Adaptive mutations drive much of the evolutionary diversity observed in nature. Understanding the relationship between genotype, phenotype, and fitness in adaptive mutations is therefore essential to understanding adaptation and evolutionary processes in general. As fitness is dependent on the environment, it is necessary to study the fitness of adaptive mutations across a range of conditions to characterize fitness pleiotropy. We recently developed a DNA barcoding technology to quantify the fitness effects of ~4,800 independently evolved yeast clones. Each clone contains on average a single adaptive mutation and was isolated from an experimentally evolving population under glucose-limited batch culture conditions. We selected several hundred clones possessing adaptive mutations and performed whole-genome sequencing to comprehensively characterize the genetic basis of adaptation-driving mutations in this low-glucose condition and build a genotype-fitness map. In this work, we used this DNA barcode technology and our previously identified adaptation-driving mutations to study both fitness pleiotropy and understand how these mutations affect cell physiology. We re-measured fitness for all 4,800 independent clones under varying growth conditions, where we systematically varied either the amount of time the cells spend in exponential growth or stationary phase within each batch growth cycle. We found strong evidence of fitness pleiotropy across both sets of experiments, as well as instances of antagonistic pleiotropy where many of our adaptive clones became deleterious if they spend too long in stationary phase between batch culture cycles. In addition, fitness pleiotropy was dependent on both the gene and in some cases the mutation type (i.e. missense, nonsense or frameshift) of the specific adaptive mutation. Our sample size and ability to isolate single adaptive events gives us confidence that all of these results are driven by pleiotropy and not by passenger mutations in these strains. Our findings also showed that all of our adaptive mutations modulated multiple phases of the yeast growth cycle, including the exponential and stationary growth phases. We validated our results using detailed physiological studies measuring the growth of the adapted strains in monoculture and growth when competed against the ancestral strain. Surprisingly, our adaptive mutations appeared to generate antagonistic phenotypic effects even within the exponential growth phase, where many adaptive mutations grew more slowly than the wild-type clone early in exponential growth but grew faster than wild-type late in exponential phase. Our results highlight the complex physiological changes that underlie even single adaptive events, and suggest that the genotype-phenotype-fitness map can be modulated by even slight changes in the environment.