**1. Introduction**

Populated estuarine regions worldwide have been subject to a variety of stressors, including the introduction of invasive species, loss of tidal habitat, anthropogenic alterations to the natural hydrologic cycle (including freshwater diversions), impacts to sediment transport resulting from upstream watershed land use modifications, and other water quality impairments [1]. These stressors can adversely affect the estuarine habitat for resident and anadromous aquatic species. Today, there is growing interest in many parts of the world to restore estuaries to more pre-development or natural conditions [2–5]. Although restoration planning must account for multiple interacting stressors, for estuaries subjected to significant hydrologic alterations, restoration of a more natural hydrology and salinity regime is key. To support such restoration planning, pre-development reference

**Citation:** Hutton, P.H.; Meko, D.M.; Roy, S.B. Supporting Restoration Decisions through Integration of Tree-Ring and Modeling Data: Reconstructing Flow and Salinity in the San Francisco Estuary over the Past Millennium. *Water* **2021**, *13*, 2139. https://doi.org/10.3390/w13152139

Academic Editor: Nigel W.T. Quinn

Received: 2 June 2021 Accepted: 27 July 2021 Published: 3 August 2021

**Publisher's Note:** MDPI stays neutral with regard to jurisdictional claims in published maps and institutional affiliations.

**Copyright:** © 2021 by the authors. Licensee MDPI, Basel, Switzerland. This article is an open access article distributed under the terms and conditions of the Creative Commons Attribution (CC BY) license (https:// creativecommons.org/licenses/by/ 4.0/).

conditions may need to be defined, although no formal methodology is proposed in current U.S. regulations. Directly observed data representing reference conditions in a developed estuary are difficult to obtain, especially when the development has occurred over centuries. However, some pre-development characteristics can be inferred from proxy data, notably estimates of precipitation in the estuary watershed through tree-ring measurements of long-lived tree species.

This work seeks to support restoration planning in the San Francisco Estuary, the largest estuary on the Pacific coasts of North and South America, by characterizing the region's pre-development hydrologic and salinity conditions over the past millennium. The estuarine region includes a series of interconnected embayments, rivers, sloughs, marshes as well as the delta formed by the Sacramento and San Joaquin Rivers (hereafter referred to as the "Delta"), which together drain a watershed of 75,000 square miles, more than 40% of the area bounded by the state of California [6,7]. Following European settlement of California in the mid-18th century and the subsequent Gold Rush (circa 1850), the estuary and its watershed have been subject to extensive changes, including land-use conversion to agriculture and urbanization, construction of water storage and diversion facilities on major rivers, channelization and modification of riparian and tidal habitats, and out-ofbasin exports of water [7–9]. The estuary is currently the focus of much scientific attention because of its importance to aquatic ecosystems and because large parts of the state's urban and agricultural economies are dependent on water supplies from the Delta [7,10,11].

Freshwater flow to the estuary (termed "Delta outflow") has been identified as a vital planning component for regional sustainability. Delta outflow and salinity have been managed for several decades through the regulation of upstream reservoirs and out-of-basin exports. Maximum salinity levels are prescribed at various locations in the Delta; the broader salinity regime is regulated as the position of the 2 parts per thousand bottom isohaline from Golden Gate (measured in km), commonly referred to as X2 [12–14]; see Figure 1 for isohaline positions). Despite ongoing regulatory efforts, the abundance of many Delta fish species continues to decline from the first formally recorded levels in the 1960s [7,15–17]. In response to these declines, additional freshwater flow and salinity regulations are being considered for future implementation [18]. An improved understanding of the estuary's hydrology and salinity characteristics prior to development, and differences from contemporary conditions, will support decisions related to its future management.

The broader region delimited by the San Francisco Estuary and its upstream Central Valley watershed benefits from the availability of extensive data to reconstruct past flow and salinity conditions. These data include flow and salinity measurements, over a century or more, that represent the intensification of development in the region (e.g., [14]). These data also include tree-ring measurements to characterize watershed precipitation over the past two millennia (e.g., [19–21]). The specific research objectives of this work are to refine and update tree-ring-based reconstructions of Central Valley runoff over the past millennium and reconstruct Delta outflow and salinity over similar millennial timeframes using our runoff estimates within a modeling framework informed by previously published work. This integrated evaluation provides a time-resolved characterization of the estuary's flow–salinity behavior that allows comparison between pre-development and contemporary conditions.

This work builds on previous research that either (i) relies on a contemporary hydrologic sequence to estimate outflow and salinity changes using different modeled representations of the region's level of development [22] or (ii) relies on contemporary salinity data to estimate salinity changes using a tree-ring based hydrologic sequence [23]. By using a tree-ring based hydrologic sequence in conjunction with a modeling approach that estimates pre-development estuarine flow and salinity responses, this work attempts to represent the actual range of flow and salinity conditions over long time horizons and is expected to better support regulatory decision making by providing a baseline to inform future flow regulations and restoration actions in the estuary. Furthermore, this work places

the wet and dry flow patterns recorded in the estuary over the past 150 years in the context of flow variations estimated over the past millennium from the tree-ring proxy record.

**Figure 1.** Study location map showing the locations of tree ring sites used in the analysis. Circles mark sites contributing to the short (60 sites) and long (13 sites) reconstructions. Circles sized proportional to percentage of variance explained in regression models for single site reconstructions (SSRs). Sites contributing to long reconstruction marked with green; those as well sites marked with red contribute to the short reconstructions. Nine sites (gray) were screened out and not used in later reconstruction steps.
