Chemical ecology of fish-killing microalgae: Toxins, toxin production and the effects on the marine food web

Research output: Book/ReportPh.D. thesis

  • Sofie Bjørnholt Binzer
Some microalgal species are able to form high biomass blooms in coastal areas that negatively affect the marine environments. These blooms can cause mortality of the marine fauna at all trophic levels, including fish and shellfish. This thesis concerns members of two important harmful algal genera, Karlodinium and Prymnesium, and their toxins. Karlodinium armiger, known from recurrent annual harmful blooms in the Mediterranean, produces karmitoxin, a toxin that resembles karlotoxins, produced by the closely related K. veneficum. The toxic effects of K. armiger cultures on different life stages of blue mussels were studied using different algal concentrations. The effect on mussels was severe, and adult mussels rejected the K. armiger cells and refused to ingest them. The mussels died within 24 h of exposure to ecologically relevant algal concentrations and a clear dose-response relationship was observed. The mussel embryos and larvae were more sensitive and died at lower algal concentrations than the adults. Micropredation of K. armiger on the early life stages of mussels was observed. The effects on fish larvae were different, as swarming behavior or attachment of K. armiger cells in high numbers on fish larvae was not observed, even at high K. armiger concentrations. The toxic effects of K. armiger cultures on juvenile fish and fish larvae were also investigated and dose-response relationships were established.
A method for quantitation of karmitoxin in culture samples was developed which allowed for the first measurements of quantified toxin concentrations during laboratory experiments. Purified karmitoxin lysed rainbow trout gill cells, and caused mortality in both fish larvae and copepods in a dose-dependent manor. Purified karmitoxin induced physical damages to fish larvae similar to what was caused by live K. armiger cells. However, twice the amount of pure toxin was needed to induce similar toxicity as was observed with live cells. Although a loss of karmitoxin of twenty percent was observed during the experimental exposure, this could not explain the discrepancy. Other factors may as well influence the toxicity of live cultures. It is possible that there are other unknown toxins at play, in addition to karmitoxin, or that live cells facilitate toxin transfer towards the fish larvae and thus, increase the toxicity of a live culture. To study the toxic mechanism of K. armiger further, a K. armiger culture was treated with HP-20 resin, which absorbs the extracellular karmitoxin. Approximately 37% of the total karmitoxin concentration in the algal culture was removed by HP-20, without a reduction in cell concentration. This coincided with decreased toxicity towards fish larvae; there was high mortality of fish larvae in the untreated controls, whereas the fish larvae which were exposed to cultures treated with HP-20 were immobilized, but survived the 12 h exposure. Altogether, these results suggest that karmitoxin released by K. armiger to the surrounding water constitutes the main toxic mechanism in fish kills by K. armiger.
Three fish representatives expressed different sensitivities towards K. armiger and karmitoxin in these studies. Juvenile rainbow trout were around six times more sensitive than fish larvae when exposed to live K. armiger culture, and gill cells were about three times more sensitive than fish larvae when exposed to purified toxins. These differences are important to keep in mind when interpreting mortality data of different fish substitutes exposed to K. armiger and possibly also other microalgae that produce broad targeting toxins.
Production of karmitoxin by K. armiger and the toxicity of the K. armiger culture towards fish larvae were investigated in K. armiger batch cultures grown with two different nitrogen sources (prey and ammonium). The karmitoxin production and cellular karmitoxin content were in the same range whether cultures were grown with prey or ammonium. Karmitoxin production took place during exponential growth, and stopped when the cultures entered stationary growth phase. Thus, no accumulation of karmitoxin was observed in the cultures. Fish larvae were killed within 2 h of exposure to the highest cell concentrations of K. armiger cultures, irrespective of the nitrogen source.
Prymnesium parvum, the other species investigated in this thesis, is a cosmopolitan species and one of the most important microalgae associated with fish kills. The toxins produced by P. parvum are known as prymnesins, and each strain produces one type (A-, B-, or C-type), depending on the length of the carbon backbone in the prymnesin molecule. Prymnesin production seems to play an important role for P. parvum as all 26 P. parvum strains investigated in a screening study produced prymnesins. A global distribution was found for the three toxin types and no apparent biogeographical pattern was observed. Interestingly a phylogenetic analysis revealed that three clades could be described based on the three toxin types. Thus, each toxin type has monophyletic origin. No prymnesins were detected in any of the other five Prymnesium species tested.
A quantitation method was developed for prymnesin B in algal pellets and water samples. For the first time, Prymnesin concentrations were assessed in pH/carbon limited batch cultures of the two P. parvum strains (K-0081 and K-0374). Strain K-0081 was found to contain about five times more toxin than K-0374, corresponding well with the higher toxicity also found for this strain. For both strains the cellular prymnesin content varied by a factor of two and a half during the growth period. The highest cellular toxin concentrations were found in the exponential growth phase, during prymnesin production. The lowest cellular prymnesin concentrations were found during stationary/death phases, accompanied by no and even negative net prymnesin production, indicating no accumulation of prymnesin in these cultures. When using different separation methods 64-89% of the prymnesins were found to be associated with the biomass. Nevertheless, the proportions of biomass associated versus excreted prymnesins remained constant during a full growth cycle.
Original languageEnglish
PublisherDepartment of Biology, Faculty of Science, University of Copenhagen
Publication statusPublished - 2019

ID: 230477283