Received: 13 August 2010 Revised: 5 October 2010 Accepted: 7 October 2010 Published online in Wiley Online Library: 00 Month 2010 Dear Editor, How to preserve termite samples in the field for carbon and nitrogen stable isotope studies? The measurement of stable isotopes of carbon ( 13 C) and nitrogen ( 15 N) is a powerful tool in ecological studies, since it can indicate food preferences of organisms. [1] This is possible because these isotopes are available in the environment and are acquired during feeding. Thus, the isotopic composition of animal tissues reflects the diet gained and assimilated throughout the animal’s life. [2] Particularly for termites, whose diets vary subtly within a decomposition continuous from sound wood to highly humified material, isotope analyses can be rather revealing. In such cases, carbon isotope signals indicate the source (e.g. C3 or C4 plants), [3] whereas nitrogen isotopes generally reflect the decomposition degree of the diet [4,5] (although this is not straightforward for wood-feeding termites which are able to fix N 2 from the atmosphere). [6] The analyses demand that specimens should be processed soon after collection to prevent deterioration and, therefore, isotopic changes. Termites, however, are highly prone to decompose due to their weakly chitinised and very fragile bodies. For this group, therefore, more attention to prevent deterioration and isotopic alteration is needed. Currently, for isotopic analysis, termites are immediately dried at 608C for 24 h [1,7–9] or frozen after collection. [10] Very often, however, collections occur in remote regions, away from well- equipped labs. In such places, fast processing is unfeasible, as this demands the availability of techniques for specimen preservation to avoid sample decomposition. The conventional technique for preserving termites is their immersion in ethanol 80% which maintains the morphological characteristics in the long term. [11] Being an organic compound with liposolvent capabilities, ethanol could affect the carbon content and, hence, be unsuitable for preserving samples for carbon isotopic analyses. [18] This theoretical unsuitability of ethanol, however, has not been always confirmed, for vertebrates and invertebrates. [12] Ethanol-preserved tissues of quails, [13] sheep, [13] turtles [14] and caddisflies, a Trichoptera, [12] have shown no alteration in carbon isotopic signature. Alternatives would include inorganic (carbon-free) preserving substances, of which sodium chloride (NaCl) seems the cheapest, the best known, and also available worldwide. This work, therefore, aimed to establish a technique for preserving termites which would suit field work in remote regions while still allowing d 13 C and d 15 N isotopic analyses. To do so, the isotopic ratios and the C/N ratios of stored samples of Cornitermes cumulans termites were compared with values for samples analysed immediately after collec- tion. The stored samples were kept in vials containing (i) NaCl solution and (ii) ethanol. Our rationale is that stored samples should vary in their carbon and nitrogen content, because (i) ethanol-preserved samples are subject to lipid loss, whereas (ii) NaCl-preserved samples should keep their original carbon and nitrogen content. Therefore, we hypothesise that freshly processed samples should not differ from NaCl-preserved samples and would differ in their carbon content from ethanol-preserved ones. The experiment was performed using worker termites (third instar and beyond) from three field colonies of Cornitermes cumulans (Kollar) (Isoptera, Termitidae), in Vic ¸osa, state of Minas Gerais, in southeastern Brazil. Cornitermes spp. are Neotropical termite species occurring in several habitats, including forests, ’cerrados’ (Brazilian savannas) and man-modified habitats, such as pastures or even gardens within cities, where they feed on living and dead grass and herbs. [15] As soon as mound fragments were taken to the laboratory, termite specimens were extracted and allocated to the appropriate treatments, as described below. Voucher speci- mens were preserved in 80% ethanol, labelled and identified by comparison with the collection of the Termite Section of the Entomological Museum (UFVB) of the Federal University of Vic ¸osa, Vic ¸osa, Brazil (UFV). This work was carried out from December/2009 to February/2010. From each termite mound, 30 samples of five workers each were collected to compose ten replicates for each of the processing techniques below. Freshly processed workers were placed in vials with distilled water, immediately frozen at –148C for about 2 h and then freeze-dried. This is thought to be the optimal procedure for isotopic analyses, as it avoids sample degradation. Preserved in ethanol workers were placed in vials with 80% ethanol, at room temperature, for 49 days until washed in distilled water and placed in vials with distilled water to be freeze-dried. The ethanol concentration was confirmed by an alcoholmeter, after adding distilled water to 92.88 GL commercial sugar cane ethanol (Miyako do Brazil Industria e Comercio Ltda, Guarulhos, Brazil). This is the traditional method for maintaining termite samples in collections, and it is expected to be only a partially suitable procedure, as it is bound to affect the d 13 C isotopic signal, because the alcohol could leach lipids and add carbon to the sample. Preserved in NaCl workers were placed in vials containing a brine solution of table salt (300 g/L), at room temperature for 49 days, until freeze-dried. A brine solution was obtained by adding salt to distilled water, at room temperature, until the water could no longer dissolve any more salt. This is thought to be the best procedure to our aims, as it is cheap, straightforward, and prevents sample decomposition without masking the isotopic signals, as the compound contains no carbon or nitrogen. To proceed with the isotopic analyses, all the samples were freeze-dried for 48 h, ground, sieved (mesh ¼ 100) and placed in tin capsules. The d 13 C and d 15 N isotopic ratios, and the C/N ratio in each sample (1.5 mg), were determined in an isotope ratio mass spectrometer (ANCA-GSL 20–20, SerCon Ltd., Crewe, UK), at the Laboratory of Stable Isotopes, Soils Department, Federal University of Vic ¸osa. Rapid Commun. Mass Spectrom. 2011, 25, 243–246 (wileyonlinelibrary.com) DOI: 10.1002/rcm.4820 Letter to the Editor Rapid Commun. Mass Spectrom. 2011, 25, 243–246 Copyright ß 2010 John Wiley & Sons, Ltd. 243