Barry Rice

- doing science on a terrestrial planet -

Are Genlisea traps active? A Crude Calculation
(Carnivorous Plant Newsletter, 1994, 23:2, 40.)

I find the structure of the Genlisea trap very interesting, although they are not well understood in the botanical literature. Many questions remain unanswered. Are prey attracted or do they just wander into the traps? How is the prey retained in the utricle (the trap's digestive chamber) and how is it digested? How are the digested nutrients retained and then assimilated? With more time and research, these questions will be answered.

The form of Genlisea traps is well known, and I described some of its features in the previous article. In this article I concentrate on a single aspect of the Genlisea trap, an aspect that would seem to be a flaw in the trap's construction. An observation has been made by Juniper, Robins, and Joel in the book, The Carnivorous Plants (hereafter JRJ), which may point to the plant's solution to this flaw--namely that the trap is actually active. To complete my conjectures, I present an approximate calculation exploring whether this is plausible.

It seems CP are fairly efficient digestion mechanisms. Dionaea traps allow prey to escape if the prey are too small to be worth digesting. JRJ report that Drosera erythrorhiza absorb a full 76% of the available nitrogen in insect prey. Yet consider the fate of a rotifer (to choose a likely nutrient source) swimming along the interior of a Genlisea trap. Because of strategically located trap hairs it can only swim towards the utricle where it dies and is broken down for absorption. But what of the chemicals released by the dissolving rotifer, before they are absorbed by the plant? There are no one-way valves at the entrance of the utricle (as there are in Utricularia bladders), and inward pointing hairs have no effect on individual molecules. So what prevents a significant portion of the valuable nutrients from diffusing through the utricle entrance, out of the trap, and away from the plant?

How does Genlisea prevent a wasteful loss of nutrients from the trap? Or does it simply operate inefficiently? JRJ make an observation which may be important (pg 126). They note that utricles contain not only the digested carcasses of prey, but also particles of dirt. The traps of Genlisea hang downward, so it is difficult to explain how sinking or drifting dirt particles could find their ways into the utricle. After settling into the spiral trap entrance, the particles would need to inexplicably rise into the trap mouth, through the trap tube, and into the utricle. Instead of that unlikely scenario, is it possible these bits of detritus have been sucked into the trap by the plant's effort? Perhaps the plant is expelling water from the trap through the utricle walls. New water from outside the trap would flow up the trap tube to replace the water removed from the utricle. The expulsion would be comparable to the phase in which water is removed from the interior of a sprung Utricularia bladder and is excreted into its surroundings. This is not too implausible since the two genera are closely related and the traps of both genera contain similar internal and external glands. The purpose of this expulsion might be to suck valuable nutrients into the cell walls, and thus prevent their escape from the trap. Genlisea traps may be active and not passive.

I decided to make a few simple calculations to see if it is even wildly possible that a Genlisea trap could function as a pump. Could it remove water from its utricle at a rate sufficient to overcome the molecular speed of nutrients diffusing down the trap tube to the trap bifurcation, and then into open water? Being a scientist, I know that approximate calculations provide insight to basic phenomena. You can get a rough idea of what is going on, or if a mechanism is possible--then let the next group of researchers worry about the details! To treat this problem I needed to calculate two velocities. First, what is the velocity of liquid being sucked through the trap tube to the utricle? Second, what is a typical velocity at which nutrient molecules diffuse out of the trap? If the velocity of fluid up the tube (Vf) is greater than a molecule's diffusion speed (Vd) then the plant could overcome diffusion and thus maximize its efficiency. If you find math uninteresting or paralyzing, skip the next three paragraphs and read the one starting with "I don't expect you...." for the results.

First I estimated the flow velocity through the tube. JRJ note work by various researchers who measured that Utricularia bladders expel about 40% of their fluid volume in approximately 20 minutes. Assuming a spherical bladder 1 mm in diameter, this corresponds to 1.74×10-7 cm3 sec-1 of water pumped through its surface area. Some research suggests the glands scattered over the entire exterior surface of the bladders are responsible for removing the internal bladder fluid. Since similar glands are found on the exterior of the Genlisea utricle, it is plausible they remove water from the trap in the same way. Modelling a typical large African Genlisea utricle as a sphere 4 mm in diameter, it would have sixteen times the surface area of the Utricularia bladder and could pump water sixteen times faster. As this water is sucked through the narrow trap tube, which has an inner diameter of about 0.05 cm, it would produce a flow velocity of Vf= 0.0014 cm sec-1.

And what is the diffusion speed of nutrient molecules through water? This is a little more complicated. A molecule of mass m and at temperature T (in Kelvins) will have a molecular velocity Vm approximately given by m(Vm)2 = 2kT, where k is Boltzmann's constant. For a typical nutrient like the phosphate ion (PO4)-3 at T=25°C, Vm=2.3×104 cm sec-1. As this ion races among the water molecules, it will travel only a short distance L before colliding with one. This distance is called the mean free path. (The mean free path can be estimated using L3=m/p, where m and p are the molecular mass and density of H2O.) The time for a particle to traverse a mean free path is given by t=L/Vm. Because of all these molecular collisions, the ion will not travel in a straight line. Instead it will randomly wander around. It can be shown that after n molecular collisions, the ion will have wandered a distance X from its starting point, where X2=nL2. For it to wander about 1.5 cm (the length of the trap tube for a large Genlisea) the ion will suffer 2.3×1015 collisions! To wander this distance will take the phosphate ion a total amount of time equal to nt, so I can write the effective diffusion velocity as:

Vd = n1/2L/(nt) = L/(n1/2t) = Vm/n1/2.

For our nutrient ion, this gives a diffusion velocity of Vd=0.00048 cm/s.

My velocity calculations were admittedly crude and did not consider a wealth of interesting details. But unless I made a fatal blunder and neglected an important effect, the details that would make these calculations many times more difficult are unlikely to change the results too much. I note for example that I did not treat the effects of intermolecular forces at all. But these forces would only conspire to decrease diffusion velocities, and therefore make the trap even more effective. I think the strongest criticism against my argument is that the methods of water excretion in both Utricularia and Genlisea traps are not understood. In spite of its greater size a Genlisea trap might pump fluid only at the same rate as a Utricularia trap. But still the flow and diffusion velocities would be roughly comparable and the pumping mechanism would be useful for the plant. After all, diffusion is a random process and the diffusion velocity I calculated is only a typical value for a molecule--there will always be faster and slower particles. So the precise value of Vd is not important. For this reason, I am not too concerned with my choice of a phosphate ion as the test particle--Vd is modified only by the square root of the nutrient's molecular mass. I would be very surprised if all my approximations would combine to change the ratio of velocities I calculated by as much as 100.

I don't expect you necessarily followed that calculation. But the point is the following: simple estimates show that a Genlisea trap may be fully capable of generating a current into its stomach with a speed three times faster than the speed at which nutrient molecules could escape. This tactic would allow Genlisea to extract a greater percent of nutrients from its prey. Perhaps the water-sucking phase of a Genlisea trap only occurs when the trap is signalled by some mechanical or chemical means, analogous to the 20 minutes of water expulsion Utricularia bladders experience after they have been sprung. In fact, a Genlisea would have to draw fluid through its utricle for 18 minutes to completely change the fluid in its tube. It is striking that this is about the same time period as for a Utricularia bladder's water suction phase. Maybe Genlisea swallows!

Finally, while these calculations are interesting and even evocative, they do not prove anything. It might just be that despite any calculations Genlisea is a passive carnivore. Proof must await the laboratory and not the calculator. But an experimental investigation to prove or disprove the hypothesis that Genlisea is active would be relatively easy to perform. Place a chemically killed but structurally intact Genlisea trap next to a live and functioning one. Observations of how quickly dyes migrate through the tubes of each trap should reveal if the live trap is drawing dye into its utricle faster than the dead trap. Unfortunately I have neither the facilities nor the familiarity with biological lab methods to do this experiment to my own satisfaction, so I will leave that job to someone else. Clearly, this is a field of study that is in need of solid experiments for information and insights into the mechanism of this fascinating plant.


10 November 2007