Appendix 1. Measuring HCN gas - Methods detail.

Using small plastic cup to concentrate HCN gas:
This procedure was developed to measure relatively low concentrations of HCN gas produced by crushing Passiflora foliage. Measurements are repeated every five minutes until the concentration of gas falls below the sensitivity of the meter; the resulting data describes the rate of HCN release as well as the total quantity released. In general, the total HCN values should be a good indicator of the amount of cyanogenic glycoside in the plant tissues as long as there is sufficient β-glucosidase enzyme to cause hydrolysis. Reducing the amount or activity of β-glucosidase enzyme should slow the rate of release but should not affect the final total quantity. Procedure 1described below is for measuring HCN in three different leaves along a branch; modify for other materials or as needed. Comparison with Procedure 2 below reveals that about half the HCN gas is lost during Procedure 1, perhaps by diffusion through the plastic cup and tubing as well as during cup evacuation every five minutes (see below). Despite this lack of accuracy, the technique is sufficiently precise and repeatable to create informative charts and comparisons (see Appendix 2. Gallery of HCN profiles for different Passiflora species).

"Gasalert Extreme" HCN meter, manufactired by BW Technologies/Honeywell. Set meter in "stealth" mode. Open meter & remove vibration device. These devices slowly lose sensitivity over time and use. Meters were replaced according to the following schedule: unit 42 purchased March 2011 and retired February 2014; unit 95 purchased January 2012 and retired September 2014; unit 60 purchased June 2015 and retired December 2015. Differences between units and decay of sensitivity introduced some error to the measurements, but again, the losses were small compared with the variation being measured.

"The Sampler" gas pump, manufactured by BW Technologies; includes1/4 inch flexible clear tubing and connector device attaching to meter (see photo below).

AA batteries
3 snap cap airtight plastic cups, about 143cc volume with lid on, made by "Glad"
1/4 inch diameter drill bit
flame to heat drill bit and pliers to hold hot bit
milligram balance
small mortar and pestle
notebook, watch/clock and pencil

using hot drill bit, melt hole in side of glad cups (3)
cover hole with small tab of transparent tape
label cups and corresponding lids 1-3
measure tare weight of cup, lid, and tape combination
create data form in notebook (see below)

Sample Preparation:
collect three leaves from Passiflora branch and bring to lab
start "Sampler" pump and connect pump to HCN meter. Make sure both are turned on.
cut off 1-10 cm^2 of leaf material and place by cups 1-3
record date, species, location of sample, leaf age or #, cup #, and tare weight of cup into notebook for each cup

Sample processing:
record time zero, and immediately begin processing the three samples.
Sample 1:Fold leaves into as small a bundle as possible, and hold in bottom of mortar with fingertip. Quickly crush with pestle.
Remove crushed leaves quickly (within 10-15 seconds) and put in cup. Snap on lid. Immediately go to next sample.
Repeat samples 2 and 3.
Now you have three samples in sealed cups, outgassing HCN into the air above the sample confined within the cup.
Weigh the three cups and record the gross weight of each.
Turn on "sampler" pump and leave running.

Sample measurement:
At 5 minutes past time zero (t1), peel off tape from cup #1 and insert 1/4" tube from "sampler" pump.
Watch meter. The readings start low and build over 10-20 seconds, reaching a peak value. Quickly record the peak value on the form next to to the time t1 (time zero + 5 minutes).
Quickly remove the tube, replace the tape over the cup hole, open the cup lid and air out the cup for 3 seconds. Replace the lid.
Repeat with the next cup (cup#2) at 6 minutes past time zero, and with cup#3 at 7 minutes past time zero. The one minute interval is usually enough time for the meter to return to zero; if not, add extra seconds to the interval until zero or small values (e.g..3-.5 ppm) are reached. The following readings can then be shortened accordingly, coing back to the original schedule.

Repeat 3 sample measurements starting at 10 minutes past time zero (t2), 15 minutes, (t3) and at 5 minute intervals until the ppm readings drop below 0.5 ppm.
Terminate measurements when the ppm readings drop below 0.5 ppm.
Enter data into spreadsheet. To calculate μM HCN /g plant tissue, multiply the sum total ppm for that sample (ppm1 + ppm2 + ppm3 + . . . + ppmn), times the volume of the sample chamber (for Glad cups this is 0.143 l), divided by the volume of one mole of gas (22.4l). This value is then divided by the sample weight in grams to give μM/g HCN.

Lab notebook column headings: date, sample ID info, leaf age, cup#, tare wt. (g), t0, gross wt. (g), net wt. (g), t1, ppm1, t2, ppm2, t3, ppm3,t4, ppm4 t5, ppm

Spreadsheet column headings include all the lab notebook headings as well as sum of all ppm values and the μM/g HCN value.

(above) HCN gas measurement apparatus. Shown: plastic cup concentrator with hose to "sampler" gas pump. Sampler connects to HCN meter by a special connector supplied with the unit.





(right, above) using closed glass container for measuring HCN gas

(right, below) measuring HCN gas by active, feeding H. sara larvae


Measuring HCN gas in a closed glass bell jar. An alternative method is to put the HCN meter and sample pump in a sealed glass container with crushed sample and wait until the ppm readings stabilize. Because of the large volume of the chamber required to enclose the meter(in this case a 9.6 liter glass bell jar) this method was not very sensitive to low concentrations or small samples. However, in cases where samples were large enough, this method was useful for monitoring the time course of HCN release on a finer scale (every 20 seconds) than the 5-minute intervals in procedure 1. This procedure also has the advantage of improved accuracy, as no gas is lost during the measurement process. Comparison of the two procedures suggest that about 50% of the gas escapes detection during Procedure 1, and that values should be approximately doubled to yield correct quantities. I have not done this correction, however. The numbers presented in all graphs and tables are the original measurements for each procedure. The data in Appendix 2. Gallery of HCN profiles for different Passiflora species graphically illustrates the precision of these procedures.