UC Berkeley

In this dissertation I investigate the impacts of artificial and natural watershed disconnection on the management and ecology of valued native California fishes. I explore the subject at regional and local scales and pay particular attention to the consequences of river discontinua on juvenile rearing and habitat use patterns for multiple populations of Pacific salmonids with conservation and economic importance. Scientific inquiry in my study systems was initially motivated by the intention to comprehend the potential direct and indirect links between connectivity loss and collapsed hatchery-supported Chinook salmon (Oncorhynchus tshawytscha) commercial and recreational fisheries and recurring widespread animal mass mortality events involving the take of threatened O. mykiss at one of California's few remaining estuary-saltmarsh habitat complexes. My research coincided with a severe and historic drought and the timing offers a special opportunity to better understand the potential effects of climate change on salmon hatchery management and estuarine fish ecology in California.

From the 1940s to the 1970s, salmon hatcheries were constructed in the foothill zones of California's Sacramento and San Joaquin river watersheds in order to mitigate for lost spawning and rearing habitat upstream of large impassable dams. While initially intended to reduce harvesting pressure on wild stocks, hatchery-produced Central Valley fall run Chinook salmon (CVFRCS) now dominate the population complex and comprise the majority of the California and Southern Oregon commercial salmon fishery. The population collapsed from 2006-2009 and the commercial and recreational fisheries were completely closed in 2007 and 2008 for the first time in their 90 year histories. The proximate cause for the collapse was warm water temperatures and low food abundances in the nearshore zone when smolts exited the Golden Gate in 2005. However, the environmental conditions were not unprecedented and, given that hatchery salmon currently sustain the fisheries, a closer inspection of hatchery management practices was warranted. Surprisingly, basic hatchery release data was largely in hard copy form at the time of the collapse and, thus, largely inaccessible to scientists until June 2015 when my second chapter and accompanying electronic database were published in the San Francisco Estuary and Watershed Science online journal.

In chapter two I synthesize information from 139 state hatchery annual reports and pre-existing state and federal electronic records and present an analysis of temporal trends for the total number of CVFRCS released and release timing, size, and location from 1946-2012. I find that over two billion juvenile CVFRCS salmon were released during the 67-year period and increased optimization of incubation and rearing hatchery 'habitats' over time permitted distinct shifts in management practices. Individuals have been trucked and planted off-site in the San Francisco Estuary (SFE) with increasing frequency since the early 1980s in a literal effort to circumvent poor rearing and survival conditions in the Sacramento and San Joaquin rivers and delta. Since 2000-2001 there has been a dramatic shift towards planting fish in very large quantities at similar times, sizes, and locations. Indeed, approximately half or more of the total number of fish currently released annually from all hatcheries combined are springtime releases of "advanced smolts" in the SFE. This rapidly growing life history type was not observed before the mid-1980s and advancements in medical, feed, filtration, circulation, heating, aeration, and other technologies help facilitate its production. Borrowing from economic theory, my results indicate that CVFRCS population resiliency may have been compromised in recent years by reduced diversity of release life history types (akin to "assets" in a financial portfolio). Continued restoration of inland habitats in conjunction with a re-diversification of release phenotypes would enhance system biocomplexity and theoretically reduce risks to the fishing industry by providing more stable supplies of adult fish for market and recreational purposes.

Intermittent estuaries (IEs) are a special estuarine type characterized by the temporary formation of a coastal sandbar at the river mouth which produces shifts between fully open (complete tidal exchange), partially open (muted tidal exchange), and closed (no tidal exchange) ecosystem states. These highly variable and complicated ecosystems are commonly encountered in small to medium sized watersheds in regions possessing wet/dry seasonality and energetic wave climates such as those found in South Africa, Australia, and North America's Pacific Coast. During the closed state, trapped salts combined with the loss of powerful tidal mixing forces produces density-stratification in the water column and increased hydraulic residence times. The enhanced benthic-pelagic coupling processes can deplete dissolved oxygen levels in the estuary and, depending on season, lagoon waters can either be much warmer or cooler than adjacent ocean waters. Water volume generally increases during the closed state, however, and there is a pressing need to better understand how fishes respond to water quality and habitat quantity tradeoffs, especially during drought periods when sandbar-closed conditions persist for longer durations.

Longitudinal studies that determine fish assemblage and population changes over time in response to drought-induced sandbar regime shifts are especially lacking in the published literature and assemblage studies in California IEs north of Point Conception, where a biogeographical split exists, are nearly nonexistent. In chapter three I present fisheries data from 27 months of continuous mid-month sampling at a large Central California IE (Pescadero Creek Estuary, San Mateo County, CA, USA) experiencing regular breach-induced fish kills caused by widespread anoxia. My sampling commenced three months before the official onset of drought and conditions become progressively drier over time. I find that the Pescadero IE fish assemblage exhibits marked seasonality and is characterized by low species diversity and high species dominance. After 19 months of drought the system shifted from one open or partially open for approximately 75% of the year to one closed for 75% of the time or more. Small-sized estuarine residents possessing rapid life cycles and protracted repeat spawning life histories appear to benefit most from prolonged sandbar closure. Not surprisingly, marine migrant species that utilize the IE for rearing, spawning, predator refuge, or a combination of these factors were most negatively impacted by marine disconnection. Standardized catches of O. mykiss remained elevated during summer months regardless of ecosystem condition. An acute breach-induced fish kill on 11-November, 2011 terminated the steelhead rearing period and produced high levels of species turnover. However, the assemblage rebounded to normal levels two to three months post-kill once water quality in the estuary was restored by repeated tidal exchanges. Overall, I observe strong regularity of seasonal icthyoassemblage shifts despite dramatic changes to the estuarine habitat over time. This observation is almost certainly due to specialized physiological adaptations of commonly encountered fishes to widely fluctuating environmental conditions.

Oncorhynchus mykiss was the most dominant species sampled overall and, while summertime standardized catch rates for O. mykiss remained elevated during all three years, the somatic growth rate analyses I present in chapter four suggests that juveniles likely experienced sublethal stress after one year of drought. Through the combined use of mark-recapture methods and a newly presented scale pattern microanalysis technique, I demonstrate that juvenile O. mykiss summertime body growth rates were fastest when the Pescadero IE was fully or partially open in 2011 and 2012 and slowest when it was completely closed in 2013. While growth of the 2013 cohort was probably limited in part by density-dependent effects, the lagoon nevertheless provided abundant aquatic habitat during a period when upstream habitats were heavily degraded. The novel scale method offers an appealing non-lethal alternative to other biochronological techniques and, thus, is especially suited for conservation-listed fishes. The method has most utility for detecting broad-scale differences in individual growth patterns and life history trajectories and may serve as a useful bioassessment procedure. My analysis reveals that the "twice smolting" phenotype is likely a popular and successful life history strategy for Pescadero steelhead as a result of rapid estuarine growth rates and migration opportunities during favorable environmental conditions.

In chapter five I utilize recent advances in sophisticated acoustic telemetry technology in order to elucidate fine-scale microhabitat use and movement patterns in the Pescadero IE for multiple steelhead from the 2013 rearing cohort. I observe that vertical and lateral habitat compression is correlated to degraded water quality both at depth and upstream in the estuary. The fish exhibited strict preferences for shallow (<1.5 m) and protected microhabitats where the sandy substrate/food production zone occurred within the lighted and oxygenated freshwater epilimnion. Fish use of the upper estuary declined linearly when mean daily water temperatures surpassed 18.0ºC and dissolved oxygen concentrations declined below 7.0 ppm. The concentration of individuals in restricted zones of suitable water quality likely contributes to the density-dependent growth effects observed in chapter four. Efforts that alleviate anoxia and hypoxia in lagoon waters should increase the total amount of aquatic habitat available to steelhead and other fishes in California IEs and increase fish production. Strategic addition of complex large woody debris structure near preferred microhabitats could improve habitat conditions immediately. Efforts that prevent breach-induced fish kills from occurring until early winter when double smolts have moved upstream would maintain life history diversity and, thus, enhance population resiliency.

In summary, by synthesizing and analyzing pre-existing and field-obtained long-term data sets, I am able to elucidate important aspects of the historical, community, population, and behavioral ecology of native California fishes with high economic and conservation value. The analyses presented in this dissertation can help inform management decisions during an anticipated period of rapid and extreme environmental change.