Samples of the Bering Sea euphausiid Thysanoessa inerrn~s (Kroyer), collected at 2 stations in the Bering Sea, and consisting primarily of first-year adult females and last-furcilia-stage sub-adults, were studied to determine the percentage infected by the apostome ciliate endoparasite Collinia beringensis (Capriulo & Small). In both groups of krill, infection percentage was found to be extremely high, at 98 and 68 % respectively. The higher rate of 98 % was found in the sample composed of larger and older euphausiids. Such infestation rates suggest that these ciliates may be important modifiers of euphausiid production. Euphausiids (krill) represent a significant proportion of the biomass in the world's oceans and are an important food source for numerous species of whales, fish, seals and birds (Mauchline & Fisher 1969, Mauchline 1980). Given their key ecological role, it is important to identify factors which would negatively impact euphausiid biomass and productivity. Capriulo & Small (1986) reported the discovery of a previously unknown species of apostome ciliate (Collinia berjngensis) living endoparasitically in the haemocoel cavity of the Bering Sea euphausiid Thysanoessa inermis (Krayer). Some of the krill contained extremely high numbers of these ciliates, packed so tightly that normal ciliate cell shapes were altered (Fig. 1). Such high densities of parasites in these planktonic crustaceans suggest the likelihood of physiological impairment to the infected host, perhaps leading to reduced growth, egg production and even death. To begin to assess the impact ot these parasites on krill populations we examined the percentage infection within 2 separate samples of Thysanoessa jnermis collected from the Bering Sea on 2 May (Stn 2012) and 8 May (Stn 2033) 1981, by Dr Sharon Smith and D. Ninivaggi, as part of Brookhaven National Laboratories' involvement in the National Science Foundation funded PROBES project. Samples were taken at night via horizontal tow with a 1 mm mesh plankton net, from surface water (Stn 2012) and 15 m depth (Stn 2033), and fixed in 10 % buffered formalin. The 2 euphausiid samples were characterized with respect to species composition, size, developmental stage and sex (Table 1). Fig. 1. Thysanoessa inermis infected by Collinia beringensis. Photomicrograph shows the protargol-stained ciliate endoparasite (C. beringensis) from the haemocoel fluid of the krill. Note the irregular cell shapes caused by the extremely high ciliate densities found in some hosts O Inter-Research/Printed in Germany 204 Mar. Ecol. Prog. Ser In order to determine the percentage of Thysanoessa inermis infected, 50 individuals were randomly taken (without regard to sex, developmental stage or size) from each sample. Each specimen was rinsed in water, placed under a dissecting microscope on a clean glass slide and cut open in the cephalothorax region. The krill was then removed, leaving the expelled body fluid, which was retrieved with the aid of a Pasteur pipette, on the slide. This fluid was treated with Lugol's solution (iodine-based fixative) to enhance cell contrast for microscope viewing. Slides were then examined under a compound microscope and apostome ciliate presence or absence was recorded. All dissecting instruments were rinsed off between surgical procedures to prevent contamination. Results of this work indicate extreme infection rates in the 2 samples studied of 98 and 68 O/O respectively (Table 2) . The higher rate of 98% occurred in the sample containing the older and larger (mean carapace length 4.52 mm as compared to 3.98 mm, Table 1) krill. Such rates are in marked contrast with those reported elsewhere in the literature for apostome ciliate krill ectoparasites (3 to 16 %; Lindley 1978). The physiological and ecological consequences for the knll harboring Table 1. Thysanoessa mermis and Thysanoessa longipes. Krill population characteristics for nighttime horizontal plankton tow (1 mm mesh net) samples from Stns 2012 (2 May, surface water) and 2033 (8 May, 15 m depth) in 1981 in the Bering Sea. Measurements are based on exammation of 102 and 105 randomly chosen individual knll from Stns 2012 and 2033, respectively. Mean carapace lengths (CL) are presented & SD Species % Pop. n Mean CL Sex (stage)