The induction of systemic resistance involves the activation of latent resistance mechanisms in plants against pathogens in response to the treatment with biotic or abiotic agents. From the evolutionary point of view plants developed a latent defense system, that can be activated, with the goal of saving energy, contrary to the constitutive resistance that represents a real cost for the plant, since independently of the pathogen presence the plant invests its limited resources in the production of these defense factors. The costs are defined as “all negative effects on plant fitness that result from the expression of a defence trait when a plant grows under evolutionarily relevant conditions”. Thus, the induced resistance under natural conditions will represent cost only in the pathogen presence. The cost is compensated by the time delay in the defense expression so that resources will be allocated for this purpose only when necessary. According to the allocation resource hypothesis when the pathogens are present the investment in defense must worth and the induced plants must be benefited . Plants that invest their resources to defend themselves in the absence of the pathogen will pay off with costs that will reflect in productivity, since the metabolic changes that led to resistance have an associated fitness cost which could outweigh the benefit .

The cost of the induced resistance in plants is difficult to be measured because it would be necessary to have a susceptible host in a state completely not induced. When a seedling starts to interact with its environment, genes will start to be activated in response to it. For example barley plants when grown in a microorganism free environment exhibited increased susceptibility to powdery mildew, suggesting that in nature resistance can be induced by microorganims surrounding the plant. Thus, plants at the field level always would be in an intermediated state of induction . Because of the difficulty to measure the total cost of the induced defenses, researchers have determined only the costs of a specific inducer by comparing induced and not induced plants. This approach can lead to mistakes when results from different researchers are compared, even where opposite results can be observed.

Contradictory results can be generated when species of varieties from different plants are compared as well as from environmental changes. The environmental influence was shown applying acibenzolar-S-methyl (ASM) in Arabidopsis plants grown under different conditions of water and nitrogen supply and, working with wheat, where it was seen that the cost was dependent upon nitrogen supply as well as the stage when plants were induced. Thus, small changes in the experimental approach and in the inducer used, associated to variations among different plant-pathogen interactions and the use of high inoculum density can influence the results and the understanding of them. In order to compare different systems and extrapolate the results to natural conditions it must be considered the inducer concentration, the application methodology, the absence of other inducers that can be present in the extracellular fluid from the challenging microorganism, the host plant species and its stage of growth .

The differences in costs can also be associated to the activation of different signaling pathways. Each pathway can be independently activated by specific elicitors, and each one will induce the expression of a specific group of genes that represents a fitness cost for the plant. Among these expressed genes some are not related to defense . For example, coffee plants treated with ASM showed the additional expression of 302 genes, but only 16% of them were related to resistance mechanisms. Besides the costs related to just one pathway, interactions among pathways can happen and the results will be positive when the benefit is higher than the cost, negative when increases in metabolic costs happen and protection does not occur or neutral. In these context, it is still important to consider that when one induced plant is challenged by the inoculation of a pathogen, the magnitude of gene expression is increased and other genes are expressed when compared to a noninduced and challenged plant . Thus, all these can increase the difficulties in planning an experiment to evaluate the cost of induced resistance.

Although the costs directly related to gene expression are the most important ones, it is also interesting to consider that the production of toxic compounds to stop the intruder, as phytoalexins and reactive oxygen species, can also cause damage and reduce plant productivity. Antimicrobial compounds can also cause an ecological cost when interfering in mutualistic relationships between plants and symbionts, as mycorrhiza and rizobia, avoiding that the plant receives the benefits from these interactions.

Our research work with biotic ( Bacillus cereus ) and abiotic (ASM) inducers, using bean plants, has been showing that as the number of ASM application increases the dry plant biomass and productivity decreases, but this is not seen with the biotic agent, although both inducers can cause increases in plant respiration. The negative effects on plant productivity usually occur when chemical inducers are used repeatedly or in higher doses, mainly in the absence of the pathogen. Thus, the authors suggested that to get the most out of the resistance inducers the number of applications and the dosages must be optimized, and the plant stage that respond better to the resistance induction needs to be defined.

Finally, taking in consideration the above aspects, we can say (2005), that in some cases we can be walking on a fine line between cost and benefit , where the cure may be as bad as the disease itself.