Emergence in natural systems is the development of complex behaviors that result from the aggregation of simple agent-to-agent and agent-to-environment interactions. Emergence research intersects with many disciplines such as physics, biology, and ecology and provides a theoretical framework for investigating how order appears to spontaneously arise in complex adaptive systems. In biological systems, emergent behaviors allow simple agents to collectively accomplish multiple tasks in highly dynamic environments; ensuring system survival. These systems all display similar properties: self-organized hierarchies, robustness, adaptability, and decentralized task execution. However, current algorithmic approaches merely present theoretical models without showing how these models actually create hierarchical, emergent systems. To fill this research gap, this dissertation presents an algorithm based on entropy and speciation - defined as morphological or physiological differences in a population - that results in hierarchical emergent phenomena in multi-agent systems. Results show that speciation creates system hierarchies composed of goal-aligned entities, i.e. niches. As niche actions aggregate into more complex behaviors, more levels emerge within the system hierarchy, eventually resulting in a system that can meet multiple tasks and is robust to environmental changes. Speciation provides a powerful tool for creating goal-aligned, decentralized systems that are inherently robust and adaptable, meeting the scalability demands of current, multi-agent system design. Results in base defense, k-n assignment, division of labor and resource competition experiments, show that speciated populations create hierarchical self-organized systems, meet multiple tasks and are more robust to environmental change than non-speciated populations.