Autonomous systems are increasingly deployed in safety- and mission-critical domains where humans and robots must operate as a team to complete complex tasks. Existing requirements for Human-Robot teamwork remain fragmented across disparate sources, with no unified framework that addresses complexities of collaborative human-robot tasks. We address this gap by presenting a taxonomy of Human-Robot teamwork requirements derived from analysis of standards, regulatory guidance and the literature, both academic and industrial. We extracted 361 requirements from 14 cross-domain sources. Through iterative classification and refinement, we develop a two-level hierarchical taxonomy comprising 6 high-level categories and 21 lower-level subcategories that distinguish information provision, relational control, decision support, safety mechanisms, performance monitoring, and foundational system capabilities. Our taxonomy addresses both conceptual clarity (defining what constitutes Human-Robot teamwork requirements) and practical applicability (supporting requirement elicitation in real systems). To assess the validity of this framework, we conduct an expert evaluation with several domain specialists. We further examine how requirements distribute across Human-Led, Robot-Led and Shared operational perspectives, revealing responsibility boundaries that shape safe collaboration. This work provides a structured foundation for requirements engineering practitioners to systematically elicit and specify Human-Robot teamwork capabilities in critical domains, enabling reliable integration of autonomous systems into cyber-physical teams.