interest for reasons besides its status as a potentially 

 threatened endemic: 1) It is consistently one of the 

 most abundant species in many of the brackish 

 sloughs of the estuary (Moyle and Daniels unpubl. 

 data; Caywood 1974), 2) most other cyprinids are 

 exclusively freshwater species, rarely found in brack- 

 ish waters, 3) its life history patterns reflect adap- 

 tation to an environment in which drought and flood 

 occur episodically, and 4) it supports a small but lo- 

 cally important hook-and-line fishery (Caywood 

 1974). 



Methods 



Fishes were collected in the Suisun Marsh between 

 January 1979 and January 1982, using 2.5 m otter 

 trawls (mesh at cod end 6 mm bar) and 3 mm bar 

 beach seines. The Suisun Marsh is the largest con- 

 tiguous tidal marsh on the eastern Pacific coast (Fig. 

 1). It comprises 34,000 ha of marshland, sloughs, and 

 shallow bays (Moyle et al. 1982a). At each locality, 

 salinity (%o), water temperature (°C), secchi depth, 

 and turbidity were recorded. Standard length (SL) 

 was measured on all fish captured; for most months a 



random sample of splittail was preserved in 4% form- 

 aldehyde solution. In the laboratory, preserved 

 specimens were measured and weighed. Scales were 

 removed from 210 randomly selected fish, from the 

 area above the lateral line, caudad to the posterior- 

 most point of the pectoral fin. These scales were 

 mounted on glass slides and projected on a mi- 

 crofiche reader screen. Measurement and back- 

 calculation follow Tesch (1968). A condition factor 

 (K = wt/SL(3 X 10 6 )) was calculated for each fish 

 (Tesch 1968). Gonads were removed and weighed; 

 fecundity was determined by weighing three sub- 

 samples from each prespawning ovary, counting the 

 number of ova present in each subsample, averaging, 

 and multiplying by total gonad weight. Stomachs 

 were removed; contents were separated, identified, 

 and weighed. When possible, the total length of prey 

 items was measured. In April 1979, fish were collect- 

 ed over a 24-h period for a feeding habit study. Full- 

 ness indices (Windell 1968) were calculated for each 

 fish as an indirect measure of activity. 



Several statistical techniques were used. One-sided 

 £-tests were used to compare means between two 

 groups; one-way analysis of variance was used to 



FIGURE 1.— Suisum Marsh, Sacramento-San Joaquin Estuary, Calif. Samples were taken from all sloughs identified. 

 Monthly (1979) or bimonthly (1980-81) sample sites are designated (*). Each site represents two tows. 



648 



