Aerobic anoxygenic phototrophy (AAP) is a metabolic process found in diverse aerobic proteobacteria across aquatic environments. Unlike classical anoxygenic photosynthetic bacteria, the bacteria that perform AAP are often obligate aerobes and are thought to use this pathway to supplement their primarily heterotrophic metabolism. The environmental and molecular factors that control AAP, however, are poorly understood. Using the model marine organism Erythrobacter longus, and the recently isolated freshwater strain Porphyrobacter LM6, we investigated the metabolic pathways and regulatory mechanisms that interact with AAP. First, we constructed deletion mutants of E. longus for several genes involved in light harvesting and the glyoxylate shunt pathway. By comparing the growth of wild type and mutants we demonstrate that light enhanced the growth of wild-type E. longus on pyruvate, glucose and butyrate minimal medium, but not in rich medium; and that the enhanced growth was product of the absorption of energy from light. We discarded that the glyoxylate shunt as the metabolic pathway responsible for light enhanced growth in E. longus, yet we confirmed that the shunt is the only for acetate metabolism in this strain and possibly other AAP strains from the order Sphingomonadales. Next, we used global transposon mutagenesis to assess gene fitness under varying nutritional conditions in Porphyrobacter LM6. The mutant libraries were grown on two different carbon sources (glucose and butyrate) in two different light regimes: 24h dark, and 14h:10h light:dark cycles. As expected, genes in central carbon metabolism had differential fitness effects in butyrate vs. glucose. Notably, the glyoxylate shunt genes isocitrate lyase and malate synthase, along with the anaplerotic carbon assimilation gene malic enzyme, appear to be important on butyrate but not glucose. We next examined the role of phototrophy. Light provided a growth advantage to wild-type cells grown in glucose but had no effect in butyrate. Consistent with this, genes encoding pigment biosynthesis and photosystem machinery were not important for fitness in butyrate, but had strong fitness effects in cells grown in the light with glucose. We determined that the anapleortic reactions performed by phosphoenolpyruvate carboxylase and phosphoenolpyruvate carboxykinase facilitate the light enhanced growth observed in glucose. Catalase/peroxidase and the oxidative stress response regulator oxyR had strong fitness effects in both butyrate and glucose, implying that ROS are a strong selective pressure for this organism. We also demonstrated that ppsR is a key regulator of phototrophy using a targeted gene knockout. These results suggest that the regulation of carbon metabolism and phototrophy are intertwined, and that, surprisingly, phototrophy is advantageous on glucose but not butyrate. Further, ROS detoxification appears to be a key pathway for survival of AAP bacteria. To further explore the impact of ROS in the physiology of AAP, we used another transposon library in E. longus. The production of bacteriochlorophyll-a (Bchla) in the presence of oxygen produces reactive oxygen species (ROS) that are detrimental for their survival in the environment. Yet, the mechanisms used by these bacteria to regulate phototrophic metabolism and overcome the effects of oxidative stress are not fully understood. As expected, we found that superoxide dismutase and catalase are important enzymes against reactive oxygen species (ROS). Mutants deficient in carotenoid biosynthesis also had low fitness under increased oxidative stress, confirming their photoprotective role in AAP bacteria. Glutathione-based systems for repairing ROS damage are vital for the survival of E. longus, as the enzymes glutathione synthase and glutathione peroxidase are required for growth. Mutants of the transcriptional regulator oxyR presented some of the lowest fitness suggesting its role as major regulator in response to oxidative stress. The mutants of catalase and glutathione reductases showed similar fitness patterns to oxyR regulon suggesting that these could be part of its regulon in E. longus. Taken together, our results demonstrate that E. longus, and likely other AAP strains, use a combination of enzymatic mechanisms and photoprotective carotenoids against reactive oxygen species (ROS). Together, our genetics and physiology results shed new light on a widespread and ecologically important metabolic strategy in aquatic systems.