Plants interact with many organisms around them, and one of the most important groups that a plant has to deal with, are the herbivores. Insects represent the most diverse group of herbivores and have many different ways of using the plant as a food source. Plants can, however, defend themselves against those herbivores, either by constitutive defences, or by traits that are induced upon herbivory. These traits, such as the formation of more trichomes or the production of secondary metabolites, can deter an insect herbivore in its decision to eat from the plant, can be toxic, or otherwise hamper the insect to feed, grow or reproduce. The way a plant responds to herbivory is very specific, depending on the feeding mode or the species of the attacking insect. Furthermore, plant responses to dual herbivory differ from the sum of responses to each herbivore alone. Also the time and order at which multiple insects arrive on a plant, influence the plant’s response. Finally, plant species or populations can show different responses to herbivory. Altogether these factors result in a plant phenotype that the attacking herbivore has to deal with. In addition to the attacker, also subsequently arriving insects will be affected by a change in plant phenotype. Because plants and insects can respond to each other in a continuous chain of interactions, an herbivore early in the season can indirectly affect the later-season community composition through the induced plant response. However, we know only little about the consequences of a dual-herbivore induced plant phenotype for subsequent feeders, and ultimately, the effects on the assembly and dynamics of an insect community as a whole. The aim of this thesis project was to study the consequences of feeding by multiple insects from the same plant, not only to a subsequent herbivore, but also to the dynamics of a whole insect community over the course of a growing season, and beyond. Furthermore, I studied how the order of herbivore arrival and timing of arrival affected both a next herbivore’s choice and performance in a greenhouse setting, as well as the development of the whole insect community in the field. In addition, I studied how plant populations vary in induced responses, have specific plant-mediated interactions among insect herbivores, and how long these induced responses influence the insect community and plant fitness. Finally, I identified non-additive effects of the history of insect attacks and plant ontogeny to the future insect community. In the first chapter of this thesis, I introduce the study system. For this project I used several populations of the wild cabbage plant, Brassica oleracea. This is an herbaceous perennial plant that flowers from the second growing season onwards, and supports a large and diverse above-ground arthropod community of more than thirty different species. The plant belongs to the family of Brassicaceae, which is known for the biosynthesis of a group of secondary metabolites, the glucosinolates. These metabolites may deter insects, although some insect species use it as a feeding cue. Two specialist insect herbivores from different feeding guilds, the caterpillar of the diamondback moth Plutella xylostella and the cabbage aphid Brevicoryne brassicae, were used to study their effect as early-season inducer of plant responses either alone or in combination. The caterpillar of the generalist cabbage moth, Mamestra brassicae, was used in bioassays to assess the effects of the induced plant phenotype by single or dual herbivory. Furthermore, in three chapters (Chapters 4, 6 and 7) I have closely studied the composition of the naturally occurring insect community throughout the season for one or two years in a common-garden field setting. In the last of these three chapters, I used the caterpillars of the specialist cabbage white, Pieris rapae, to induce plants at different moments of their ontogeny, while excluding the insect community for varying periods of time by a net or exposing plants to their natural insect community. In an elaborate literature review, I and my collaborators concluded in chapter 2 that plant responses to dual herbivory evoke different plant responses than the sum of each herbivore alone. This has consequences at all levels from arthropod community assembly to the choice and performance of individual insects. The mechanisms of plant responses to dual herbivory are found in gene expression, hormone production and other molecular processes within the plant. All these aspects of interactions between insects and plants occur and are connected at different time scales. To follow up on the question how timing plays a role in dual herbivory, we varied the time between, as well as the order of arrival of aphids and/or caterpillars on a plant. We observed that both affected the preference and performance of a subsequently feeding caterpillar (Chapter 3). Mamestra brassicae performed better on plants with a longer time interval between the first and second feeder. Also in a field setting (Chapter 4), the order of herbivore arrival early in the season affected the insect community composition later in the season in two different years, likely through a chain of indirect insect interactions. In this field study, the plant population influenced the outcome of early-season herbivory to later community dynamics. In chapter 5 we found that three plant populations in response to simultaneous aphid and caterpillar attack differed in the expression of two genes that are important for the regulation of herbivore-induced responses. Also, the production of one of two important plant hormones, salicylic acid, responded differently to single or dual herbivory in a unique pattern for each of the plant populations. These different plant responses subsequently negatively affected a next caterpillar on the same plant; M. brassicae growth was impaired on plants which had been fed upon by both aphids and caterpillars, in comparison to control plants. These field and greenhouse studies thus show the implications of dual herbivory beyond effects in the plant; it affects subsequent herbivory, and through a chain of plant-mediated insect interactions, the dynamics of a whole insect community. In the sixth chapter we show that variation in insect community dynamics can last beyond the moment that the insects were present, even across years. In this field study, the naturally occurring carnivore community influenced the carnivore community composition a year later. Importantly, the herbivore community affected plant fitness across years (but not within years). We propose that such legacy effects are mediated by plant traits, which vary upon insect induction in the first year, and affect the insect community in the next year. Finally, the history of all insect attacks to a plant up until that moment shape the future insect community by influencing the colonisation of insect species on the plant. Moreover, also plant ontogeny plays a role in shaping the insect community; plant-mediated responses to herbivory at different plant ages resulted in different insect colonisation rates. The most important conclusion from this last data chapter (Chapter 7) is that the two processes, insect community history and plant ontogeny, are non-additive and affect the colonisation of insect species in the same (synergistic) or opposite (antagonistic) direction. By framing my study results in a time line from minutes to months to years, I show in the general discussion (Chapter 8) that the consequences of dual herbivory for subsequently arriving insects are connected at different time scales. Plant responses to herbivory can occur within hours to days, which affect herbivore choices and performance in the following days and weeks. In their turn, variation of a few days in arrival time of insects may change how plants respond and prioritize their responses to insects throughout the rest of the season. The insects that subsequently arrive on a dual-herbivore induced plant may change the plant phenotype even further and through a chain of insect-plant interactions, the effects on the insects and the plant can last throughout the season, and even across seasons. Furthermore, various factors such as the species of the attacking insect and its feeding guild, the timing after previous attack and the plant age at which herbivory occurs, as well as the genotypic background of the plant, all affect the outcomes of dynamic insect-plant interactions. The results presented in this thesis thus contribute to the knowledge and interpretation of plant interactions with multiple herbivores. As plants are seldom attacked by a single herbivore, this implies that we have to take into account that multiple herbivory is not the same as the additive effects of single herbivores, and that this has long-lasting consequences for the insect community and the plant. To further understand how plants and insects have adapted to such a dynamic environment, I suggest future research to focus even more on the kinetics of plant physiological responses to dual attack, and to aim at answering the question of how predictable insect communities on a plant really are.