This paper conducts a comparative study of stochastic optimization strategies to enable multi-robot systems to search for and track an unknown number of targets. Each robot is equipped with a noisy sensor that has a limited Field of View (FoV). The robots use a distributed version of the Probability Hypothesis Density (PHD) filter to estimate both the number and states of the targets. This online target estimate is used by the different search strategies to select actions for each robot. We compare Lloyd's algorithm, a traditional method for distributed search, with four stochastic optimization techniques: Particle Swarm Optimization (PSO), Simulated Annealing (SA), Ant Colony Optimization (ACO), and Artificial Immune Systems (AIS). Each of the methods is adapted from the traditional case of finding a single global optimum to locate all the local maxima (\ie targets). We demonstrate through extensive simulations that these techniques offer superior coverage of the search area and more accurate target localization compared to the baseline Lloyd's algorithm. We also discuss the strengths and limitations of each method, assisting practitioners in the selection of the most appropriate strategy based on specific operational factors, such as the communication load in a multi-robot system.