Stream networks are spatially and temporally dynamic meta-ecosystems influenced by local and regional environmental conditions. Non-perennial streams are globally prevalent and experience a wide range of hydrologic conditions from expansion (flooding) to contraction (loss of surface flow). Drying patterns are spatio-temporally variable, influencing dynamics in stream meta-ecosystems including resource transport. Essential for many biogeochemical processes and biological communities, dissolved oxygen (DO) follows daily and seasonal cycles that can be directly linked to hydrologic connectivity. To predict how biogeochemical processes and communities respond to changes in drying and flowing conditions, we need to assess how hydrologic connectivity across the network alters the exchange of resources and organisms through non-perennial streams. Here we use data collected from stream temperature, intermittency & conductivity (STIC) sensors placed throughout the network as well as water level and DO sensors at the outlet of three different watersheds across continental United States to assess how network contraction 1) differ among watersheds from different regions and 2) drive whole-ecosystem responses. Using our sensor network data and directed acyclic graph theoretic approaches, we calculated summative network connectivity metrics (e.g., active network length, global centrality) to characterize how three non-perennial stream networks behave during a drying event. Except in our spring-dominated system, where we observed frequent net losing conditions in the mainstem, headwaters tended to dry first. Each network dried at different rates, with some drying slower and others reaching a threshold and drying quickly after, directly altering DO dynamics. At time periods where there is a lower proportion of the network flowing, we generally see lower daily minimum DO concentrations and increased daily DO amplitude, although these relationships vary through time and across the watersheds. Linking stream network connectivity and ecosystem structure and function is essential for our global understanding of how drying alters stream meta-ecosystems.