When autonomous surface vessels conduct systematic seafloor surveying using a swath sonar system, the fixed angular coverage leads to differing horizontal coverage width with depth. Traditionally, missions have been pre-planned using fixed line spacing accounting for the minimum expected depth or with the intention of leaving gaps in some areas. A path planning algorithm and accompanying behaviors have been developed to adaptively survey a region for complete coverage with minimal gaps. Subsequent paths are planned based on measured depths in the field and this process continues until a defined area is completed or a depth threshold reached. The algorithm has been simulated in a custom environment and integrated into the MOOS-IvP marine autonomy system.