Field Manual for Sampling Forage Fish Spawn in Intertidal Shore Regions

With the listing of many Puget Sound salmon stocks as threatened or endangered, the issue of maintaining salmon forage fish stocks has been identified as a high priority by the San Juan' County Marine Resources Committee (SJC MRC). All the important forage fishes, i.e. surf smelt, Pacific sand lance, and Pacific herring, depend on nearshore habitats for spawning and rearing. Protection of nearshore habitats utilized as spawning and rearing areas for forage fish will be needed if salmon recovery is to be successful. Recovery of bottomfish, within SJC was also identified in 1996 as a key priority by the SJC MRC. These species have since become a high priority throughout Puget Sound because six stocks have been identified for potential listing as threatened or endangered species. The same forage fish species of interest in salmon recovery will be vital for the success of any program to restore bottomfish stocks.

Washington Department of Fish and Wildlife (WDFW) presently attempts to protect all known, documented Pacific herring, surf smelt, and Pacific sand lance spawning sites from impacts of shoreline development. "No net loss" regulations for the protection of known spawning sites of these species are included in the wording of the Washington Administrative Code ''Hydraulic Code Rules" (WAC 220-110), which are applied by WDFW marine habitat managers during considerations for granting Hydraulic Permits for in-water shoreline development proposals. However, the forage fish habitat protection regulations only apply to shorelines where spawn has actually been detected by WDFW or other qualified surveyors. Thus it is critical for overall protection of these habitats that spawn deposition site inventories be complete and comprehensive. Not all outwardly suitable-appearing shorelines seem to be used by spawning forage fishes. On the other hand, large areas of formerly productive spawning habitat have been degraded or destroyed by shoreline practices in the absence of a database (or concern) regarding forage fish spawning activity.

Surveys to identify spawning areas were conducted by WDFW between 1989 and 1999, which documented 14 surf smelt spawning beaches, and 8 Pacific sand lance spawning beaches (Penttila 1999). WDFW was conducting a systematic survey of forage fish spawning beaches from 1991- 1996 throughout Puget Sound, but lost funding for the effort in 1997, just as the San Juan County beaches were to be surveyed. As a result of the diminished program, only a small portion of the potential beach spawning habitat has been surveyed (Penttila 1999).

Surf smelt in the San Juan area spawn year-round, with no particular spawning season more dominant than another (Penttila 1990, 1999, Figure 1). Eggs, about 1 millimeter in diameter, are deposited in the upper intertidal zone on mixed sand and gravel beaches (Figure 2). After spawning, the eggs are dispersed across the beach by wave activity, so more of the beach is used for incubation than is used for actual spawning. Surf smelt can spawn on the same beach through the year, so eggs are likely to be present at any time. For example, at Hunter Bay and N. Shaw Island index sites, smelt eggs were found during 13 of 16 visits from February 1989 to May 1990 (Penttila 1990).

WDFW conducted field surveys of spawn visible to the eye from 1989 to 1990 and "bulk sampling" (i.e., composited sediment samples from potential spawning beaches) from 1993 to 2000 to identify surf smelt spawning areas within San Juan County. The bulk sampling method consists of collecting beach samples and subjecting the sample to laboratory examination for egg presence. This method is considered a much more accurate measure of spawning activity than the visual method. A total of208 visual samples and 286 bulk samples were taken during the survey periods. Most of the visual surveys were on Orcas, Lopez and Shaw islands, while bulk sampling was primarily on San Juan, Orcas and Lopez islands (Penttila 2000). The distribution of sampling is illustrated in Figure 3. As presented above, fourteen beaches within San Juan County have so far been identified as supporting spawning by surf smelt (Penttila 1999, Figure 4). The visual sampling method is considered relatively inefficient for identifying spawning locations, thus WDFW recommends that the locations surveyed in 1989-1990 that did not yield eggs should be resurveyed using the bulk method (Penttila 2000).

Results of bulk sampling indicate that not all beaches with appropriately-sized sand and gravel are used for spawning. Usage appears greatest on beaches with over-hanging vegetation. Overhanging vegetation provides shade, which reduces egg mortality caused by desiccation. The shading is likely to be particularly important for the portion of the stock that spawns from late spring to early fall, when low tides are during the day and exposure to warm, dry air is greatest.

The intertidal nature of Pacific sand lance spawning was not known until 1989 (Penttila 1999). Pacific sand lance appear to use the same spawning substrate as surf smelt, as eggs from both species are often in the same sample. Pacific sand lance, however, will also use pure sand beaches that are not utilized by surf smelt. Fresh spawn appears as shallow, circular pits on the upper beach (Figure 2). The pits disappear rapidly after spawning as wave action re-works the beach sediment. Spawning by Pacific sand lance is during the winter, from early November through February (Figure 1). Development of the 0.6-0.8 mm eggs takes about 4 weeks, depending on temperature, thus incubating eggs could be present into late March.

The bulk sampling method described for assessing ' surf smelt spawning is also used to document Pacific sand lance spawning. The visual method is virtually useless for detecting Pacific sand lance eggs because these eggs are covered with sand grains and are essentially undetectable with the naked eye. Eight Pacific sand lance spawning areas were found during the bulk sampling conducted from 1993 to 2000, with the distribution as depicted in Figure 5.