How best to sample termites – or the importance of loo rolls

By Paul Eggleton

Based on research for the review paper, in press: “A.B. Davies, C.L. Parr and P. Eggleton. A global review of termite sampling methods. Insectes Sociaux.”

The three co-authors (from left to right: Kate Parr, Andrew Davies and Paul Eggleton) giving out a certificate to a student at a termite identification course in South Africa.

Social insects have always been difficult to sample. Do we count the colonies or the individuals; the organism or the superorganism? The bumble bees buzzing in your garden, are they really separate beings or just flying bits of an individual that is, in fact, much bigger?  But at least bees and wasps fly, so can be sampled using nets or traps that catch insects on the wing, and ants are usually running around on the ground, so can be sampled using traps that they fall into. When it comes to termites, only the reproductive caste flies, and all the wingless individuals, the soldiers and the workers are underground in the soil or hidden in dead wood.  This wouldn’t matter if termites were rare and unimportant things, but, in fact, they make up a huge amount of global animal biomass and are known to be extremely important for ecological processes. This poses a real problem for termite quantitative sampling, which is vital if we are to understand the role of termites in ecosystems. We discuss this in our recent paper, where we review termite sampling methods.

Termites are found predominantly in the tropics, in areas that traditionally have been hard to access. They are at their most abundant in tropical rain forests and savannas, places that are hot and often extremely humid.

Soldiers of Macrotermes bellicosus on the surface of dead wood, Kenya

The most obvious way to sample termites is to extract them from their mounds or nests. Termite mounds are easy to find, and although some of them may have a hard outer wall, are easy to get into. However, this will not give an accurate estimate of the species present as only a small fraction of termites have conspicuous mounds. Most colonies live underground, with no visible external structures. In fact, here are really only two ways to sample termites effectively – sampling them directly by picking them out of where they live or attracting them with baits.

A toilet roll out in the field for monitoring porposes (Sabah, Malaysia)

In dry areas, such as dry savannas and deserts, toilet roll baits are the most effective way of assessing termite diversity and activity. They are essentially cellulose, which is like a chocolate snack for a termite. In fact, termites generally gobble up the outer tissue rapidly and leave the inner roll behind. The termites can be taken directly out of the roll and the state of the roll can be assessed for termite activity. This works because in drier areas there is generally only wood feeders and there are few termites that can be extracted from the dry, hard soil. The main problem is that they are attractive to larger animals, particularly people, and they can, fairly often, be removed or disturbed by pigs or dogs. Toilet rolls are a major part of my group’s field work tools, and shopkeepers are often surprised when you buy their whole stock of loo roll in one go. They must wonder why on earth we need so much!

Toilet rolls waiting to be deployed (Sabah, Malaysia).

In wetter areas the problem there is the opposite – there are too many termites that live in the soil and are not attracted to cellulose baits. Soil-feeding termites become increasingly important as the environment becomes wetter and more stable. They feed directly on the soil and are important elements of the nitrogen cycle. Most of these termites are found living underground, with no obvious nest structures and so digging up the soil is the only way to find them. This is also related to the habitat complexity, as soil feeding termites are predominantly found in tropical rainforest, which show great horizontal and vertical variation, across the ground and up the trees. This means there are more places to search in tropical rain forests than in drier savannas – the soil, in mounds and nests, in and on trees, in dead wood, in the soil, and in the buttress roots of trees.

The authors’ student Fidele, sampling termites using the ‘transect method’ in a Gabonese rainforest.

We recommend some standardised methods in the paper, and essentially propose a standard method, and modifications of it that were developed in the late 90s, by my colleague, David Jones and me.  This involves intensive searching of a 100 m x 2 m “transect”, and takes, a rather gruelling, 20 person hours to complete.  In areas, such as wet savanna, where there are fewer trees, and many fewer places to search, we have reduced this to a fraction of this time, while still retaining the same general principles.

There are some habitats where sampling is harder. Grasslands tend to have surface foraging grass-feeding species that do not come to toilet roll baits and have unpredictable foraging patterns, usually at night. Sampling termites using a head torch is probably a step too far for even the most dedicated termitologist.