A Dissertation submitted to the Faculty of Science, University of the Witwatersrand, Johannesburg, in fulfilment of the requirements for the degree of Master of Science. Johannesburg, 2012 / Various insect taxa are known to modify bone with their mandibles, including members of the orders of Dermestidae, Tenebrionidae, Calliphoridae, Tineidae and Termitidae. Despite bone modification being a known behavioural trait of many of these taxa, little work has been done to record the distinctive ways in which they modify the bone surface, and a lack of concise descriptions of modification suites inhibits decisive identification and interpretation. The most widely inferred causal agents in palaeontological literature are either termites or dermestid beetles, whilst cockroaches as potential bone modifying agents have not yet been considered. The primary aims of this investigation were to establish whether or not cockroaches and dermestids modify bone, and if so in what ways, develop an interpretative framework to aid future researchers in the identification and differentiation between the variously reported agents of bone modifications, test whether or not the agents will modify bone of varying densities (thin cortical, thick cortical, compact and cancellous bone) or in a particular state of preservation/condition (fresh, dry, weathered or fossilised), and investigate whether or not the occurrence of insect modifications on bone can be used as a proxy to establish a broad climatic signature based on their known thermal physiological limits. A single experimental trial of 18 bone specimens were exposed to the African cockroach Periplaneta americana for a period of six months and a further four experimental trials (totalling 80 bone specimens) were exposed to the Coleopteran Dermestes maculatus for periods of four months each under the absence or presence of substrate and variable feeding conditions. Experiments were conducted within an insectary at 28° C, 40 % humidity and 12 hour light/ 12 of darkness. Subsequently, all specimens were viewed using an Olympus SZX 16 Multifocus microscope fitted with a digital camera at magnifications between 7 and 115x. Three modification types were identified for P. Americana, namely discolouration, destruction of bone and gnawing. A total of five modification types were established for D. maculatus including the occurrence of surface tunnels, destruction of bone, bore holes, surface pits (Classes 1–3) and gnawing. Three distinctive surface pits morphologies were identified; Class 1 pits are highly variable but most often semi-circular to elliptical shallow depressions with a U-shape profile with striations radiating around the outer circumference of the depression. Class 2 surface pits are semi-circular shallow depressions with randomly orientated striations occurring over the entire feature. Class 3 surface pits are irregular shaped depressions with complex profiles not associated to gnawing striations. Broad climatic signatures for both of these agents were developed based on their known physiological thermal limits. The indistinct modification signature of P. americana in combination with limited occurrence and frequency patterns may prove difficult to identify from an archaeological or palaeontological context. Periplaneta americana and D. maculatus do significant damage to aves bones, which could result in their under representation in the archaeological and palaeontological records. The highly distinctive signature as well as occurrence and frequency patterns of modifications produced by D. maculatus has enabled the reinterpretation of existing palaeontological analyses, suggesting that dermestids are in fact not responsible for reported instances in which they are suggested as the causal agent during the Mesozoic and Cenozoic.