Info

Gr/mod corresponds to the percentage of the modality belonging to the group, and mod/gr corresponds to the percentage of the group belonging to the modality. Modality label as in Table 5. n denotes the number of observations

Gr/mod corresponds to the percentage of the modality belonging to the group, and mod/gr corresponds to the percentage of the group belonging to the modality. Modality label as in Table 5. n denotes the number of observations alternative prey were available and Phaeocystis was selected against ("Against"). The predator-to-prey size ratio modality "P:p1" (size ratio < 4) also has an important contribution to the first axis (10%), showing that small copepods (such as Centropages hama-tus, Acartia clausi and Pseudocalanus elongatus) tend to reject Phaeocystis in situ. On axis 2, the two modalities "Stationary" and "For" have the greatest contribution (18% and 20%, respectively). This axis separates grazing experiments where Phaeocystis in a stationary growth phase was selected positively by predators in situ. In the first factorial plane we also observed that Phaeocystis concentration and predator-to-prey size ratio in lab experiments were higher than what was generally observed in situ (Fig. 1A) and that higher carbon-specific ingestions were mostly related to laboratory experiments (Fig. 1B).

The combined analysis of the MCA and HCA clearly distinguished four groups of observations (Fig. 1B and Table 7). Groups 1 and 2 bring together 97% of the grazing experiments performed in the field and are separated from groups 3 and 4, which comprise 96% of the lab experiments with Phaeocystis provided as single prey. Group 1 consists of field experiments with small-size copepods selecting against P. globosa. Group 2 includes all field experiments with positive selection for P. pouchetii (93% of the group), and is characterized by a slightly higher predator-to-prey size ratio (87% of them had a Pip-ratio of 4-16, "P:p2"). In these experiments Phaeocystis was in the stationary growth phase and in relatively low abundance ("Ab1" = abundance between 0.1 and 125 p,g C l-1). The group 3 is characterized by lab grazing experiments of Temora spp. on P. globosa. The predator-to-prey size ratio in these laboratory experiments was higher (89% of them had a P:p ratio of 64-256, "P:p4") than what was commonly observed in grazing experiments on natural plankton. Similarly we observed higher carbon-spe-ciWc ingestion than what was observed in situ for this copepod genus (group 1). Group 4 brings together the lab grazing experiments on P. pouchetii by larger copepods (Calanus spp.). This group includes the highest Phaeocystis abundance and predator-to-prey size ratio tested in lab experiments and grouped the highest carbon-specific ingestion estimates.

In conclusion, the MCA and the HCA analyses suggested that: (1) the lowest grazing rates from the Weld (often zero) were recorded for small copepods (Acartia, Pseudocalanus, Temora and Centropages) in blooms of Phaeocystis globosa colonies, (2) large copepods (e.g., Calanus) feeding on Phaeocystis pouchetii had higher grazing rates, especially in lab studies, when no alternative food was present and (3)

P. pouchetii in a stationary growth phase was positively selected by large copepods.

It is important to note that grazing data for large copepods (Calanus spp. and Metridia longa) were only available for P. pouchetii, whereas grazing studies with small copepods (Acartia, Centropages, Pseudocalanus and Temora) were limited to P. globosa. Thus, the presently available data do not allow us to determine whether the different results for P. pouchetii and P. globosa are due to differences between the two species, or differences between the predators. In order to test for species-specific differences, we need to compare grazing on both species simultaneously using similar methodologies in future studies.

Quantitative results from data on crustacean grazing

Because the statistical analysis suggested a large difference between feeding rates in laboratory and field investigations, we compared the average specific ingestion rates in all available field and laboratory experiments (Table 6), and for five groups of predator-to-prey size ratios separately (Fig. 2). These comparisons revealed that: (1) overall feeding (carbon specific ingestion) on Phaeocystis spp. was significantly lower [p < 10"8, analysis of variation (ANOVA)] in all field (average 2.5% d"1) studies with natural plankton, compared to laboratory studies (average 11% d"1), (2) the highest average carbon-specific ingestion rate (23% d"1) on Phaeocystis spp. was found when the predator-to-prey size ratio (P:p) was 4-16 in lab studies, whereas there was no clear size ratio trend in field studies, and (3) ingestion was low (<2% d"1) when P:p was <4 in both lab and field, indicating that the upper effective size limit for prey equivalent spherical diameter was V of the predator prosome length.

Thus, crustacean grazers show a much lower grazing on Phaeocystis in the field than in the laboratory. One reason for this discrepancy could be that fewer than 5% of the laboratory studies offered the cope-pods alternative prey to Phaeocystis whereas alternative prey were available in the field. Alternatively, laboratory cultures might not display the same chemical grazing cues (and possible grazing deterrents) as Phaeocystis growing in natural plankton. Hapto-phytes may lose their inhibitory effects in vitro, and toxicity may be species-, strain- or growth-condition-

Fig. 2 Crustaceans (all stages of copepods and krill) grazing on Phaeocystis pouchetii, P. globosa or P. antarctica, in laboratory and field experiments. Average daily carbon-specific ingestion rates (percent ^g C ^g C"1 d"1) are presented for five different predator-to-prey size categories [derived from predator prosome or carapace length and average equivalent spherical diameter (ESD) of the Phaeocystis prey, solitary cells or colonies]. Error bars denote the standard error (SE) of the average; the number of observations (n) for each category are given above each column. Note the high daily ration for P:p size ratio 64-256 in the field is due to a single (30%) value (with that value excluded, the rate is 3.0 § 1.0 %)

Fig. 2 Crustaceans (all stages of copepods and krill) grazing on Phaeocystis pouchetii, P. globosa or P. antarctica, in laboratory and field experiments. Average daily carbon-specific ingestion rates (percent ^g C ^g C"1 d"1) are presented for five different predator-to-prey size categories [derived from predator prosome or carapace length and average equivalent spherical diameter (ESD) of the Phaeocystis prey, solitary cells or colonies]. Error bars denote the standard error (SE) of the average; the number of observations (n) for each category are given above each column. Note the high daily ration for P:p size ratio 64-256 in the field is due to a single (30%) value (with that value excluded, the rate is 3.0 § 1.0 %)

specific (Edvardsen and Paasche 1998). Support for this hypothesis comes from laboratory studies that failed to recreate toxicity in the lab that was observed in the field, even with haptophytes isolated from highly toxic blooms, such as the Chrysochromulina polylepis bloom in 1988 (Nielsen et al. 1990). Likewise, the antipredation effect revealed by Estep et al. (1990) in actively growing field-collected Phaeocystis pouchetii apparently disappeared during the first 12 h in experimental containers. It may be that laboratory or other in vitro studies underestimate the negative effects of potentially toxic haptophytes in situ, but this requires explicit evaluation.

Ideally, future investigations on the ecology of zooplankton feeding should preferably be conducted in situ, although laboratory experiments are needed to investigate the effects of potentially important signaling substances between Phaeocystis and different predators in presence of realistic alternative prey.

Grazing by protozooplankton and other microzooplankton

General trends in grazing by protozoan microzooplankton on Phaeocystis are difficult to assess since the number of quantitative investigations is limited (Tables 3 and 4). Very few studies have investigated

Table 8 Microzooplankton grazing on Phaeocystis globosa solitary cells during blooms in the English Channel off northen France May 2003 and April 2004 (JC Nejstgaard, AF Sazhin and LF Artigas unpubl.)

Date

Prey type (|im)

r2

H (d1)

G (d-1)

Grazing impact (% SS d-1)

Apr 2003

Chl a (>0.45)

Was this article helpful?

0 0

Post a comment