The termites and their fungal symbionts respire in the same way we do: obtaining chemical energy for work by oxidizing food with oxygen, producing carbon dioxide waste in the process. As in ourselves, the higher the colony's rate of energy consumption (or its metabolic rate), the faster it must exchange of oxygen and carbon dioxide with the atmosphere.
The Macrotermes colony stands out in the termite world as having a very high rate of metabolism: various estimates put the colony's metabolic rate at 55 watts up to more than 200 watts. To put this number in perspective, the resting metabolic rate of a typical human is in the range of 80-100 watts. To sustain these high metabolic rates, oxygen must be fed into the colony at a high rate, just as we require. Similarly, carbon dioxide waste must be rapidly vented out of the nest, just as it must be from our own bodies. In ourselves, these high rates of gas exchange are sustained through muscle powered ventilation of the lungs. Without adequate ventilation, our bodies would suffocate. The same is true for the colony: it must be ventilated, or it will suffocate.
Ventilation involves the bulk movement of air. Air has mass, and it is viscous: ventilation therefore requires work to be done to overcome its inertia and viscosity. In our own bodies, the work of ventilation is done by muscles powered by metabolic energy. In the Macrotermes nest, two energy sources power the work of ventilation: kinetic energy in wind, and heat energy derived from metabolism. The mound is the venue where these two energy sources are modulated and integrated to provide reliable ventilation.
Scroll down to read about how the mound interacts with wind. The next page examines the effects of metabolism-induced buoyant flow.
When wind flows over the ground, or past any immovable object, there is a transfer of momentum from the moving air to the object. This translates into a gradient in wind velocity called a boundary layer (right). Because the mound is built upward from the ground, it spans the boundary layer: mound surface close to the ground encounters slow winds, while mound surface high off the ground encounters faster winds. This allows the mound to tap the the energy in the boundary layer's velocity (=momentum) gradient.
The gradient in wind speed also allows termites to adjust the capture of wind energy to power ventilation. By building the mound to different heights, for example (left) the mound captures different quantities of wind energy: the higher the mound, the more energetic the winds. If, say, the captured wind energy is insufficient to ventilate the nest properly, the termites can correct this by simply building the mound higher, extending it into more energetic winds. This is why, for example, mounds in wooded areas tend to be taller than mounds in more open areas.
The mound captures kinetic energy in wind by converting it to potential energy, i.e. pressure. These dynamic pressures, as they are called, are readily predicted by the Bernoulli principle.
When winds intercepting the mound are made to slow down, as they are at the mound's windward surface (left), the lost kinetic energy is transformed to a positive pressure (pink arrows). Where winds intercepting the mound are made to speed up, as they will when they pass around the mound, the added kinetic energy is compensated by a reduction of pressure, i.e. a suction pressure. Suction pressures are particularly strong at the mound's lateral surfaces with respect to wind (middle). The mound's leeward surface (right) will also experience suction pressures, although these will be weaker than at the mound's lateral surface.
Measuring these pressures with a micromanometer shows precisely this pattern. High off the ground (+2.5 m: red bars), where wind speeds are high, wind-induced pressures at the upwind surface are positive, roughly 10 Pa on average. Lateral to the wind surface, dynamic pressures are strongly negative, roughly -35 Pa on average. Weaker suction pressures prevail on the downwind side, roughly -10 Pa. Close to the ground (+0.5 m: blue bars), this pattern is repeated, but at smaller scale, as is expected from the slower wind speeds close to the ground.
Thus, the mound's interception of wind imposes a complex pressure field on the mound's surface. Because the mound surface is porous, these pressures can drive air across the mound surface, forcing it inward on the upwind surfaces, and drawing it out at the lateral and downwind surfaces (the actual flows of air within the mound are outlined in more detail here.)
This wind-induced pressure field at the mound surface affects pressures within the mound. The vertical gradient in wind-induced pressure translates, for example, into a vertical pressure gradient within the surface conduits (right) that is correlated with wind speed. These vertical pressure differences can drive bulk flows of air within the mound, particularly in the surface conduits.
In the natural world, winds are mostly turbulent, that is they exhibit a substantial chaotic variation in both wind speed and wind direction. One can dissect this variation using a device known as a frequency spectrum, which derives from a mathematical technique known as a Fourier series. In a nutshell, any time series (like a temporal variation of wind speed) can be dissected into a series of component sine waves--the Fourier series--which, when added together, reproduce the original time series. Thus, turbulent winds consist of a mixture of high-frequency components (winds whose sinusoidal variation is rapid), intermediate-frequency (winds whose sinusoidal variation is more moderate) and low-frequency components (winds that vary only slowly with time). This temporal variation of winds is called the transient, or AC, component. High-frequency components dominate the frequency spectrum of turbulent winds, while lower frequency components dominate the frequency spectrum of steadier winds. Laminar steady winds, with no time-dependent variation, have no frequency spectrum. This is known as the steady, or DC, component.
The pressure field around a mound that is intercepting turbulent winds will also have a frequency spectrum. This poses a conundrum for the function of the mound as an organ of physiology. To serve a physiological function well, an organ of physiology must perform reliably. In a conventional "inside-the-skin" organ, reliability is ensured by the reliable provision of energy supplied in ATP. When the energy supply is chaotic, as it potentially could be in "lung" driven by chaotic energy in winds, the physiological function may not be so reliable.
The mound solves this problem by acting as a "low-pass filter" for the broad frequency spectrum of turbulent wind energy. A filter is a device that selectively blocks particular frequencies of the transient component of an energy field. A "high-pass" filter, for example, only lets through the high-frequency components, and blocks the low-frequency ones. A "low-pass" filter does the opposite: admits the low frequency components and blocks the high frequency ones. An extreme "low-pass" filter admits only the steady component. This filtering of high frequency components is also known as "damping."
The network of tunnels in the mound acts as a damper for the high-frequency components of turbulent wind. This is seen in the graph to the left, which depicts the results of a tracer gas pulse-chase experiment. Here, a bolus of propane tracer gas is injected into one of the surface conduits (open arrow), and its movements are followed using combustible gas sensors on the chimney (red trace) and two surface conduits (green and magenta traces). Wind speed is depicted in the blue trace (light blue is wind speed every 15 seconds, dark blue is the 5 minute rolling average wind speed). At the injection site (green trace), the tracer concentration rises rapidly, then declines as tracer is "washed out" of the mound air space. As the tracer migrates from the injection site, it appears later in the chimney (red trace). The damping effect of the mound is evident as the much "choppier" trace in the surface conduit (green) as air flow is driven by the rapid variation in wind speed. In the chimney (red), however, the trace is very smooth: the rapidly varying components have been filtered out, leaving only the steady and very low-frequency components to drive the washout of tracer.