Cattle also went into the bottom land to get water, especially during droughts when large holes were dug into the riverbed to provide livestock and people with access to the perennial groundwater for drinking . Reclamation projects stimulated by new state laws promoting land use converted marshy areas around valley’s sloughs into agricultural land during 1850-1870 . According to records from American land surveyors and the British naturalist William Brewer who traveled the river corridor in the mid-1850s and early 1860s , the Salinas was a desolate, dry bed of wind-blown sand and treacherous quicksand at that time, yet newspapers also reported on devastating floods, such as in 1852 when the Salinas plain was inundated. Overall, a fundamental transformation in the landscape was wrought by Europeans, even as regional climate remained the same with no long term trend throughout that period , although the characteristic feature of the climate has been frequent recurrences of droughts and floods . Similar to the observed threshold change in sedimentary response to post-European land transformation in the Chesapeake Bay region , the pre-European Salinas landscape was likely resilient to natural climatic events , even with Ohlone fire practices, but after European land transformation the landscape became highly sensitive to perturbation.Today, the Salinas River is heavily changed due to modern land use and flow regulation. Excessive surface irrigation and the regional drought of 1880 were the major stimuli for the onset of groundwater pumping , fueled by the wood from the valley’s remaining trees. By 1901 pumping was well underway, with wells drawing water from as deep as 75 m below the ground surface and lowering the water table below the ground by 3-5 m . Agriculture is now the largest anthropogenic disturbance in the watershed in terms of area,blueberries in containers followed by urbanization and dam emplacement .
Three major dams were constructed on the mainstem and two major eastern tributaries from 1941–1965, impounding the runoff of some 1,970 km2 , or about 17% of the total watershed, primarily in the mountainous subbasins in the wetter western mountains . Urbanization has increased significantly in the basin over the past century but represented only ~2% of land area by 2010.The Salinas River watershed of central California was used as the test case for this study because of the preponderance of data documenting diverse hydrologic events in the watershed, and its history as a preeminent test bed for watershed-scale sediment flux over the last few decades. This work provided the initial forensic setting from which to explore the roles of wildfire and agricultural activity in a semi-arid, mountainous watershed. Previous studies found evidence for the primary role of the Salinas River in supplying sand to Monterey Bay and the significant reduction of coarse sediment export to the coast due to damming on the Salinas and neighboring Rivers . External controls on decadal to inter-decadal scale fluvial sediment flux patterns have been further investigated in terms of El Niño Southern Oscillation cycles . Integrated expressions of external and internal factors in the form of antecedent hydrologic conditions that are affected by ENSO have also been examined as controls on suspended sediment discharge regimes . One previous study has also addressed the importance of wildfire in controlling the sediment export from a major Salinas tributary . Inman and Jenkins conducted a regional scale study on suspended sediment flux from central and southern California coastal rivers with a focus on episodic events and their relationship to regional climate cycles.
They found that large events dominated sediment transfer from the rivers in this region, including the Salinas, and that decadal scale wet and dry cycles lead to concomitant increases and decreases in suspended sediment flux to the ocean. Their approach to calculating suspended sediment discharge through the lower Salinas utilized a rating curve constructed from data collected by the USGS gauge station near Spreckels, CA during water years 1969-1979, which they applied to monthly averages of daily water discharge from 1944-1995, resulting in an estimated average annual suspended sediment discharge of 1.7 Mt yr-1 . Farnsworth and Milliman also examined the role of large discharge events in the estimation of total suspended sediment load at gauge S1, and used the same set of S1 USGS data to compute a power law rating curve that was then applied to daily water discharge data from 1930-2000 for an average annual suspended sediment discharge of 3.3 Mt yr-1 . Gray et al.thoroughly investigated suspended sediment dynamics in the Salinas River in relation to climatic and hydrological drivers. It was found that CSS-Q behavior was influenced by antecedent hydrologic conditions at event, annual, interannual anddecadal time scales , and that the temporal trend in discharge-corrected suspended sediment concentrations was negative over the 1967-2011 period of record . Notably, no change in the relationship between precipitation and discharge was found over this time period . By taking these factors and the temporal dependence of sediment behavior into consideration, the average annual suspended sediment flux was estimated as 2.1 to 2.4 Mt . The Arroyo Seco drains ~ 700 km2 of the wetter western mountains in the Salinas River watershed. The majority of the Arroyo Seco remains largely undeveloped and is undammed, unlike the other major Salinas River subbasins in this region. For this reason, the Arroyo Seco is a significant contributor of water and sediment to the Salinas River .
Warrick et al.found that sediment export from the Arroyo Seco was highly controlled by the sequence of wildfire and subsequent high precipitation/discharge events. Nearly complete burning of the upper Arroyo Seco watershed in the summers of 1977 and 2008 facilitated a pairwise comparison of post fire sediment flux past gauge A3 , in relation to storm timing and intensity differences . A rare , high intensity wet season during the 1978 water year led to an annual sediment flux ~ 100x pre-fire, while moderate precipitation in 2009 and 2010 resulted in only 5x and 9x increases, respectively. Warrick et al.estimated that the Arroyo Seco’s post-fire sediment flux may have caused the total Salinas River sediment to export to double in 1978.Discharge and a number of fluvial constituents, including suspended sediments, are monitored in the lower Salinas River at S1 and S2 . This study was based on 286 suspended sediment samples collected by the USGS between 1967 and 2010 , and 44 by the authors between 2008 and 2011 from gauging stations S1 and S2 and the Davis Street bridge crossing 4 km downstream from S1 . Potential biasing of the suspended sediment data set in terms of hydrologic regime representation and temporal distribution was examined and discounted in Gray et al. . Paired samples were collected by the authors from the water surface at cross-channel stations of one-quarter, one-half, and three-quarters wetted channel width , which were then processed for CSS and particle size distribution as per Gray et al. . Suspended sediment samples for this study were collected from the surface of river flow. For this reason coarse suspended sediment particles were expected to be underrepresented. Geographical and temporal wildfire data recorded by the California Department of Forestry and Fire Protection Fire Resource Assessment Program from 1911- 2010 were clipped to the geographic extent of the Salinas watershed . Contributing areas behind dams were masked for the time periods of their operation,planting blueberries in pots as the trapping efficiency of these dams was previously estimated as > 90% for fine sediment . The areal extent of fires in undammed portions of the watershed was then summed by year for further computations of effective burn area . The lasting, time-dependent effect of a given wildfire, or set of wildfires, on the landscape can be modeled as “effective” burn area, which is the initial burn area modified by standard exponential decay functions operating over the elapsed time since the fire . The range of relevant half-life values for the EBA decay function found for semiarid southern California systems is between 0.5–14 years, with 1.4 years identified as the best fit for the Arroyo Seco . In this study the EBA approach was applied to annual burn area data with a range of t½ values from 0.5–10 years in 0.5 year steps. As most Salinas fires occur during hot, dry summers, the EBA associated with a given CSS sample was calculated by summing of effective burn area contributions from all previous years, with the year before the water year of the sample treated as t = 0, while the annual burn area from the water year of the sample was excluded.
Spatial coverage of crops by year from 1965-2011 was obtained from the Monterey County Agricultural Commissioner’s Office Crop Reports and sorted into ‘field’ and ‘row’ crops . Data on the areal coverage agriculture by irrigation technique from 1993-2010 and urbanization from 1984-2010 were extracted from Monterey County, 2015b.Testing for the control of wildfire and agricultural activity on sediment production in the lower Salinas River was based on comparing the temporal trends of LOESS residuals for both fines and sand with the temporal trends of metrics for effective wildfire burn area, bulk and crop specific agricultural areas, and areas under given irrigation types. Temporal trend analysis was conducted with Mann-Kendall and linear regression. It was hypothesized that the decreasing CSS–Q relationships found in the lower Salinas were caused by one or a combination of the following changes in sediment controls over the sample period: decreased wildfire activity, decreased agricultural land area, changes in agricultural composition to less erosive crops, or changes to less erosive irrigation techniques. All data sets were examined for temporal trends in light of the 1967-2011 base period of CSS data, although the irrigation technique data set did not begin until 1993. Factors with statistically insignificant temporal trends and/or trend directions opposing those expected in light of decreasing CSS–Q were eliminated as potential controls. A correlation test was also performed between LOESS residuals and the wildfire burn area metric EBA, as short term responses in basin scale sediment production would be expected from wildfire disturbances. A significant positive correlation between the wildfire and suspended sediment magnitudes would be considered indicative of wildfire as a potential dominant control on decadal scale trends in sediment production, since wildfire has been generally found to increase the production of sediment to a greater degree than water in the steep, brush dominated environments typical of the primary source areas in the Salinas River . Correlation tests were not performed on the agricultural metrics as they were expected to only produce decadal to inter-decadal sediment production responses rather than abrupt shifts in sediment supply dynamics due to the slow rate of change of these factors. As Warrick et al.had found that wildfire affected ~ 100 times more sediment yield from the Arroyo Seco subbasin when followed by a winter of intense storm events, EBA was also examined in concert with peak daily Q for each year. The values of LOESS residuals for years with high EBA values were then compared with consideration given to the peak daily Q experienced by these years, to determine if the convolution of wildfire and subsequent large storms was responsible for temporal patterns of suspended sediment behavior at the annual scale.Beginning in the mid-1960s dry field products like barley and animal feed declined and root products such as sugar beets and potatoes all but disappeared . Intensively irrigated row crops and utilization of land for multiple cropping seasons expanded, including rapid increases in leaf lettuce beginning in the early 1980s, while grape production expanded rapidly from 6–141 km2 between 1971 and 1974, mostly on converted field crop and grazing lands. As a result of these crop changes and the limited use of efficient drip irrigation , groundwater withdrawals are estimated to have increased during the first half of the suspended sediment record . Monterey County, which is mostly contiguous with the Salinas watershed, began recording groundwater withdrawals in 1993 . Between 1993 and 2010 total ground water extraction actually decreased from 0.62 to 0.57 km3 , although irrigated agricultural land area increased from 702 to 732 km2 and total crop area increased from 1270 to 1588 km2 .