ability is limited by their small size; while they can migrate vertically on a diurnal or tidal cycle, they cannot swim against tidal currents, but rather they move passively with horizontal movements of water. Therefore it is often better to sample zooplankton and characterize their habitat according to salinity rather than location (Laprise and Dodson 1993). This way of looking at zooplankton is helpful when analyzing the food supply of delta smelt, which also move with the water.
The long-term data show several periods of substantial change in the last 38 years. Many species or groups of species are now at much lower population levels than they were when monitoring started. Declines have occurred throughout the estuary, except possibly Central Bay, but have been most severe in the freshwater delta and the low-salinity zone.
From 1972 through 1986 the zooplankton species composition of the upper estuary was stable except for the introductions of three species of copepod from Asia (Orsi and Mecum 1986). The introduction and subsequent spread of the overbite clam in 1987 caused an immense disruption of the food web in brackish to saline waters between San Pablo Bay and the west-central delta, and several zooplankton species declined sharply (Kimmerer et al. 1994, Kimmerer and Orsi 1996, Orsi and Mecum 1996). Between 1988 and 1994 a series of additional introductions essentially filled in the gap in the summer food web left by the earlier declines (Kimmerer and Orsi 1996, Orsi and Ohtsuka 1999). Since 1994 the food web has seen no further major introductions, yet some declines continue, and most of the species in the low-salinity zone are introduced (Orsi and Ohtsuka 1999, Winder and Jassby 2011).
Most of the introduced species probably arrived in ballast water, although Winder et al. (2011) reported that droughts may have facilitated the spread of some introduced species. Regulations requiring exchange of ballast water at sea since 2000 seem to have reduced the frequency of invasions. A study conducted in 2002-2003 found some potential invaders in ballast water of ships entering the estuary, but their numbers were low and in some cases their condition was poor, suggesting that they were unlikely to overcome the rigors of their new habitat to establish new populations (Choi et al. 2005). The lack of invasions could also be a matter of chance, since a successful invasion requires several coincident conditions that may be met only infrequently (Choi and Kimmerer 2009).
Many of the changes discussed above occurred within the low-salinity habitat of juvenile delta smelt (Bennett 2005). The overbite clam clearly had a substantial effect through grazing on phytoplankton, resulting in poor feeding conditions for some zooplankton. The clam also consumes larval stages of some zooplankton (Kimmerer et al. 1994). The zooplankton species introduced after the clam became abundant have had several advantages over the previously abundant species. First, anchovies abandoned this