Scarabaeid pests in South Africa and especially KwaZulu-Natal are characterised by a very long larval life cycle and short pupal and adult periods. However, it has nearly always been the adults of the species that have been identified, with very little attention paid to the larval identification of the species. This is unfortunate as it is nearly always the larval stage that is found to be associated with crop damage. Accurate identification of the species of these larvae is important for the management of scarabaeid pest species, as it unlocks the necessary information on the biology and ecology of many species, which allows the adaptation of control methods for different species. Inadequate keys for the taxonomy of larvae of these groups, as well as the lack of morphological taxonomists working on these groups have been identified as constraints. When a species is difficult to identify using traditional taxonomic methods, DNA diagnostic tools can be useful. Chapter 2 investigated the feasibility of identifying scarabaeid larvae using mitochondrial DNA data. Variation in the base pair sequence of the mitochondrial cytochrome c oxidase sub unit I (cox 1) gene was used. DNA sequences of cox 1 from scarabaeid larvae collected from sugarcane fields were compared with sequences from scarabaeid adults of known species in order to identify the species attacking sugarcane. Neighbour-joining and maximum parsimony analyses of 658 bp cox 1 sequences identified groups of larvae that linked to adult specimens. The major groupings delimited specimens belonging to the subfamilies Dynastinae, Melolonthinae and Rutelinae. Within-group sequence divergence ranged from 0 - 3.4 % and divergence between sister groups ranged from 2.6 - 25.1 %. The recorded divergence range within and between tribes was 0 - 21.3 % and 17.3 - 28.5% respectively. Similarly, the divergence range observed within and between genera was 0 - 19.2 % and 17.1 - 25.4% respectively. The maximum sequence divergence observed within subfamilies was 23.7 % and divergence between subfamilies ranged from 16.8 - 26.7 %. Examination of pairwise sequence divergence levels as well as node support allowed 68% of the unidentified larval specimens to be associated with identified adult specimens. Phylogenetic analysis matched identified adult mtDNA with unidentified larval mtDNA. This allowed the identification of those larvae through morphological characteristics unique to certain species. To create a field key to the subfamilies of Dynastinae, Melolonthinae and Rutelinae the most useful character distinguishing larvae of different species was the raster but additional morphological characteristics were included. These relationships between larval and adult scarabaeid specimens from sugarcane were examined using various phylogenetic tools. The data set included a total of 19 morphological characters as well as 166 partial cox 1 gene sequences. Maximum parsimony analyses were performed on morphological, molecular and combined data. The same morphological and molecular data sets were run both separately and as a combined analysis with MrBayes. In both types of analyses the morphological data performed poorly and crude groupings resulted, dividing taxa to tribe level only. Molecular data showed greater resolution than the morphological data and taxa were separated into groups equivalent to species and morphospecies designated in Chapter 2. A partition homogeneity test indicated that both data types could be combined. It is recommended that both morphological and molecular data be utilised in identification of scarabaeid sugarcane pests and that a character-based approach be implemented. Further molecular data from other genes should be included to test the accuracy of these results. The keys produced during this study will allow workers to focus on a single species biology, and subsequently allow an analysis of between species interactions, and within species control. These advances are a start to the improvement of knowledge of the species composition of scarabaeid larvae in sugarcane fields, thus making management and biological control of these pests a greater possibility. Further recommendations for future work are discussed in Chapter 5. / Thesis (M.Sc.)-University of KwaZulu-Natal, Pietermaritzburg, 2008.