Shrub-Flammability

Project goals

• To test relative importance of crown and leaf traits for shrub flammability.
• To determine how well do shoot-level flammability measurements correspond to the flammability category rankings developed by the fire hazard categories listed in the National Fire Protection Association’s Firewise program?

Methods

We have collected 17 shrub species from multiple counties in Texas (species description is in species_table_2022.csv). For each sample for each species, we have collected alike, a paired 70 cm branch from the same plant. The collected branches were stored in big coolers and brought them back to Schwilk Lab. One of the branch get burned to measure the flammability traits and other branch left unburned. We have also collected a sub-sample of leaf and twig with leaves from the same plant which was used to measure the leaf traits and field moisture content.

• Burned sample : Each burned sample went through 36 hours of bench drying inside Schwilk Lab at Texas Tech before burning. The temperature and humidity of the drying period was recorded by the hobo data logger. We measured the canopy volume during the bench drying period and burned them after 36 hours of drying. We also measured the canopy moisture content and leaf moisture content of the burned sample by separating some twigs with leaves and leaves right before burning.

• Unburned sample : We measured the leaf stem mass ratio by separating the leaf (including the rachis and petiole) from the stem from the unburned sample.

• Sub-sample : We measured the leaf mass per area, leaf area per leaflet, leaf length per leaflet and field moisture content from the sub-samples which was collected from the same plants from where the paired branch collected.

Plant traits and their descriptions

Canopy Traits

• Canopy density : The canopy density of the burned sample measured during the bench drying period of the burned branch. We assumed that the 70 cm branch would have a particular geometric shape among three possible shape : cylinder, truncated cone and the combination of two truncated cone. For those samples which have a cylinder shape, we measured the diameter of the branch at three different position (top, bottom, middle). Then we averaged the diameter and calculated the volume according to the formula of the volume of a cylinder. For those samples which have truncated cone shape, we measured the diameter at the top and at the bottom. For those samples which have the shape of the combination of two truncated cone, we measured the maximum diameter, distance of the maximum diameter from the bottom and the diameter at the bottom and at the top portion of the branch. Then we calculated the volume of the truncated cone according to the formula of a truncated cone. For all the types of shape, we calculated the canopy density by dividing the total mass of the burning sample in dry basis by the canopy volume.

• Leaf stem mass ratio : We have collected two 70 cm branch from the same plant for each sample. We made sure that the two branch look alike and assumed that those two branch have almost similar amount of leaf stem mass ratio. One of the branch was used to measure the leaf stem mass ratio. After getting the samples back in the lab from field, we separated the leaf (rachis and petiole included) from the stem and took their fresh weight separately. Both the leaf and stem went through 48 hours of drying period at 65 degree celsius and took their dry weight and calculated the leaf stem mass ratio in dry weight basis.

• Total weight of the burning sample : We took the fresh weight of each burning sample right before burning. Then we converted the fresh weight into dry weight by multiplying the ratio of the dry weight and fresh weight of the twig with few leaves which was separated from the burning sample right before burning to measure the canopy moisture content. After that we added the dry weight of the twig with few leaves and the leaves (which was separated from the burning sample right before burning to calculate the leaf moisture content) with the dry weight of the burning sample to get the total weight of the burning sample in dry basis.

• Canopy moisture content : A twig with few leaves were separated from the burning samples right before burning. After taking the fresh weight, it went through 48 hours of drying inside oven at 65 degree celsius and then calculated the moisture content as percentage in dry basis.

Leaf Traits

• Leaf mass per area : We have used five leaves which was separated from the sub-samples to measure the LMA. For most of the species we have used five leaves and some of them have less than five or more than five. We took the fresh weight of the leaves and then they went through 48 hours of drying inside oven at 65 degree celsius. The leaf area of the leaves were measured in Licor 3100 leaf area meter. We used the dry weight of the leaves in calculating the leaf mass per area. Three species from Juniperus group, we calculated their leaf area manually. We took 10 branchlets for each samples. We assumed that the shape of the branchlet is a cylinder. We measured the diameter and length of each branchlets by slide calipers and meter scale respectively and calculated their area based on the formula of the surface area of a cylinder. After taking the fresh weight of the 10 branchlets , they put inside the oven for 48 hours at 65 degree celsius.

• Leaf area per leaflet : The total leaf area of all the leaves used to measured the LMA, divided by the number of leaflet to get the leaf area per leaflet.

• Leaf length per leaflet : For most of the samples, we have used five leaves to measure the leaf length of each leaflet (including the petiole). We used the meter scale to measure the length of the leaflet and averaged the total length of all the leaflet to get the leaf length per leaflet.

• Leaf moisture content : Few leaves were separated from the burning samples right before burning. Those leaves were went through 48 hours of drying inside oven at 65 degree celsius. The moisture content was calculated as percentage in dry basis.

Apart from all the canopy traits and leaf traits, we also measured the field moisture content. We separated some twig with leaves from the sub-sample and took their fresh weight. After 48 hours of drying inside oven, we took the dry weight and calculated the field moisture content as percentage in dry basis.

Flammability traits and their descriptions

After 36 hours of drying period, the burning samples put on grill for two minutes for pre-heating. Right before the pre-heating period, the weather data (Temperature, humidity and wind speed) were recorded with a Kestrel 3000 pocket weather meter. After two minutes of pre-heating ,an ignition blow torch turned on until a sample got ignited.

• Flame duration : The duration of the flame sustained. The duration between the time when the ignition started and when the flame stopped.

• Flame height : Measured by a meter scale attached with the barbecue apparatus. An assistance continuously kept eye on the flame and estimated the maximum flame height of each burning sample.

• Mass consumed : Calculated by the difference of the mass of burning samples before and after burning.

• Volume burned : The percentage of volume of the burning samples estimated visually. After each trial, at least two person together estimated the percentage of the voulme of the burning samples to reduce the error.

• Ignition delay : The time taken by each samples after two minutes pre-heating to get ignited.

• Self ignition : If a sample get ignited during pre-heating, the sample scored 1 for self ignition, else 0. If a sample had self ignition and the flame sustained after two minutes pre-heating , the sample got 0 as ignition delay. Some of the samples had self ignition during pre-heating and the flame ceased after few seconds, they ignited by the blow torch. Therefore they scored 1 as self ignition and had their ignition delay period as well since they are ignited by the blow torch.

• Maximum temperature : When a sample get ignited, the maximum temperature of the disc near the flame recorded by the temperature gun. Most of the sample doesn't have their maximum temperature. However, we have also three thermo-couples data loggers at three position (both sides and at the middle) which were recording the temperature of the entire trial at one second interval. We calculated the peak temperature from the data loggers.

We have also calculated the total duration of the trial where the temperature were greater than 100 degree celsius. The sum of the total temperature where the trial temperature were greater than 100 degree celsius from the thermo-couples data loggers.

Data

Each data file in data in shrubflam repo has as associated machine readable metadata file. If a data file named canopy_measurements_2022.csv, then it has an associated metadata file as canopy_measurements_metadata.csv. The metadata file describes each variables information: name of the variables, the data type, the unit and a description how the variables measured.

Data analysis and scripts

All the data analyses were performed in R with all scripts in scripts/. To reproduce the results, please read run-all.R script. The task need to perform before running each script is documented at the beginning of each script.

Hobo data loggers, aluminum discs and gas flow

The gas flow from the Blue Rhino gas cylinder was 20.35 gram per minute. The position and distance of the data loggers:

• Height : The height of the data loggers from the grill was 20 cm for each data loggers. The height of the aluminium discs was also 20 cm from the grill to the middle of the disks.

• Position : Both the left and right hobo data loggers is 18.5 cm from the both sides and the mid hobo data loggers approximately 44 cm away from both sides. The distance of the disk from both sides is approximately 30 cm from both sides.

Both the position and height of the data loggers and discs displaced time to time during some trials due to the massiveness of some samples.

List of missing data

• Bench drying hobo data : We have bench drying hobo data for 5 trials date. The 05_19_2022.csv and 05_20_2022.csv is the weather data of the lab from the day after the first trial.

• Wind speed data : We have three missing wind speed data of 129 trials which is the number of trials after removing the species which has less than three samples and two samples, one of them has missing flammability data and another one is biased.

• Air temperature : We have 37 missing data for the air temperature of the trials from Kestrel 3000 pocket weather meter. The air temperature of the first trial date is from the hourly average temperature from https://weatherspark.com/h/d/4735/2022/5/20/Historical-Weather-on-Friday-May-20-2022-in-Lubbock-Texas-United-States\#Figures-Temperature.

• Humidity : We have only few humidity data for few trials from Kestrel 3000 pocket weather meter. Almost entirely missing for all the trials.

However, we have temperature and humidity for 89 trials from hobo data loggers.