In the study of the feeding behaviour and nutrition of free-ranging mammalian herbivores, determining what the animals are eating, its quality and quantity can be difficult to accomplish. The measurement processes themselves may disturb the animals’ normal foraging behaviours which can be a major problem in rangeland, forest and other semi-natural environments. Furthermore, animals are likely to select mixtures of plants and their components which differ from the available vegetation. Quantitative measures of diet composition, digestibility, faecal output and intake in individual grazing or browsing animals have depended on the use of faecal markers. These are materials measurable in faeces that originate from the diet (internal markers), or are absent from the diet, but administered by oral dosing (external markers). The ‘ideal’ faecal marker needs complete recovery in faeces, simple and accurate quantitative measurement, inertness in having no effect on the animal or its diet, and similar physical characteristics (in terms of particle size and density) to the digestive tract contents. No individual material or chemical entity has been found which fulfils all of the ‘ideal’ marker attributes. For example, lignin, indigestible acid-detergent fibre (IADF) and ‘chromogen’, have been used as internal markers, but since they are not discrete compounds, analytical methods are empirical., resulting in inconsistent faecal recoveries.

Virtually all higher plants have an outer surface layer of wax, which is usually a complex mixture of aliphatic lipid compounds whose composition differs between plant types, and different parts of the same plant. Plant waxes can be analysed as discrete compounds, are relatively inert, and because the patterns of individual compounds tend to differ between plant species, they offer the potential of enabling measurement of the contributions of specific plant species to the diet. Leaves and floral parts tend to have the highest concentrations; roots have very low levels. The main classes of plant waxes are straight and branched chain alkanes, alkenes, long–chain fatty acids and esters, long–chain fatty alcohols; long–chain fatty aldehydes and ketones and b–diketones. Analysis is usually carried out by a stepwise process of solvent extraction, purification and gas chromatography (GC). Straight–chain alkanes (n–alkanes) have been the most commonly used marker to date, being present as mixtures with chain lengths ranging from 21 to 37 carbon atoms. Over 90% of n–alkanes have odd–numbered carbon chains, with C29, C31 and C33 alkanes being dominant in most pasture species. Recovery in faeces of plant waxes is high but not complete, and is related to chain length, the longer the chain, the higher the recovery. Correction factors have been measured in a number of herbivore species.

The first application of plant-wax n-alkanes as faecal markers was to determine herbage digestibility in ruminants. Subsequently it was realised that dosed synthetic alkanes could be used to determine faecal output, and hence dosed and herbage alkanes could be concurrently used to estimate intake. This offered substantial advantages over other methods. Furthermore, differences between plant species and parts in their patterns of individual alkanes can be exploited to enable quantitative determination of diet composition from the patterns found in faeces. Since differences in the relative faecal recoveries of individual markers could modify the faecal marker pattern, recovery corrections may be necessary. This approach was first used for measuring the composition of simple dietary mixtures and the intake of dietary supplements to be determined. The use of long-chain fatty alcohols and fatty acids as additional markers offers the potential for more complex diets to be evaluated. The fact that plant-wax alkanes remain attached to particulate dietary residues throughout the ruminant gut, means that they are also good markers for determining the rate of passage of material along the digestive tract

Estimates of digestibility and faecal output obtained from respective natural and dosed n-alkanes will be biased, unless corrections are made to account for incomplete faecal recoveries. However, intake estimates will be unbiased if the faecal recoveries of the two markers are the same. Studies in sheep, cattle and goats have shown plant C33 and dosed C32 alkanes to have very similar faecal recoveries and thus give unbiased estimates of herbage intake. The alkane method for estimating intake offers advantages over other techniques. It gives individual-animal intakes and can be used where animals are receiving feed supplements. Also, GC analysis allows both plant and dosed markers to be determined at the same time, which limits analytical time, error and bias. Since the ratio of the concentrations in faeces is used, it is not necessary to obtain absolute faecal concentrations.

Alkenes and branched-chain alkanes have been investigated as additional markers for diet composition estimation since they can be quantified in the same GC analysis as n-alkanes. Although alkenes, which tend to be associated with floral plant parts, have low recoveries (25-40%), they can be useful diet composition markers since their recoveries are little affected by chain length. Faecal recoveries of the branched-chain alkanes, from Agrostis capillaris herbage, were slightly lower (60-65%) than the respective n-alkanes (C30 and C32) of equivalent carbon number (85-90%). These alkanes are rare in forage species, and their practical usefulness as markers for quantitative composition estimation has yet to be tested.

Long-chain fatty alcohols have been shown to be effective diet composition markers. In most plants fatty alcohol concentrations are higher than those of hydrocarbons, and there can be profound differences in composition between species. They may be of particular value for diets containing plants with low alkane concentrations. Faecal recoveries in sheep, like n-alkanes, increase progressively with chain length from about 60% to 90%. It has been shown experimentally that the use of alcohols, together with n-alkanes, is likely to give a better estimate of diet composition in a given situation than n-alkanes alone.

The very long-chain fatty acids of plant cuticular wax (C20-C34), originally suggested as digestibility markers, may also have potential as diet composition markers. Like n-alkanes and long-chain fatty alcohols, the faecal recoveries of plant-wax fatty acids in sheep increase with carbon chain length. Comparisons with n-alkanes and fatty alcohols suggested that the fatty acids were inferior as diet composition markers. This may have been due to the fact that the fatty acid extracts analysed by GC were relatively impure, containing a number of unidentified compounds. The reliability of plant-wax fatty acids as markers may be improved with more effective analytical procedures.

There are a number of ways of calculating the diet composition from marker patterns in the faeces and potential dietary components. A simple approach is to determine a solution from a matrix of simultaneous equations; the number of dietary components must equal the number of markers used. Because, for simple dietary mixtures, there may be more available markers than dietary components, difficulties may arise in making the best choice of marker. Least-squares optimisation methods allow the number of markers to exceed the number of diet components, and thus (in theory) make better use of available information.

Since the concept of using faecal marker patterns for making quantitative diet composition estimates is relatively new, the associated mathematical and statistical procedures used to date have been rather crude and simplistic. There is potential to make more effective use of the marker data by using more sophisticated computational approaches. These include a range of multivariate techniques, including: a) Principal component and discriminant analysis; b) The weighting of the contribution of different markers, since with the leastsquares optimisation procedure, markers with the highest overall concentrations contribute most to the composition estimate, even though some markers with low concentrations may have large relative differences between dietary components. It would logical to weight markers in favour of those having the greatest relative variation across dietary components, and those providing the least compositional information could be weighted against; c) Statistical procedures are needed to evaluate the quality of diet composition estimates. The minimisation procedures described earlier take no account of any within-component variation in marker composition, and the effect of such variation on the quality of resultant diet composition estimate is not known. Attempts are being made to develop statistical procedures which will provide details of confidence intervals of compositional estimates resulting from particular plant species mixtures of known within- and between-species variability in marker composition.

For reliable estimation of digestibility, intake and diet composition, the feed sample must be representative, with respect to its plant wax marker concentration, of the material ingested by the animals under investigation, not necessarily all of the material present. Since marker concentrations can differ for different plant parts and plant species, care must be taken in sampling the vegetation for analysis. Although oesophageal-fistulated animals have been used to collect samples of ingested vegetation, hand-plucked grass samples have been found to be adequate for uniform grass swards,. In heterogeneous vegetation environments, especially when browse species are present, herbivores are likely to be highly selective. It is thus wise to make preliminary observations of animals’ ingestive behaviours, so appropriate parts of the dietary plants are sampled.

The use of plant waxes, initially with alkanes, as markers in the study of the diets of domestic ruminants is finding increasing use in other herbivores, both domesticated and wild (moose, fallow deer, mountain hares, pigs, rabbits, horses, donkeys, giraffes). Such methodologies have been applied in non-mammalian herbivores, including birds (pigeons, and ostriches) and reptiles (tortoises). Synthetic alkanes have been used as markers to estimate digestibility and intake in fish. Although tests have not yet been carried out, plant wax marker methods may even be applicable to non-vertebrate herbivores, such as caterpillars, slugs and snails. Expansion of the technique to include plant wax compounds other than alkanes will broaden the spectrum of animals and systems which can and could be studied. Alkanes can be used to estimate the botanical composition of plant mixtures, including mixed root mats, and since these compounds can remain in soil for a considerable period, they may also be used to describe the vegetation history of an area by analysis of soil strata. There may be potential for this approach to be extended into archaeological and forensic studies. Insects and spiders contain hydrocarbons (mainly branched-chain alkanes) in their cuticular wax, and preliminary tests have indicated that these compounds are recoverable in the faeces of bats and insect-eating birds; thus they could be used to determine the insect species composition of the diets of insectivores. Analysis of these waxes is relatively simple, and although good GC equipment is required, there is scope for laboratories which do not posses the equipment to make the initial extractions and purification (simple but laborious), with the final GC analysis undertaken by a collaborating laboratory. Thus there are many uses of plant and animal waxes as markers, and potential applications are probably limited only by our imagination.