Climate change might severely disrupt the effective maintenance of agricultural communities under biocontrol management. Most of pests and natural enemies used to control them are arthropods and, as ectothermic organisms, many of their physiological functions, and the features of many of the biotic interactions individuals will establish with other individuals during their life span, are climate-dependent. The future success of biological pest control, thus, might depend on the capacity of predicting how each agricultural community will respond to a changing climate. The main goal of this thesis was to generate knowledge on the effects of changes in abiotic conditions on the way how biotic interactions occur among species that inhabit agricultural communities under biological pest control, and how these effects determine the trophic structure and dynamics of these communities. To achieve this goal, in the present thesis I studied a mite community present in avocado agro-ecosystems in South-eastern Spain. Because this community is composed by a low number of species, it is an optimal study model to understand the effects of abiotic conditions on interactions between pests and their natural enemies. The avocado mite community is composed of the herbivore pest Oligonychus perseae (Tuttle, Baker & Abatiello) (Acari: Tetranychidae) and two species of phytoseiid mites: Neoseiulus californicus (McGregor), which is specialized in tetranychid mites, and Euseius stipulatus (Athias-Henriot), which is preferentially pollen feeder. The work presented in this thesis was addressed from two different approaches. On the one hand, the effects of abiotic conditions on the strength and direction of biotic interactions in the avocado mite community were evaluated at the individual level, and at three combinations of temperature and relative humidity: mild (M), hot and dry (HD) and very hot and dry (VHD). On the other hand, it was studied the impact of abiotic conditions on the community dynamics at M and HD conditions, i.e., how the observed effects on species interactions at the individual level were rendered at the population level, determining the trophic structure and dynamics of the community. The work started evaluating how abiotic conditions affected the strength of predator-prey interactions between either E. stipulatus or N. californicus, and their herbivore prey O. perseae, in the presence and the absence of alternative food (i.e. pollen). Results revealed that changes in abiotic conditions had different effects on predator-prey interactions depending on the species of predator. At M conditions both species of predatory mite preyed on O. perseae females, whereas at HD and VHD conditions only N. californicus fed on the prey. Furthermore, the strength of interaction between N. californicus and O. perseae varied among abiotic conditions, being the highest one at M conditions. The presence of alternative food (i.e. pollen) influenced in predator-prey interaction strength between E. stipulatus and O. perseae, but not between N. californicus and O. perseae. Euseius stipulatus preyed on O. perseae females in the absence of pollen at M conditions, but clearly preferred to forage on pollen when was available. Also, oviposition rates of E. stipulatus were increased in the presence of pollen at the three abiotic conditions. Therefore, the addition of pollen as alternative food for E. stipulatus might promote the growth of its populations and favour pest control through apparent competition between the pest and pollen. Next, I studied the effects of abiotic conditions and presence of alternative food on intraguild interactions between the two predatory mites, E. stipulatus and N. californicus, with O. perseae as shared prey. Results showed that the trophic structure of the avocado mite community changed with changes in abiotic conditions, resulting in community configurations that would be detrimental to pest control. At M conditions, results revealed that the community resembled a trophic chain, with juveniles of N. californicus preyed on O. perseae, and females of E. stipulatus preyed on juveniles of N. californicus. On the other hand, juveniles of E. stipulatus fed on O. perseae and induced antipredator behaviour in females of N. californicus, which did not interact with the pest at all. Increasing unfavourable abiotic conditions shifted the structure of the community to one dominated by exploitative competition between the two species of predators. However, results also revealed that adding pollen to the system would likely enhance pest population control, because trophic interactions between predators ceased due to E. stipulatus preference for pollen, what strengthened the predator-prey interaction between N. californicus and O. perseae, at the three abiotic conditions. Finally, mite predator/prey population dynamics were evaluated in the presence and the absence of alternative food at two different abiotic conditions, M and HD. Contrary to expectations, results revealed that the addition of pollen did not reduce negative trophic interactions occurring between predator species; instead, E. stipulatus excluded N. californicus at both abiotic conditions, independently of the presence or absence of pollen. Also, in spite that the addition of pollen to the system favoured a numerical response of E. stipulatus populations at M conditions, population increase did not translate into a better O. perseae population control. Furthermore, at HD conditions, the growth of E. stipulatus populations was negatively affected by high temperatures and drought. Therefore, results from this thesis evidence that abiotic conditions modify the way in which species interact, affecting trophic structure and dynamics of communities. This implies that in a changing climate the successful use of biocontrol agents will need to take into account the influence that abiotic factors exert on interactions occurring among the components of each specific agricultural community.