the physiological data included in Input_ds.csv (loaded as the ds_d and ds_d_Nu objects, with data for all three species, Neckeropsis undulata, Plagiochila raddiana, and Symbiezidium transversale, and only for N. undulata)

As it is explicit in PoiCarb1.1.RMD, the R objects WCmax, WCm, k_start, and k are derived from WCD_pars and correspond to the target species maximum water content, the initial water content forced to maximum water content, the initial desiccation coefficient forced to that of the target species, and the desiccation coefficient of the target species.

Crvs: An R list containing the input data in four data.tables: Assimilation, Dessication_Speed, Water_Content_Light, and Water_content_Dark, which correspond to the files Input_light.csv, Input_ds.csv, Input_Assimilation_WaterContent.csv, and, Input_Respiration_WaterContent.csv respectivelly.

PAR_fun: R user function to calculate bryophyte Assimilation from the instant light input (PAR) and from the light response curve parameters (a, b, c). This function corresponds to Equation 1 in Nikolić et al. 2024.

A_models: R list containing a nonlinear regression model (nls R object) form and its coefficients (a,b, and c). This nonlinear regression model, in turn, is derived from the physiological data included in Input_light.csv(loaded as the l_d and l_d_Nu objects, with data for all three species, N. undulata, P. raddiana, and S. transversale, and only for N. undulata).

climate_input_data: As it is explicit in PoiCarb1.1.RMD, is an R object derived from the imported Input_Canopy_Climate.csv into R (object climate_input) containing the climate variables relevant for running PoiCarb1.1, namelly: Date (yyyy-mm-dd), time (date and time in the format yyyy-mm-dd hh:mm:ss), Hour (in 24 h format), average photosynthetic active radiation density (PAR, in micro mol/(s m²), relative humidity (RH), average air temperature (Te in °C), total precipitation (rain.mm in mm), total minutes with wet leaves (LWMWet_Tot), and vapor pressure deficit (VPD in kPa).

A_pars: R list containing a matrix with the parameters of the light response curve for the target species, where the model parameters (Maximum assimilation rate, a, apparent quantum yield, b, and Light compensation point, c) are in the rows and their statistical estimate (Estimate), standard error of the estimate (Std. Error), statistics (t value) and significance (Pr(>|t|) are in the columns. This light response curve is modeled from the physiological data included in Input_light.csv(loaded as the I_d and l_d_Nu objects, with data for all three species, N. undulata, P. raddiana, and S. transversale, and only for N. undulata).

WCR_pars: R list containing a matrix with the parameters of the theoretical rehydration curve for the target species, where the curve parameters (maximum water content, inflexion, and time) are in the rows and its theoretical estimate in the column. This rehydration curve is theoretically derived from the physiological data included in Input_ds.csv.

wc_fun: R user function to calculate bryophyte assimilation from the thallus water content (WC) and from the water response curve parameters (a, b, c). This function corresponds to Equation 2 in Nikolić et al. 2024.

AWC_models: R list containing a nonlinear regression model (nls R object) form and its coefficients (a,b, and c). This nonlinear regression model, in turn, is derived from the physiological data included in Input_Assimilation_WaterContent.csv(loaded as the wc_d_l and wc_d_l_Nu objects, with data for all three species, N. undulata, P. raddiana, and S. transversale, and only for N. undulata).

AWC_pars: R list containing a matrix with the parameters of the modeled water response curve for the target species, where the model parameters (optimum water content, a, water content coefficient, b, and maximum carbon assimilation, c) are in the rows and their statistical estimate (Estimate), standard error of the estimate (Std. Error), statistics (t value) and significance (Pr(>|t|) are in the columns. This desiccation curve is modeled from the physiological data included in Input_Assimilation_WaterContent.csv(loaded as the wc_d_l and wc_d_l_Nu objects, with data for all three species, N. undulata, P. raddiana, and S. transversale, and only for N. undulata).

wc_dr_fun: R user function to calculate bryophyte respiration from the thallus water content (WC) and from the water response curve parameters.

RWC_models: R list containing a nonlinear regression model (nls R object) form and its coefficients (a,b, and c). This nonlinear regression model in turn, is derived from the physiological data included in Input_Respiration_WaterContent.csv (loaded as the wc_d_d and wc_d_d_Nu objects, with data for all three species, N. undulata, P. raddiana, and S. transversale, and only for N. undulata).

RWC_pars: R list containing a matrix with the parameters of the modeled water response curve in dark conditions (maximum respiration rate, a, respiration coefficient, b, and respiration point,c). This water response curve in dark conditions model in turn, is derived from the physiological data included in Input_Respiration_WaterContent.csv(loaded as the wc_d_d and wc_d_d_Nu objects, with data for all three species, N. undulata, P. raddiana, and S. transversale, and only for N. undulata).

WC_models: R list containing a nonlinear regression model (nls R object) form and its coefficients. This desiccation curve is modeled from the physiological data included in Input_ds.csv (loaded as the ds_d and ds_d_Nu objects, with data for all three species, Neckeropsis undulata, Plagiochila raddiana, and Symbiezidium transversale, and only for N. undulata)

DeH_Fun: R user function to calculate bryophyte desiccation from the desiccation curve parameters.

ReH_Fun R user function to calculate bryophyte rehydration using a Michaelis-Menten equation (Morales-Sanchez et al. 2022).

Model

The code implementing the model is presented in the file PoiCarb1.1.RMD

The user can run this code with the provided Inputs.RDATA file as well as with their own derived Input.RDATA. It is recommended to run this code as an R.Project after saving both the code and the input file in the same directory.

If the user uncomments the code line 148, the output will contain a test with 140 time steps. Inputs.RDATA contains a total of 41921 time steps (~293 days with a 10-minute time resolution) so PoiCarb1.1. takes ~20 minutes to generate an output.

Output

Three RDATA files are provided per each target bryophyte species.

For Neckeropsis undulata:
Output_Nu_PoiCarb1.1_Ha.csv
Output_Nu_PoiCarb1.1_La.csv
Output_Nu_PoiCarb1.1_Ma.csv

For Plagiochila raddiana:
Output_Pr_PoiCarb1.1_Ha.csv
Output_Pr_PoiCarb1.1_La.csv
Output_Pr_PoiCarb1.1_Ma.csv

For Symbiezidium transversale:
Output_St_PoiCarb1.1_Ha.csv
Output_St_PoiCarb1.1_La.csv
Output_St_PoiCarb1.1_Ma.csv

Each Output corresponds to each of the alternative assimilation lines derived for each species (see PoiCarb1.1.RMD and Berdugo et al. submitted)

Visualization

A Dashboard (HTML file) presents the detailed diel figures of the three outputs generated for Neckeropsis undulata for the modeled 290-day period

REFERENCES

Berdugo M B, N Nikolić, L Guérot, S Brach, J Bendix,G Zotz, P Porada, M Bader. Good days, bad days: weather as a driver of bryophyte carbon-exchange dynamics in tropical rainforest canopies

Nikolić, N., Zotz, G., & Bader MY. 2024. Modelling the carbon balance in bryophytes and lichens: Presentation of PoiCarb 1.0, a new model for explaining distribution patterns and predicting climate‐change effects. American Journal of Botany e16266: 1-21** https://doi.org/10.1002/ajb2.16266

Morales-Sánchez, JÁ, Mark, K, Souza, JPS, & Niinemets, Ü. 2022. Desiccation–rehydration measurements in bryophytes: current status and future insights. Journal of Experimental Botany 73: 4338-4361. https://doi.org/10.1093/jxb/erac172

We collected fresh plant material, dense green clusters of canopy bryophytes, in eight canopy trees, including the one instrumented with the canopy climate station. These samples, found in the inner crown, were transported to the lab and processed within five days of the collection date. Sample processing included identifying species composition and carefully separating colonies per species. To identify species, we prepared vouchers and deposited them in the herbarium of the Pontificia Universidad Católica del Ecuador – QCA (Quito) – and in the cryptogamy collection at the Herbarium of the Muséum National d'Histoire Naturelle – PC (Paris). There, mosses were identified using Churchill (1994), Buck (2003), and Florschütz-de Waard (2011), and liverworts using Gradstein & Ilkiu -Borges (2009) and Gradstein(2021). Eleven species composed the bryophyte samples: seven liverworts, Bazzania hookeri (Lindenb.) Trevis, Ceratolejeunea cornuta (Lindenb.) Steph., Plagiochila disticha (Lehm. & Lindenb.) Lehm. & Lindenb., P. raddiana, an unidentified Plagiochila species, Symbiezidium barbiflorum (Lindenb. & Gottsche) A. Evans, and S. transversale; and four mosses, Groutiella apiculate (Hook.) H.A. Crum & Steere, Neckeropsis undulata, Sematophyllum cf. subsimplex (Hedw.) Mitt., and Zelometeorium patulum (Hedw.) Manuel.

To describe how carbon assimilation varies in response to light intensity (light-response curves) and how water content influences carbon exchange under both light (water-content response curve of net assimilation) and dark (water-content response curve of dark respiration) conditions, we measured CO₂ exchange of monospecific layers of green shoots in a 3.5-cm-diameter cuvette with a portable gas exchange system (GFS-3000; Heinz Walz GmbH, Effeltrich, Germany). Measurements were performed under constant CO₂ concentration (400 ppm) and temperature (26°C) within the cuvette; relative humidity entering the cuvette was set at 80%, but the relative humidity within the cuvette is not controlled by the gas exchange system (Walz 2005). Light intensity (photosynthetically active radiation) was constant at 300 mmol m^-2 s^-1 for water-content response curves and was increased step-wise for light response curves (0, 10, 25, 50, 100, 150, 200, 250, 300, 400, 500, 600, 800 and 1000 mmol m^-2 s^-1, based on the protocol of Wang & Bader, 2018). For water-content response curves, samples were weighed after each measurement to determine the fresh weight, which was then used to calculate the water content after the experiment as water content (WC) = (fresh weight – dry weight) / dry weight. In addition, the drying speed of the samples was determined by weighing them hourly, leaving them in outdoor conditions near the YSS lab building in between measurements. The starting point was at the maximum water-holding capacity (WHC), determined after fully wetting and lightly shaking the samples.

Although we processed the 11 collected species. However, the analysis targets N. undulata, the dominant bryophyte in this system (Berdugo et al. 2022), P. raddiana, and S. transversale, two co-occurring species that exemplify the range of maximum assimilation rate to dark respiration physiological observed in tropical bryophytes (Wagner et al. 2014). For each set of measurements of these three species, we derived species-specific response curves or drying curves by fitting the observations to predefined nonlinear model types (Nikolic et al 2023) and improving the adjustment of the parameter least-squares estimates with a bootstrapping approach. These regression models provided us with species-specific parameter estimates and their uncertainty (standard error) to simulate bryophyte carbon exchange and desiccation rate.