Field Methods
Larval sturgeon were sampled by deploying appropriate sampling gear in non-random locations along the San Joaquin River between April and July. Targeted sampling locations were selected that were near areas where aggregations of sexually mature sturgeon had been found. Gear types included:
1) Benthic D-nets deployed from a boat (D_net, n = 313 sets)
2) Benthic D-nets deployed from a rotary screw trap frame (D_frame, n = 227 sets)
3) Custom double-frame nets anchored to the bottom (Double_frame, n = 120 sets)
4) Large drift nets deployed from an rotary screw trap frame (Pontoon_net, n = 48)
5) Otter trawls (Otter_trawl, n = 77)
D-nets
Benthic D-nets were used to sample for larvae in the San Joaquin River. These nets and deployments were similar in design to nets that have been successful in capturing larval White Sturgeon and Green Sturgeon on the Sacramento River (Kohlhorst 1976; Brown 2007; Poytress et al. 2009). The benthic D-net was 0.8-m wide by 0.6-m tall. Steel bar stock was added to the bottom of the net frame to ensure that the frame oriented properly and remained in contact with the river bottom during sampling. The net was made of 3.2-mm nylon mesh for the entire 3.0-m net length and an inner layer of 1.6-mm mesh that was sewn into the last 46-cm of the net. The length was tapered to fit a PVC sample bucket in the cod end (16.5-cm diameter, 46-cm length). Site selections for larval sampling were determined based on egg collection locations during previous years.
In 2013, sampling was from a moored boat (D-net in Gear column of SJR_Larval_WST_Set.csv). The survey vessel was moored using Columbia River-style anchors and oriented into the current. The D-net was deployed by a 25-m length of 8-mm diameter Amsteel® rope attached to the top of the frame and the net was deployed and retrieved with a custom-fabricated aluminum boom and anchor winch. Designated sampling times were between 2000 and 0100 hours. These hours were selected because recent work with larval Green Sturgeon on the Sacramento River indicated that migration activity is highest during night-time hours (Poytress et al. 2009). Sets were for approximately 30 min.
In 2015, 2016, and 2017 (no larval sampling occurred in 2014), D-nets were deployed from a rotary screw trap frame moored at two anchor locations (D-frame in Gear column of SJR_Larval_WST_Set.csv). One mooring was a steel cable attached to a tree on shore and the other was a Columbian River-style anchor deployed in-channel. D-nets were deployed at approximately the same time, one off each side of the screw trap frame, using a pulley system attached to a custom davit. In 2015, nets were fished approximately 12 hours per set. Debris load was an important factor in determining length of set-time, so in 2016 and 2017, nets were set for approximately 30-min intervals during higher debris loads and 60-min intervals at lower debris loads.
Environmental data were collected during the setting and retrieval of the net. Prior to the first nightly net deployment GPS coordinates and turbidity were measured. During each tow, depth was measured with a stadia rod and flow was measured with a General Oceanic® Model 2030 flow meter that sampled flow at the center point of the net mouth.
Double-frame Nets
Custom made double-frame nets (Double_frame in Gear column of SJR_Larval_WST_Set.csv) were deployed to survey for the presence of White Sturgeon larvae from April 5 to June 10, 2016. Double-frames were modified from the triple-frame drift nets used by the Confederated Tribes of the Colville Reservation on the Upper Columbia River (Howell and McLellan 2014 ; Matt Howell, Coleville Confederated Tribes, personal communication). We constructed the frames from welded 2.5-cm square steel tubing with two 0.6 x 0.8-m openings to support two 4.0-m drift nets with 1.6-mm mesh. At the downstream end of each drift net, we connected a 22.7 l collection bucket with seven columns of 3.2-mm holes drilled along the sides to mitigate obstruction from heavy debris loads. Double-frames were anchored within the river using a bridle connected to a Columbia River-style anchor deployed 10-15 m upstream.
Double-frames were deployed in the evening (1700–2000 hours) and retrieved 12–15 hours later (0700–1000). Depth (m), water temperature (°C), and GPS coordinates were recorded during deployment and retrieval. Upon retrieval, collection buckets were removed and contents were rinsed into a 20-L bucket for transport to the laboratory. At the laboratory, samples were filtered and sorted for fish eggs, larvae, and invertebrates. Captured fish and invertebrates were stored in a 95% ethanol solution for later identification.
Large Drift Nets (Pontoon Nets)
In 2015, a second, large drift net (Pontoon_net in Gear column of SJR_Larval_WST_Set.csv) was also deployed from the rotary screw trap frame. The large drift net consisted of a rectangular frame (2.0-m wide by 1.0-m high) attached to a net with 4.8 mm mesh, designed and constructed by Red Bluff Fish and Wildlife Office in Red Bluff, CA. The mesh net funneled into a stainless steel live-box with 3.2 mm holes. The large drift net was set and retrieved using a steel cable and a winch system.
Preliminary deployment of the large drift net, conducted May 4 to May 6, occurred between 1900 and 0000 hours with net retrieval every 30 min. On May 6, catch and debris load were determined to be extremely low allowing us to adjust sampling techniques. At this time, the sampling deployments were altered to two 12-hr sets with gear retrieval at 0700 and 1900 to capture both a day and night drift.
Otter Trawls
Modified otter trawls (Otter_trawl in Gear column of SJR_Larval_WST_Set.csv) were used to survey for White Sturgeon larvae from July 6 to July 22, 2016. The modified trawl design is used throughout the Mississippi River Basin to sample for small-bodied fishes, including juvenile sturgeon (Herzog et al. 2009). The entire trawl body is made of 3.2-mm heavy delta style mesh (Herzog et al. 2005), while cod ends have a 17.5-mm inner mesh allowing smaller fish to separate from larger fish and debris to reduce sampling mortality.
Trawl tows occurred during evening hours (2000–0000 hours). The net was set off the bow of the boat with 11–15 m of rope attached to the boat, depending on depth, and towed in reverse downstream between for 1–6 minutes. Trawl locations generally consisted of long stretches free of snags, leading into or out of deep pools. Start time was recorded when ropes were pulled tight and sampled up to five minutes or until conditions (e.g., snags, boaters) forced sampling to cease.
Environmental Data
Sample effort and environmental data were collected in conjunction with larval sampling. Sample effort data consisted of the starting and ending dates, times, and latitude and longitude coordinates for individual sets. GPS and sonar depth readings were obtained using a Lowrance side-scanning sonar unit (Lowrance HDS-10 Gen2) and supplemented with GPS coordinates from a handheld unit (Garmin Etrex 10). Environmental data collected typically included surface measurements of starting and ending water temperature (°C; YSI Model 55 or Lowrance Structure Scan HDS 10), dissolved oxygen (mg/l; YSI Model 55), and turbidity (NTU; Model 2100P Hach turbidimeter).
Fish Processing
Larval fish were stored in 95% ethanol and transported back to the laboratory for identification. Preserved fish were processed using a dissecting microscope (Olympus® Model SZX2-ILLK). We measured most identifiable and intact fish to FL in mm and identified to species, except small centrarchids and gobiids, which were identified to family.
