Marine Biology (1994) 119:105-113 Springer-Verlag 1994

D. W. Coats K. R. Bockstahler G. M. Berg J. H. Sniezek Dinoflagellate infections of Favella panamensis from two North American estuaries

Received: 3 September 1993 / Accepted: 3 November 1993

Abstract Favella panamensis Kofoid and Campbell, producers to higher trophic levels and act as a major source 1929 is seasonally abundant in meso- to polyhaline waters of nutrient regeneration. In contemporary models of the of Chesapeake Bay and Indian River, Florida, USA, where marine planktonic food web, ciliates represent a two-way it reaches densities of 103 cells 1-1 . During the summers link of primary production to the larger zooplankton (Sherr of 1986-1992, F. panamensis populations of the two estu- et al. 1988). They form a direct connection by repackag- aries were commonly infected by the parasitic dinoflagel- ing the nanophytoplankton into a size fraction that is more late Duboscquella aspida Cachon, 1964. The intracellular readily grazed by the macrozooplankton and represent an phase of the parasite reached maturity in -21 h (30 ~ and indirect connection by forming the terminal link in a mul- consumed -35% of the host's biomass. Infections were not tiple step pathway between secondary bacterial production typically lethal to F. panamensis, but sometimes forced the and zooplankton, the "microbial loop" (Azam et al. 1983). host from its lorica. Several D. aspida were found in the Metazoan consumers are not, however, the only preda- cytoplasm of many hosts, and the number of parasites in- tors of planktonic ciliates. A variety of heterotrophic pro- fection-1 was directly related to infection level. Parasite tists including fungi (Ball 1969), sarcodines (Caron and prevalence averaged 24.0 and 11.5% with mean number of Swanberg 1990; Swanberg and Caron 1991), dinoflagel- parasites infection-1 being 1.5 and 1.3 for Chesapeake Bay lates (Hansen 1991; Schnepf and Elbr~ichter 1992), and and Indian River samples, respectively. D. aspida was es- other ciliates (Dolan and Coats 1991 a, b) utilize ciliated timated to remove up to 68% of host standing stock d -1 protozoa as a food source. Even some typical phytoplank- with a mean of ~ 10% for all samples. The average impact ton species that were conventionally viewed as strict pho- of parasitism on F. panamensis populations was somewhat totrophs are now known to feed on small choreotrich cili- less than would be expected from copepod grazing. ates (Bockstahler and Coats 1993 a, b). Grazing pressure exerted on ciliates by protozoan predators may at times ri- val that generated by larger zooplankton. For example, epi- demic infections of Eutintinnus pectinis by the parasitic Introduction dinoflagellate Duboscquella cachoni may remove in ex- cess of 50% of the host standing stock d -1 in Chesapeake Ciliated protozoa are widely recognized as playing a crit- Bay (Coats and Heisler 1989). The annual impact of this ical role in the trophodynamics of planktonic food webs parasite on host populations was estimated to equal that of (Stoecker and Capuzzo 1990; Gifford 1991; Reid et al. the dominant metazoan grazer, Acartia tonsa. 1991; Pierce and Turner 1992). These small heterotrophs Several other heterotrophic dinoflagellates act as intra- mediate the transfer of energy and matter from primary cellular parasites of loricate and aloricate ciliates (Cachon and Cachon 1987). Among these, Duboscquella aspida has been most thoroughly studied with infections reported for a number of tintinnids including Coxliella lacinosa, Codo- Communicated by J. R Grassle, New Brunswick nella campanula, Favella ehrenbergii, and Eutintinnus D. W. Coats ([]) K. R. Bockstahler J. H. Sniezek frakndii (Cachon 1964). The life cycle ofD. aspida, as de- Smithsonian Environmental Research Center, scribed for infections in E ehrenbergii (Cachon 1964), con- Box 28, Edgewater, Maryland 21037, USA sists of an intracellular vegetative phase followed by an extracellular reproductive/dispersal phase. During the tro- G. M. Berg Marine-Estuarine and Environmental Sciences, phont (i.e., vegetative) stage, the parasite increases in size University of Maryland, College Park, and forms a large yellowish mass within the cytoplasm of Maryland 20742, USA the host. This intracellular phase of the parasite's life cy- 106 cle is terminated when the trophont ruptures through the and photographed. These direct observations provided data on de- cortex of the host, and in the process phagocytizes a large velopmental events and the duration of sporogenesis. Information on the population dynamics of DuboscqueIla aspi- portion of the ciliate's cytoplasm. Cachon noted that emer- da was obtained through the study of a natural host-parasite assem- gence of the trophont through the host's cortex was often, blage. A 120-liter cylindrical plastic tub was filled with Indian Riv- but not always, fatal to F. ehrenbergii. Once outside the er water containing parasitized Favella panamensis and maintained host, the parasite undergoes a rapid series of divisions to at ambient temperature (~ 30 ~ in a flowing estuarine water bath. The isolated population was exposed to the natural light regime and produce numerous biflagellate dinospores that initiate new sampled at 2- to 4-h intervals over 34 h. For each sample, 4 liters of infections if ingested by another susceptible host. water were concentrated on 20-gm mesh Nitex, back-flushed to While Duboscquella aspida is not host specific, it ap- 25 ml using multiple rinses, and preserved in Bouin's fluid. F. pana- pears to parasitize Favella spp. more frequently than oth- mensis abundance, number of host Ioricae without ciliates, and num- er tintinnids, with infection levels in F. ehrenbergii some- ber of loricae containing parasite sporogenic stages were determined by examining 2-ml aliquots of preserved sample (=320 ml whole- times approaching 100% (Cachon 1964). Cachon also sug- water) using inverted microscopy (x200). Parasite prevalence, num- gested that epidemic outbreaks of D. aspida in the plank- ber of parasites infection-1 , and the relative occurrence of four se- ton of Algiers harbor caused abrupt decreases in the abun- quential stages (I-IV, respectively) in the intracellular, trophic phase dance of F. ehrenbergii, even though some host organisms of D. aspida were determined from hematoxylin preparations with 100 hosts examined sample-1. The abundance of parasite trophonts were capable of surviving infection. at each sampling time was calculated by multiplying mean number Here we document the occurrence of Duboscquella as- of intracellular parasites host-I by host density. pida in populations of Favella panamensis from two North An estimate for the duration of the parasite's trophic phase was American estuaries, Chesapeake Bay and Indian River, derived from data on the temporal occurrence of different intracel- lular stages of during this experiment. Data Florida. Data on parasite and host morphology, parasite Duboscquella aspida analysis was similar to that used by Heinbokel (1988) for determin- generation time, and in situ infection levels are used to es- ing duration of ciliate reproductive stages. The percentages of para- timate the potential impact of this parasite on host popu- sites represented by stage I-II and III-IV infections were plotted sep- lations. arately, and the time interval during which those percentages exceed- ed their respective mean values was determined by linear extrapola- tion. The mid-point of each interval provides an objective estimate for the time of maximum occurrence of each stage, and the differ- Materials and methods ence between the two mid-points equals half the intracellular devel- opment time. Parasitism of Favella panamensis Kofoid and Campbell, 1929 by The impact of parasitism on FavelIapanamensis populations, ex- the dinoflagellate Duboscquella aspida Cachon, 1964 was studied pressed as the percent of host standing stock removed d -1, was cal- by examining specimens collected from Chesapeake Bay and Indian culated as: River, Florida, USA. Samples from Chesapeake Bay were taken VvzrXNXD -1, at routine stations located along the major axis of the Bay during cruises in 1986 to 1991. Station protocol followed established where Vp/r is the mean value for the volume of the mature parasite procedures (Coats and Heisler 1989) and included conductivity-tem- divided by total parasite-host volume (=0.35); N is the number of perature-depth profiles, Niskin bottle samples at eight to ten depths, parasites (=infection level multiplied by parasites infection-I); and and vertical net tows (30-gm mesh). Plankton collections containing D is the development time for the intracellular phase of the parasite sufficient numbers of E panamensis for the present study were in days adjusted for sample temperature using a Q10 of 2. only obtained from the meso- to polyhaline portion of the Bay Living and stained specimens were viewed and photographed us- south of the Choptank River (i.e., between 38034 ' and 37o07 ' N lat- ing Zeiss microscopes equipped with brightfield, phase contrast, and itudes). differential interference contrast optics. Intracellular parasite stages Samples from the Indian River, a shallow polyhaline estuary on were measured on hematoxylin stained cells using a filar microme- the southeastern coast of Florida, were obtained in June and Septem- ter; means are reported with standard errors (+ SE) and sample siz- ber 1991, and in August 1992. On each occasion, 18 stations even- es (n). Cell volumes for hosts and the post-phagotrophic stage of Du- ly spaced along a 40-mile transect from Sebastian Inlet (27o52 ' N boscquella aspida were calculated using appropriate geometric for- lat.) to St. Lucie Inlet (27o10 ' N lat.) were sampled over a 1- to 2- mulae and cell dimensions obtained from photographic images of wk period. Whole-water and net tow (35-gin mesh) samples were living specimens. collected near the surface of each station, and measurements were made for temperature and salinity. Host abundance was determined by inverted microscopy (x200) using replicate 50-ml aliquots of whole-water samples preserved Results with a modified Bouin's solution (Coats and Heinbokel 1982). For assessment of parasite prevalence, host cells collected in net tows or Parasite life history by screening 2 to 4 liters of sample onto 20-gm Nitex were fixed in modified Bouin's and stained with acidulated alum hematoxylin (Ga- ligher and Kozloff 1971) or by Protargol silver impregnation (Mon- Four morphologically distinct stages in the trophic phase tagnes and Lynn 1987). Parasite prevalence was obtained by scoring of Duboscquella aspida were identified in hematoxylin the number of Duboscquella aspida present in each of 100 Favella stained specimens. Very early infections, stage I organ- panamensis sample-I . For observations of living parasites, infected ciliates were isms, were spherical to ovoid cells (9.6_+0.5 by 8.3+0.4 gin; washed by micropipetting cells through several changes of 0.45-gm n---28) that had a prominent cytoplasmic vacuole and a large filtered water and individually transferred to wet-mount slide prep- nucleus (6.4 _+0.4 gm diameter; n=28) containing a single arations. Wet mounts consisted of ~ 0.1 ml of filtered estuarine wa- nucleolus (Fig. 1). Stage II parasites (Fig. 2) were notice- ter sealed beneath a coverslip and enclosed within a vaspar ring (50% petroleum jelly; 50% paraffin) of sufficient thickness to provide an ably larger (17.5_+0.7 by 13.2_+0.4 gin; n=32), had two or ample air space. Isolated specimes were placed in humidity cham- more cytoplasmic vacuoles and multiple nucleoli (nuclear bers, held in a Percival incubator at 25 ~ and periodically examined diameter=10.1_+0.3 gm; n=32), hut were otherwise simi- 107 lar to stage I cells. Mid- to late infections (stage III) were Host-parasite population dynamics characterized as individuals that had highly vacuolated cy- toplasm and numerous pseudopod-like protrusions of the Information on the population dynamics of Favella pana- cell surface (Fig. 3). These irregularly shaped cells meas- mensis and Duboscquella aspida was obtained by moni- ured 30.0+_0.9 by 25.7+1.0 gm and had a nuclear diameter toring a natural host-parasite assemblage during a 34-h in- of 16.9_+0.6 gm (n--30). As the trophont approached ma- cubation study (ambient temperature -30 ~ The study turity (stage IV), all pseudopod-like structures were re- was initiated in late afternoon (t7:00 hrs, -3 h before sun- sorbed and the cytoplasm took on an homogeneous smooth set) using Indian River plankton that contained 240 hosts to finely granular appearance (Fig. 4). Stage IV trophonts 1-1 with an infection level of 31%. Parasite prevalence de- were roughly hemispherical in profile (Fig. 5) with a max- creased to about a third of the initial level over the 7.5 h imum dimension of 46.4+1.5 gm and a nuclear diameter of the incubation (To to T7.5), then showed a step increase of 23.0_+0.6 ~tm (n=40). to -40% near the end of the first dark period, followed by The morphogenetic process resulting in the release of a second jump to -70% at the beginning of the second dark Duboscquella aspida from the host only required 3 to period (Fig. 13A). By comparison, F panamensis steadily 4 min and usually ruptured the pedicel that anchored the increased in number early in the incubation and reached a tintinnid to its lorica (Fig. 6). In some instances, the host maximum density of-1200 cells 1-I at T16 (Fig. 13A); dou- remained within the lorica as the parasite continued to de- bling time =6.7 h. No growth of the host population was velop, but often the ciliate swam out of the lorica after its evident following the first increase in parasite prevalence, pedicel was broken. The post-phagotrophic stage of the and an abrupt decline in F. panamensis abundance to -400 parasite averaged 66.1_+2.7 gm in diameter (n=21) and cells 1-1 coincided with the second rise in infection level. contained a conspicuous food vacuole (Fig. 6). This stage While the percent of hosts infected by D. aspida decreased of the parasite had a volume of 1.67x105 gm 3 (_+0.2; n=21) between To to T7.5, the absolute number of intracellular and and represented 35%+2.1 (n=21) of total host-parasite extracellular stages of the parasite present during this pe- biomass; host biovolume after emergence of the parasite riod remained relatively constant (Fig. 13B). Thus, the de- was 3.19x105 gm 3 (+_0.4; n=21). Sporogenesis (Figs. 7 to cline in parasite prevalence probably resulted from growth 9) lasted 4.9_+0.7 h (n=7) at 25 ~ and gave rise to of the host population without the spread of infections, 2000-3000 dinospores (Fig. 10) that averaged 7.0_+0.1 by rather than from the loss of parasites. 4.2_+0.1 gm (n=20). We could not count dinospores accu- Duboscquella aspida exhibited two distinct peaks in rately in wet-mount preparations, but the volume ratio for sporogenesis, with each peak followed by a sharp rise in post-phagotrophic parasites and dinospores indicated that the abundance of intracellular parasite stages (Fig. 13B). ca. 2600 dispersal cells were formed infection-I. The short lag between sporogenesis and the subsequent oc- When followed in the laboratory, the transition of the currence of new infections (3 and 6 h for the first and sec- parasite from intracellular to extracellular phase did not ond peak, respectively), along with the stepwise increase kill the host. The ability of Favella panamensis to survive in infections, suggests that dinospores of D. aspida dis- infections was also evident in plankton samples, where perse rapidly, but have a relatively narrow interval during sporogenic stages were frequently observed in the loricae which they are competent to establish infections (cf. Fig. of actively swimming hosts that, aside from being small, 13A, B). Were each of the parasites undergoing sporogen- had a normal appearance. Sporogenic stages of Dubosc- esis during the first peak (-30 1-1 at T12.5) to release 2600 quella aspida were sometimes observed in loricae that dinospores as estimated above, then successful infection were not inhabited by a host. In such cases, the ciliate may of hosts by 8 to 10% of the dinospores would account for have been killed by the infection, or may have swum out the number of intracellular parasites observed in subse- of the lorica after the pedicel was severed by the parasite. quent samples (630 to 830 1- 1 between T16 and T21,5; Fig. Parasitism did not appear to prevent reproduction of the 13B). host, as cytological preparations revealed that all stages of The presence of well defined peaks in sporogenesis of the ciliate's cell cycle including mitosis could harbor well Duboscquella aspida indicates that at least a portion of the developed infections. Interestingly, food vacuoles of post- parasite population was developing in synchrony. Addi- phagotrophic parasites sometimes contained one of the tionally, a pronounced oscillation in the relative occurrence host's macronuclei. The reproductive competency of indi- of early (stage I & II) and late (stage III& IV) trophonts vidual Favellapanamensis that have lost one or more mac- of D. aspida was evident from the onset of the incubation ronuclei is unknown. and provides further evidence for phased development of Favella panamensis was often infected by multiple par- the parasites (Fig. 13C). Mean values for percent early and asites (Fig. 11), with up to 13 Duboscquella aspida ob- late trophonts during the incubation were 49,0+6.4 and served in a single host. Multiple infections were manifest- 43.7+5.5 (n= 13), respectively (mean parasites in sporogen- ed as either several parasites of the same developmental esis =7.3+2.1%). These values are indicated by the hori- stage or parasites in two or more stages. In some instanc- zontal lines that intersect the two plots of Fig. 13C. The es, the cytoplasm of host organisms contained trophonts of intervals indicated by these lines represent the period when D. aspida in all four stages of development, while outside successive data for percent early and late infections ex- the ciliate, but still within the lorica, another parasite was ceeded their mean values. The mid-point of each interval undergoing sporogenesis. is marked by a short vertical line, and the time increment 108 109

75 between mid-points (10.3 h) is an objective estimate for is00 3 half the duration of the parasite's intracellular phase. Thus, ~, 03 the time from initial infection until D. aspida erupted ,Z" through the cortex of FaveIla panamensis, was -21 h at 7_ 1000 50 I 30 ~ i5 At the beginning of the incubation, 23% of infected hosts (=7% of all hosts) were parasitized by more than one ~ ~oo 25 "E Duboscquella aspida with an average of 1.3+0.09 (n=31) z parasites infection -1. Multiple infections became progres- o sively less common prior to the first peak in sporogenesis o 0 but were frequently encountered thereafter with a maxi- 17:00 01:00 09:00 17:00 01:00 mum of 59% (at T34) of host cells containing multiple par- ~ lO00 125 "~ asites. Values for number Of parasites per infected Favel- 7 ? la panamensis ranged from 1.0 at Tll to 2.2 at T34 and I O0 showed a strong correlation with infection level (Fig. 14). ~ 750 _J The loricae of all Favella panamensis examined prior ,~ 75 C~ o~ to the first peak in sporogenesis had a structure that was o 500 o 03 50 typical for the species (i.e., goblet-shaped with uniform .2 wall thickness and a distinctive aboral horn; see Fig. 8); "5 2s0 25 o~ o however, at T12.5, a number of E panamensis had spiral- L walled loricae that lacked an aboral horn (Fig. 12). The oc- o 0 0 o_ currence of spiralled loricae was positively correlated with -~ ' I I I I I I I ] I I I 03 17:00 01:00 09:00 17:00 01:00 parasite prevalence (r=0.85; p<0.01; n=13) and reached a frequency of 9.8% at T34. 100 Stages I & II (I) i~r.lx Stages Ill & IV (o) 80 /\ Parasite prevalence in situ and impact on host populations O3 60 O 13_ The occurrence of Duboscquella aspida in Chesapeake 40 Bay and Indian River populations of Favella panamensis was determined for samples where host abundance was O9 \o a_ 20 >10 cells 1-1. Concentrations of F. panamensis above c 10 ceils 1-1 were only encountered in the mid- to lower part 0 ' I I I I I I ' I I } ' I of Chesapeake Bay at salinity of 14 to 27%0 and tempera- 17:01 01:00 09:00 17:00 01:00 ture of 18 to 29 ~ By contrast, dense patches of hosts Time of Day were observed throughout the Indian River study area Fig. 13 Favellapanamensis parasitized by Duboscquella aspida. Incubation study of Indian River plankton; interval between sunset Figs. 1-12 Favellapanamensis infected by Duboscquella aspida. and sunrise indicated by dark area on abscissa. (A) E panamensis Fig. 1 Stage I of trophont development showing a conspicuous nu- density and parasite prevalence. (B) Abundance D. aspida present cleolus (arrow) and vacuole (v); hematoxylin stain; scale =20 gm. as intracellular (trophonts) and extracellular (sporogenic) life histo- Fig. 2 Stage II parasite with multiple nuclei (arrows) and large vac- ry stages. (C) Relative occurrence of early (stage I and II) and late uole (v); macronuclei of host (m); hematoxylin stain; scale = 20 gm. (stage III and IV) trophonts of D. aspida. Horizontal lines represent Fig. 3 Stage III trophont with numerous pseudopod like extensions interval when successive data for percent early and late trophonts ex- (arrows); parasite nucleus (n); host macronucleus (m); hematoxylin ceeded their mean values (49 and 44%, respectively); short vertical stain; scale =20 gm. Fig. 4 Stage IV in the intracellular phase of D. line indicates mid-point of each interval. Time between mid-points aspida; parasite nucleus (n); host macronuclei (m); hematoxylin (10.3 h) equals half the duration of the parasite's intracellular phase stain; scale =20 gm. Fig. 5 Stage IV trophont (arrow) just before emergence from the host; differential interference contrast (DIC) mi- croscopy of a living specimen; scale =40 gm. Fig. 6 Same specimen as Fig. 5 immediately after emerging from the host; parasite food where salinity ranged from 21 to 33%0 and temperature was vacuole (fv); pieces of the ruptured pedicel (arrows); DIC live spec- 28 to 32 ~ imen; scale =40 gm. Fig. 7 Four-cell stage of sporogenesis; parasite Data on parasite prevalence were obtained for 36 food vacuole (arrow); DIC live specimen; scale =40gin. Indian River samples of which 13 were collected in June Fig. 8 Late sporogenesis of D. aspida with host still occupying lor- and September of 1991 and 23 in September 1992. In- ica; DIC live specimen; scale = 40 gm. Fig. 9 Release of sporocytes from host lorica; phase contract microscopy of live specimen; scale fection levels were typically between 1 and 40%; howev- =40 gm. Fig. 10 Mature macrospores showing bipartite structure er, parasites were below detectable levels (<1%) on seven and transverse flagella (arrows); DIC live specimens; scale occasions and infected 69% of the hosts in one sample; =10 ~tm. Fig. 11 E panamensis infected by four trophonts (arrows) mean parasite prevalence was 14.4% (Table 1). Infec- of D. aspida; parasite nucleus (n); host macronuclei (m); Protargol silver-impregnated specimen; scale =20 gm. Fig. 12 F. panamensis tion levels for 1991 and 1992 samples were not significant- with spiralled lorica; Bouin's fixed specimen; phase contrast micros- ly different, averaging 12.5+5.3 and 15.5+2.3, respective- copy; scale =40 gin. ly. 110

2.25 20 r=0.94 I u p<

L0~ ~II o 1.25

..... & 1.00 I- ' l ' I ' I ' l ' I ' 0 I I I 10 20 30 40 50 60 70 100 Parasite Prevalence (%) T o Fig. 14 Duboscquellaaspida infection of Favellapanamensis. Cor- 75 relation between the number of parasites per infected host and par- asite prevalence during incubation of Indian River plankton E 50 "7 Table 1 Favellapanamensis infected by Duboscquella aspida. In- fections in F. panamensis collected from the Indian River and Ches- 25 t apeake Bay 3 {{ (3" Ld Host Parasite Parasites % standing i I ' I ' I ' abundance prevalence infection -1 stock 200 400 600 800 removed d-I Mean Host Density Fig. 15 Duboscquella aspida impact on Favella panamensis. (A) Indian River, Florida Proportion of host standing stock removed d-1 as a function of mean (n=36) host density. (B) Biomass consumed by the parasite expressed as an Median 175 11.5 1.3 6.9 equivalent number of host cells and plotted against mean host den- Mean 230 14.4 1.3 8.5 sity. Mean host densities are for Indian River and Chesapeake Bay SE 36 2.4 0.06 1.6 data grouped b~ host abundance into five categories: < 1001-a (n= 15); Range 10-1070 0-69 1.0-2.0 0-50 >100 to 2001- (n=10); >200 to 3001- (n=10); >300 to 4001-1 (n=5); >4001 -l (n=4). Errorbarsrepresent standard error of the mean Chesapeake Bay (n=8) Median 115 24.0 1.7 7.3 Mean 120 32.1 1.5 18.3 SE 29 11.1 0.16 8.0 ing stock removed d -1, are presented in Table 1. Mean val- Range 10-220 0-86 1.1-2.1 0-68 ues for Chesapeake Bay and Indian River were not signif- icantly different (18.3+8.0 and 8.5+1.6, respectively), and the average for all samples was 10.3+2.1; n=44. On the Only eight Chesapeake Bay samples had sufficient four occasions on which parasite prevalence exceeded numbers of Favellapanamensis to determine parasite prev- 50%, Duboscquella aspida was estimated to remove 28 to alence. Duboscquella aspida was detected in five of these 68% of the host's standing stock d -1. collections with infection level exceeding 50% on three When grouped according to host abundance, estimates occasions; mean parasite prevalence was 32.1% (Table 1). for the proportion of Favella panamensis standing stock Average infection level in F. panamensis from Chesapeake removed by Duboscquella aspida had comparable averag- Bay was about twice that of Indian River populations; how- es at mean host densities of 40 to 725 cells 1-1 (Fig. 15A). ever, mean values for the two estuaries were not signifi- The total amount of host biomass utilized by the parasite cantly different and the average for all samples was increased with mean host density and was equivalent to re- 17.6+2.7; n=44. Also, parasite prevalence was not corre- moving 2 to 70 F. panamensis d -1 (Fig. 15B). lated with host density (p>0.05) in either system. Multiple infections were common in Favella panamen- sis collected from both estuaries with as many as 53% of infected hosts containing more than one parasite (mean Discussion =20+2.9%, n=34). As in the preceding incubation study, the number of Duboscquella aspida per infected host was Loricate choreotrichs of the genus Favella are predomi- closely correlated with parasite prevalence (r=0.65; nantly neritic ciliates that are most common in temperate p<0.01; n=34). Mean number of parasites infection -1 waters (Campbell 1942; Pierce and Turner 1993) and of- ranged from 1.0 to 2.1 with Indian River and Chesapeake ten associated with dinoflagellate blooms (Stoecker et al. Bay populations averaging 1.3 and 1.5, respectively (Ta- 1981, 1984; Sellner and Brownlee 1990). Most Favella ble 1). spp. are large and densities rarely exceed 103 cells 1-~, yet Estimates for the impact of parasitism on Favella pa- they are a consistent and potentially important component namensis populations, expressed as percent of host stand- of the microzooplankton in many areas. Favella spp. are 111 known to feed on dinoflagellates in preference to other phy- abandonment of the lorica once the pedicel of the ciliate is toplankton taxa and are, in turn, a choice prey of copepods severed. Laval-Peuto (1981) has shown that F. ehrenber- (Robertson 1983; Stoecker and Sanders 1985; Ayukai gii, when removed from its lorica, forms a new lorica that 1987; Stoecker and Egloff 1987). Ciliates of this genus are has a distinctive spiralled appearance. The presence of spi- also consumed by gelatinous zooplankton, decapod zoea, ralled loricae in E panamensis was positively correlated and fish (Robertson 1983; Stoecker and Govoni 1984; with infection level and suggests that hosts forced from Stoecker et al. 1987 a, b). their loricae by developing parasites survive to form new Favella spp. have long been known to harbor parasitic loricae. dinoflagellates including Duboscquella aspida (Duboscq Multiple infections with as many as 13 parasites in an and Collin 1910; Chatton 1920; Cachon 1964), but the ef- individual host were frequently encountered in Favella fect of parasitism on host populations has been largely panamensis. The number of parasites infection-1 was pos- overlooked. Cachon (1964) was the first to point out po- itively correlated with infection level, and at maximum in- tential ecological consequences of parasitic dinoflagellates fection level, each infected host had an average of two par- when he suggested that epidemic outbreaks of D. aspida asites. In some instances, all the infections were of equiv- might regulate the abundance ofFavella ehrenbergii. More alent age (i.e., same stage of trophont development) and recently, dinoflagellate infections have been linked to low were probably established within a few hours of each oth- in situ growth rates of an unidentified species of Favella er. Hosts that survive infection and remain with their lor- (Stoecker et al. 1983). Coats and Heisler (1989) have icae would be exposed to a large number of infective stag- shown parasitic dinoflagellates to be widespread in Ches- es over a very short period and could develop multiple in- apeake Bay, where parasite-induced mortality of another fections of the same age. Alternatively, the infective dino- loricate ciliate, Eutintinnus pectinis, averaged ~ 10% of the spores of the parasite might remain clustered after leaving host standing stock d -1. the lorica and be encountered in patches. Other multiple The dinoflagellates infecting FaveIlapanamensis in the infections contained parasites of differing age. These prob- current study had morphological features that correspond- ably resulted from sequential acquisition of infections; ed closely to the description of Duboscquella aspida however, interaction among trophonts might retard the (Cachon 1964) for infection in F. ehrenbergii; however, growth of some Duboscquella aspida and shift the appar- various aspects of parasite development differed between ent age distribution of the parasites. the two host species. Some of these inconsistencies prob- Favella panamensis populations of Chesapeake Bay ably reflect the different temperatures at which observa- and Indian River, Florida were often heavily infected by tions were made (25 to 30 ~ in our study; -20 ~ for Duboscquella aspida, with maximum parasite prevalence Cachon's work). For example, the intracellular phase of of 69 and 86% for the two estuaries, respectively. These the parasite had a duration of -21 h in F. panamensis with values are substantially higher than peak parasite preva- sporogenesis requiring -5 h, whereas trophont develop- lence reported for D. cachoni infections in Eutintinnuspec- ment and sporogenesis were reported to take 3 to 4 d and tinis populations of Chesapeake Bay (Coats and Heisler 2 to 3 d, respectively, in F. ehrenbergii (Cachon 1964). Oth- 1989). Infection levels above 50% were never observed in er discrepancies may reflect strain variations in the para- E. pectinis, but were encountered in -10% of F. panamen- sites or differences in host-parasite interactions. For ex- sis samples. That E panamensis commonly survives infec- ample, Cachon (1964) reported two size classes of dino- tions and is subject to reinfection may explain the frequent spores for infections of F. ehrenbergii with each parasite occurrence of very high parasite prevalence. E. pectinis on producing either -1000 macrospores (6 to 7 gm long) or the other hand is always killed by its parasite (Coats and >50000 microspores (2 to 3 gin). Only macrospores Heisler 1989). Thus, very high infection levels in that host (7 x 4 gm) were formed by parasites of F. panamensis with species could persist for only a short time and would be in- each infection liberating 2000 to 3000 daughter cells. Al- frequently encountered. so, infections usually killed F. ehrenbergii (Cachon 1964), The impact of Duboscquella aspida on Favella pana- but were often not lethal to F. panamensis. mensis populations, expressed as percent host standing Progression of Duboscquella aspida from an intracellu- stock removed d-1, averaged 18.3 and 8.5 for Chesapeake lar to the extracellular phase of the life cycle was not typ- Bay and Indian River samples, respectively, with epidem- ically lethal to Favella panamensis under laboratory con- ic infections (>50%) cropping 28 to 68% of host standing ditions, although emergence of the parasite usually severed stock d -1. By comparison, Coats and Heisler (1989) esti- the pedicel that anchored the host to its lorica. That E pana- mated 7 to 24% of the Eutintinnus pectinis population of mensis also survives infection in the field was evident from Chesapeake Bay was removed by D. cachoni. While E pa- the occurrence of loricae containing host cells and sporo- namensis supports higher infection levels and has more genic stages of the parasite. E panamensis was sometimes parasites infection-1 than E. pectinis, trophonts of D. as- dislodged from its lorica as the parasite underwent sporo- pida only utilize -35% of the host cell, whereas those of genesis, and loricae containing only parasite developmen- D. cachoni consume the entire host. As a consequence, the tal stages were not uncommon in field samples. Cachon effect of parasitism on standing stocks of the two host spe- (1964) suggested that absence of a host cell and presence cies is roughly comparable. of parasites indicated that D. aspida was lethal to F. ehren- Estimates for impact of parasitism on Favellapanamen- bergii. Alternatively, this situation may simply result from sis populations assumed that all parasites, whether present 112 as single or multiple infections, matured to the post-phag- Bockstahler KR, Coats DW (1993 a) Spatial and temporal aspects of otrophic stage, had comparable development times, and mixotrophy in Chesapeake Bay dinoflagellates. J eukaryot Mi- crobiol (J Protozool) 40:49 - 60 utilized the same proportion of host biomass. These would Bockstahler KR, Coats DW (1993 b) Grazing of the mixotrophic di- be over-estimated should older trophonts of multiple in- noflagellate Gymnodinium sanguineum on ciliate populations of fections decrease the success or increase development time Chesapeake Bay. Mar Biol 116:447-487 of younger parasites. A more conservative approach would Brownlee DC, Jacobs F (1987) Mesozooplankton and microzoo- plankton in the Chesapeake Bay. In: Majumdar SK, Hall LW Jr., be to assume that only one parasite of each multiple infec- Austin HM (eds) Containment problems and management of tion reaches maturity, which would reduce the impact of living Chesapeake Bay resources. Pennsylvania Academy of Sci- D. aspida by an average of 34%. Our estimates also as- ences, Easton, Pennsylvania, p 217-269 sume that parasite prevalence was independent of sampling Cachon J (1964) Contribution ~ l'6tude des p~ridiniens parasites. time, but the incubation study of a natural plankton assem- Cytologie, cycles 6volutifs. Annls Sci nat (Sdr Zool) 6: 1-158 blage showed significant temporal fluctuation in infection Cachon J, Cachon M (1987) Parasitic dinoflagellates. In: Taylor FJR levels associated with phased development and spread of (ed) The biology of dinoflagellates. Blackwell Scientific Publish- infections. Decreasing infection levels in the evening were ers, Oxford, p 571-610 followed by abrupt increases in parasite prevalence at night Campbell AS (1942) The oceanic Tintinnoina of the plankton gath- ered during the last cruise of the Carnegie. Carnegie Institute of and relatively steady levels of parasitism through the day. Washington Publication 537, Washington DC If this pattern is characteristic of field populations, our es- Caron DA, Swanberg NR (1990) The ecology of planktonic sarco- timates of parasite impact should not be biased as all sam- dines. Rev aquat Sci 3: 147-180 ples were collected during the day. Chatton, E (1920) Les pEridiniens parasites. Morphologie, reproduc- tion, ethologie. Arch Zool exp gen 59:1-475 Ingestion rates for copepods feeding on spp. Favella Coats DW, Heinbokel JF (1982) A study of reproduction and other range from 10 to -264 prey copepod -1 d -1 at prey densities life cycle phenomena in planktonic protists using an acridine of 250 to 3400 1-1 (Stoecker and Sanders 1985; Ayukai orange fluorescence technique. Mar B iol 67:71- 79 1987; Stoecker and Egloff 1987). Robertson (1983) report- Coats DW, Heisler JJ (1989) Spatial and temporal occurrence of the parasitic dinoflagellate and its tintinnine ed values of 25 and 90 prey copepod -~ d -1 for Acartia ton- Duboscquella cachoni host Eutintinnus pectinis in Chesapeake Bay. Mar Biol 101:401 sa feeding on E panamensis at 250 and 1000 cells 1-1, re- -409 spectively, and 160 F. panamensis copepod -1 d -1 for Tor- Dolan JR, Coats DW (1991 a) A study of feeding in predacious ci- tanus setacaudatus exposed to 1000 prey 1-1. Utilization liates using prey ciliates labeled with fluorescent microspheres. of host biomass by Duboscquella aspida was equivalent to J Plankton Res 13:609-627 Dolan JR, Coats DW (1991 b) Preliminary prey digestion in a pre- removing an average of 2 to 70 F. panamensis d -~ at mean dacious estuarine ciliate, Euplotes woodruffi, and the use of di- host densities of 40 to 720 cells 1-1 . Thus, the average ef- gestion data to estimate ingestion. Limnol Oceanogr 36:558- fect of parasitism of E panamensis standing stock was 565 comparable to the grazing pressure exerted by a copepod Duboscq O, Collin B (1910) Sur la reproduction sexu6e d'un pro- tiste parasite des tintinnides. CR hebd S6anc Acad Sci, Paris 151: density of one individual 1-1. Since copepod densities of- 340-341 ten exceed one individual 1-1 in enriched coastal systems Galigher AE, Kozloff EN (1971) Essentials of practical microtech- (e.g.A. tonsa abundance in Chesapeake Bay during sum- nique, 2nd edn. Lee and Febiger, Philadelphia mer ranges from 4 to 20 copepods 1-1; Brownlee and Ja- Gifford DJ (1991) The protozoan-metazoan trophic link in pelagic cobs 1989), D. aspida would have a lower impact on F. ecosystems. J. Protozool 38:81-86 Hansen PJ (1991) Dinophysis - a planktonic dinoflagellate genus panamensis populations than would metazoan grazers, ex- which can act both as a prey and a predator of a ciliate. Mar cept during epidemic outbreaks when the parasite is ca- Ecol Prog Ser 69:201-204 pable of removing 68% of F. panamensis standing stock Heinbokel JF (1988) Reproductive rates and periodicities of ocean- d -1. ic tintinnine ciliates. Mar Ecol Prog Set 47:239-248 Laval-Peuto M (1981) Construction of the lorica in Ciliata Tintinni- Acknowledgments Support for this research was provided through ha. In vivo study of Favella ehrenbergii: variability of the phe- NSF grant OCE-8911316 awarded to DWC, and by the Smithsonian notypes during the cycle, biology, statics, biometry. Protistolog- Marine Station at Link Port, SMSLP Contribution #340. We thank the ica 17:249-272 Captains and crews of the R. V. "Ridgely Warfield", R. V. "Cape Hen- Montagnes DJS, Lynn DH (1987) A quantitative Protargol stain lopen", and R. V. "Cape Hatteras" for ship operations and on-deck as- (QPS) for ciliates: method description and test of its quantitative sistance; W. Lee and H. Reichert for help in collecting samples; and nature. Mar microb Fd Webs 2:83-93 Mr. T. K. Maugel for use of darkroom facilities at the Laboratory for Pierce WR, Turner JT (1992) Ecology of planktonic ciliates in ma- Biological Ultrastructure, University of Maryland, College Park. rine food webs. Rev aquat Sci 6: 139-181 Pierce WR, Turner JT (1993) Global biogeography of marine tintin- nids. Mar Ecol Prog Ser 94:11-26 Reid PC, Turley CM, Burkill PH (1991) Protozoa and their role in References marine processes. NATO ASI Series G: Ecological Sciences, 24. Springer-Verlag, Berlin Ayukai T (1987) Predation by Acartia clausi (Copepoda:Calanoida) Robertson JR (1983) Predation by estuarine zooplankton on tintin- on two species of tintinnids. Mar microb Fd Webs 2:45-52 nid ciliates. Estuar cstl Shelf Sci 16:27-36 Azam F, Fenchel T, Field JG, Gray JS, Meyer-Reil LA, Thingstad F Schnepf E, Elbr~ichter M (1992) Nutritional strategies in dinoflagel- (1983) The ecological role of water column microbes in the sea. lates: a review with emphasis on cell biological aspects. Eur J Mar Ecol Prog Ser 10:257-263 Protistol 28:3 - 24 Ball GH (1969) Organisms living on and in protozoa. In: Chen T-T Sherr BF, Sherr EB, Hopkinson CS (1988) Trophic interactions with- (ed) Research in protozoology, Vol 3. Pergamon Press, Oxford, in pelagic microbial communities: indications of feedback regu- p 565-718 lation of carbon flow. Hydrobiologia 159: 19-26 113

Sellner KG, Brownlee DC (1990) Dinoflagellate-microzooplankton Stoecker DK, Govoni JJ (1984) Food selection by young larval gulf interactions in Chesapeake Bay. In: Gran61i E, Sondstrom B, Ed- menhaden (Brevoortia patronus). Mar Biol 80: 299-306 let L, Anderson DM (eds) Toxic marine phytoplankton. Elsevi- Stoecker DK, Guillard RRL, Kavee RM (1981) Selective predation er, New York, p 221-226 by Favella ehrenbergii (Tintinnia)on and among dinoflagellates. Stoecker DK, Capuzzo JM (1990) Predation on protozoa: its impor- Biol Bull mar biol Lab, Woods Hole 160: 136-145 tance to zooplankton. J Plankton Res 12:891-908 Stoecker DK, Michaels EA, Davis LH (1987 a) Grazing by the jel- Stoecker DK, Davis LH, Anderson DM (1984) Fine scale spatial cor- lyfish, Aurelia aurita, on microzooplankton. J Plankton Res 9: relations between planktonic ciliates and dinoflagellates. J Plank- 901-915 ton Res 6:829-842 Stoecker DK, Sanders NK (1985) Differential grazing by Acartia ton- Stoecker DK, Davis LH, Provan A (1983) Growth of Favella sp. (Ci- sa on a dinoflagellate and a tintinnid. J Plankton Res 7: 85-100 liata: Tintinnina) and other microzooplankters in cages incubat- Stoecker DK, Verity PG, Michaels AE, Davis LH (1987b) Feeding ed in situ and comparison to growth in vitro. Mar Biol 75:293 - by larval and post-larval ctenophores on microzooplankton. J 302 Plankton Res 9:667-683 Stoecker DK, Egloff DA (1987) Predation by Acartia tonsa Dana on Swanberg NR, Caron DA (1991) Patterns of sarcodine feeding in planktonic ciliates and rotifers. J exp mar Biol Ecol 110:53-68 epipelagic oceanic plankton. J Plankton Res 13:287-312