Along California's North Coast, where salmon hover on the cusp of extinction, scientists and local residents seek new collaborations. Agencies, tribes, and watershed councils commission competing studies to determine links between human water use, oceanic cycles, and salmon decline. Modelers turn to ranchers' expert opinions to condition hydrologic models. Ranchers import beavers to build dams that may raise the water table. These watershed collaborations begin to transcend boundaries of human institutions, scientist / lay person, and even species. Restoring salmon-bearing streams is a project to reconfigure human relationships to water and inhabitation practices. Ininfrastructures.the western U.S., this project necessarily entails a serious grappling with ManifestDestiny legacies of Native American sovereignty, property regimes, legal doctrines, and waterMy dissertation investigates how watershed collaborations transform scientific practices,environmental subjectivities, and trans-species relations, using Salmon Creek (Sonoma Co., CA) as acase. Salmon Creek is typical of thousands of small watersheds in the Pacific West in that summerwater extractions by farmers and rural residents dry many tributaries into a series of disconnectedpools. This anthropogenic drought compounds historic beaver removal, logging, and road buildingthat have altered water, sediment, and large wood supply to the stream, limiting steelhead(Oncorhynchus mykiss) and coho salmon (O. kisutch) recovery. I argue that collaborative watershedresearch that refuses to privilege expert science over local and Indigenous knowledges can createnovel modes of scientific practice, discursive shifts, and new governance approaches. I stretch thelimits of the terms 'watershed' and 'collaboration' to encompass interactions among (1) scientistsand local knowledge holders, (2) living species and the landscapes they inhabit, and (3) humans andother species that depend on riverine ecosystems. Though disparate in methodology, the fields-Science and Technology Studies, Environmental Politics, and Eco-hydrology-share a commitment to critically re-working social-natural boundaries.Natural flow regimes-dynamic streamflow patterns that drive riverine biodiversity-arisefrom a kind of collaboration between climatic factors, geology, plants, and animals in a river basin,and are then further shaped by human ground and surface water diversions. Regarding theecosystem as a collaboratory in which humans play a role, Quantifying abiotic habitat characteristicsto determine thresholds for salmonid oversummer survival in intermittent streams investigates therole of different flow-mediated factors (dissolved oxygen, temperature, groundwater inflow, andpool volume) affect juvenile coho and steelhead occurrence in two Salmon Creek tributaries.Drawing on three years of juvenile fish surveys, synoptic water and isotope monitoring andstreamflow gauging to populate statistical models, I found that low dissolved oxygen and poolvolume limit survival; however both salmonid species can survive in spring-fed intermittent poolsthat contain sheltering logs or overhanging banks. Citizen science surveys of stream drying patternsand salmon occurrence can complement agency monitoring and should be incorporated into salmonrecovery efforts.Regarding human collaboratives of knowledge and practice, 'Thinking with salmon aboutrain tanks: stream commons as intra-actions' puts forth the argument that cultural practices of wateruse evolve in response to new understandings of other species' dependence on shared streams.Some Salmon Creek residents who install rain cisterns to curtail summer water use do so out ofconcern for salmon, and describe salmon and other riverine creatures as having rights to enoughwater to survive that are of the same status as human rights to water. Other residents are unwillingto reduce water use because the connection between their wells and the stream are poorlyunderstood and difficult to measure.dissertation contributes tothis'Rain tanks, springs, and broken pipes as emerging watercommons along Salmon Creek, CA, USA'. Residents who participate in monitoring salmon populations, water quality, and their own springs and rain tanks report that these activities have increased their sense of interdependence with other human and nonhuman neighborswho rely on the watershed's limited water sources. Drawing on Barad's (2007) concepts of apparatus and intra-action, I argue that the notion of water as an interspecies commons is co- evolving with rainwater harvesting and that collective choice frameworks that embrace both management practices and environmental imaginaries represent a coherent alternative both to state and market frameworks of water governance and to traditional adaptive management methods and discourses.foment discourses that bring humans and other species - especially beaver and salmon, which affect water and nutrient cycles and thus are considered"ecosystem engineers"- into symbiotic relations with mutual responsibilities. In the conclusion, I explore how thesepractices and local knowledge of springs, aquifers, and rainfalldevelops a method for studying up from household waterMobilizing approaches from feminist Science and Technology Studies, the introductionextends the idea of watershed collaborations to encompass humans and other species. I draw onextended interviews with scientists, policy-makers, and local residents to argue that members ofknowledge practice collaborativesconcepts of multi-species collaboratives may filter up from local collaborationsinto public water and species recovery debates, and consider limitations to more entangledapproaches to watershed governance.Salmon Creek is geographically small and removed from major river basins, yet functions asa kind of microcosm of the political, cultural, and ecological tempests these salmon recovery andecological restoration projects stir up at any scale. Salmon are simultaneously a global fisheryresource, a key subsistence and cultural resource for traditional peoples around the Pacific, ascientific project to avert extinction, and a contested site of knowledge production. In asking whatforms of collaboration are productive, and how collaborations transform those who undertake them,this research contributes to debates on practice and ethics inherent in environmental governance inthe Anthropocene era.