NOTES Spawning and Rearing of Splittail in a Model Floodplain Wetland
Total Page：16
File Type：pdf, Size：1020Kb
Transactions of the American Fisheries Society 131:966±974, 2002 q Copyright by the American Fisheries Society 2002 NOTES Spawning and Rearing of Splittail in a Model Floodplain Wetland TED R. SOMMER* California Department of Water Resources, 3251 S Street, Sacramento, California 95816, USA; and Department of Wildlife, Fisheries, and Conservation Biology, University of California, Davis, California 95616, USA LOUISE CONRAD,GAVIN O'LEARY,FREDERICK FEYRER, AND WILLIAM C. HARRELL California Department of Water Resources, 3251 S Street, Sacramento, California 95816, USA Abstract.ÐThe splittail Pogonichthys macrolepidotus, have little or no access to ¯oodplain spawning and which has been listed as threatened by the U.S. govern- rearing habitat. Although the relatively long life ment, does not produce strong year-classes unless it has span (frequently .5 years) and high fecundity of access to the ¯oodplain habitat of the San Francisco estuary and its tributaries. In this small-scale, single- this species helps to buffer the population against year study, we tested the hypothesis that managed in- low-out¯ow conditions, an extended drought dur- undation of a ¯oodplain can be used to support splittail ing the 1980s and early 1990s led to a major de- reproduction in dry years, when this habitat type is not cline in the production of young splittails (Meng readily available. Adult splittails were captured on their and Moyle 1995; Sommer et al. 1997). This decline 2001 upstream spawning migration and transferred to a 0.1-ha model ¯oodplain wetland. Our results suggest in abundance was the primary basis for the U.S. that adults will successfully spawn if they are provided government's listing of the splittail as threatened access to ¯oodplain habitat in dry years. In snorkel sur- in 1999 (USFWS 1999). Major declines have also veys, progeny showed a signi®cant association with the been noted for several other native ®sh in the es- lower portion of the water column. Young splittails (15± tuary, although the causes vary (Bennett and Moy- 20 mm fork length [FL]) concentrated in edge habitat near an in¯ow during the day but at night moved into le 1996). The splittail is the last surviving member deeper-water habitats, including open water and habitats of its genus; the only other species in the genus, with submerged vegetation. Larger splittails (28±34 mm the Clear Lake splittail P. ciscoides, became ex- FL) used a broad range of habitats both during the day tinct sometime during the previous century (Moyle and at night. Juveniles showed signi®cant schooling be- 2002). havior during the day, then dispersed at night. These Several restoration programs are underway to observations have potential implications for the design of habitat restoration projects for the splittail, the last increase the ®sh populations of the estuary and its remaining representative of its genus. tributaries (Yoshiyama et al. 2000). Floodplain restoration has been identi®ed as a potential way to support splittail and other species (CALFED The splittail Pogonichthys macrolepidotus, a na- 2000). One major restoration goal for the San Fran- tive cyprinid, is perhaps the most ¯oodplain-de- cisco estuary is to improve the connectivity be- pendent ®sh in the San Francisco estuary (Figure tween river and ¯oodplain habitat, particularly in 1; Sommer et al. 2001a). During high-¯ow periods the Yolo Bypass, the largest remaining ¯oodplain. in winter and spring, adult splittails migrate up- stream into the Sacramento2San Joaquin delta and As a result of the system of levees and weirs con- its tributaries (Daniels and Moyle 1983), where structed around its perimeter, the Yolo Bypass typ- spawning activity is apparently concentrated on ically ¯oods only in above-normal water years the seasonal ¯oodplain (Sommer et al. 1997). (Sommer et al. 2001a). Here, we use a model ¯ood- Abundance is reduced in dry years, when splittails plain wetland to test the hypothesis that managed inundation of ¯oodplain habitat during lower-¯ow years can be used to support splittail spawning and * Corresponding author: [email protected] rearing. An additional objective was to provide Received September 4, 2001; accepted February 4, 2002 diel observations of juvenile splittails; the habitat 966 NOTES 967 FIGURE 1.ÐMap showing the location of the Yolo Bypass (dark gray area). The San Francisco estuary includes the region from San Francisco Bay upstream to Sacramento. The Yolo Basin Wildlife Area is the light gray area within Yolo Bypass immediately southeast of the study site. associations of this life stage are poorly under- under natural conditions. Moreover, diel behavior stood because early juveniles typically occur dur- has not been well studied for juvenile cyprinids ing high-¯ow events, when high turbidity and ex- (Garner 1996). This information is essential for treme environmental ¯uctuations create major the successful design and evaluation of splittail sampling problems. Young and Cech (1996) have restoration projects. described the physiological tolerances and require- Our basic approach was to capture adult split- ments of young splittails in laboratory studies. tails on their seasonal spawning migration during However, little is known about the habitat pref- a dry year and relocate them to a model ¯oodplain erences and distribution of early life history stages wetland. After successful spawning and hatching, 968 SOMMER ET AL. we conducted intensive observations of the diel The wetland was ¯ooded in September 2000 before distribution of juveniles in relation to different the initiation of the splittail study. Water surface habitat types. Use of a single model ¯oodplain elevations were maintained by inundating the wet- wetland is somewhat arti®cial and limited in land with well water for 4±6 h/d, which was sup- scope; however, similar studies using captive ®sh plemented by surface runoff from precipitation in seminatural habitats have yielded useful bio- events. In addition to fresh ¯ow from external wa- logical data for other threatened North American ter sources, a screened submersible pump was used cyprinids (Blinn et al. 1998). Our objective was to recirculate water from the outlet to the inlet at to collect basic information on the spawning and a rate of 100 L/min. The wetland was approxi- rearing of splittails that could be used to generate mately oval-shaped and had a mean depth of 0.47 hypotheses for more comprehensive ®eld studies m during the study period. The depth pro®le from and habitat restoration projects. edge to center was gradual (8:1 slope) except for one side, which had a 2:1 slope. Study Site There were several notable differences between Our study was conducted with adult splittails the model ¯oodplain wetland and the Yolo Bypass. collected from the Yolo Bypass, the largest ¯ood- Based on the mean depth, area, and recirculation plain of the San Francisco estuary (Figure 1). The rate, the mean hydraulic residence time (i.e., turn- estuary has been heavily modi®ed by many factors, over rate) in the model wetland was approximately including levee construction, river channelization, 1 d, about twice as fast as that estimated for peak draining of wetlands, diversions, and introduced natural ¯ood events in the Yolo Bypass (California species (Bennett and Moyle 1996). The 24,000-ha Department of Water Resources, unpublished Yolo Bypass ¯oodplain is dominated by agricul- data). The water levels in our model ¯oodplain tural uses, but there are also substantial ``natural'' wetland had a standard deviation of 0.09 m during habitats such as seasonal wetlands and riparian and the observation period, compared with 0.20 m or upland habitats (Sommer et al. 2001a). The largest more during recent long-duration (e.g., .30 d) contiguous area of nonagricultural ¯oodplain hab- Yolo Bypass ¯ood events. Finally, water clarity itat is the Yolo Basin Wildlife Area, which is man- was much higher in our model ¯oodplain wetland. aged by the California Department of Fish and Visibility for divers (see below) was 2±5 m, while Game. The Yolo Bypass typically ¯oods in about visibility during typical Yolo Bypass ¯ood events 60% of years, when high winter and spring ¯ood- (not actually measured) is typically on the order waters enter from the Sacramento River and sev- of less than 0.5 m. eral small streams. The ¯oodplain is seasonally dewatered in summer and fall, except for perennial Methods ponds and a single tidal channel. During extended During February 2001, we collected 14 adult droughts, such as that of 1987±1992, the ¯oodplain splittails (320 6 31 mm [mean 6 SD] fork length is not inundated from its tributaries. Observational [FL]) on their upstream migration using a 3-m- studies were conducted in a 0.1-ha ¯oodplain wet- diameter fyke trap in the Yolo Bypass Toe Drain, land constructed at the Yolo Basin Wildlife Area a perennial tidal channel of the ¯oodplain. As none headquarters, which is immediately adjacent to the of the ®sh were ``ripe,'' we were unable to deter- Yolo Basin Wildlife Area. mine the sex of the ®sh at the time of collection. The model ¯oodplain was constructed in 1997, The ®sh were transported immediately to the mod- then planted with wetland vegetation and season- el ¯oodplain wetland. No other ®sh species were ally ¯ooded during September±May over the next present in the wetland before introduction of the three years. At the time of the study (2001), ap- splittail. proximately 10% of the wetland area was bordered We considered a variety of shallow-water hab- by partially submerged terrestrial vegetation, pri- itat sampling approaches, including seining, dip- marily willows Salix spp., mule-fat Baccharis sal- netting, and electro®shing (Rozas and Minello icifolia, boxelder Acer negundo, and willow herb 1997), to sample the distribution and habitat use Epilobium spp. About 15% of the wetland was cov- of juveniles produced within the model wetland.