Skip to main content

Multidisciplinary modelling approaches to understand harmful algal blooms

Final Report Summary - MOHAB (Multidisciplinary modelling approaches to understand harmful algal blooms<br/>dynamics)

The microscopic planktonic algae of the word’s oceans are critical food for filter-feeding bivalve shellfish as well as the larvae of commercially important crustaceans and fish. In most cases, the proliferation of plankton algae therefore is beneficial for aquaculture and wild fisheries operations. However, in some situations algal blooms can have a negative effect, considered from the human perspective, since they threaten public health and cause economic damage to fisheries and tourism. These episodes encompass a broad range phenomena collectively referred as “harmful algal blooms” (HABs). They include discoloration of waters by mass occurrences of microalgae to toxin-producing species that may be harmful even in low cell concentrations. A broad classification of HAB distinguishes two groups of organisms: 1) the toxin producers, which even with low biomass can contaminate seafood, causing sickness and death in humans eating the seafood, or sickness and death in the shellfish and fin-fish themselves; and 2) the high-biomass bloom species, which can cause either anoxia that indiscriminately kills off marine life, or unpleasant foam or gelatinous masses that are a nuisance for tourists who may develop allergic skin reactions after bathing.

Understanding HABs dynamics is needed before proposing any mitigation and/or management action: this understanding requires an integrated strategy that includes not only observation but also modelling. The great utility of models is that they are effective tools for describing the complex relationship among physical and biological variability in ecosystems. Models can help to understand and identify the key processes in HABs population dynamics and they are extremely useful in the construction and testing of hypotheses. Thus, models are essential tools toward defining the complexity of HAB events, with basic understanding being the first step for ultimate prediction of HABs dynamics (e.g. species in a bloom and the bloom’s intensity, geographical extent, duration, and decline).

Coupling physical effects (turbulence, shear, advection) and biological behavior (migration, physiological adaptation) holds the key to understand HABs dynamics. However, biological models are still at basic stages due to the lack of knowledge of the life history parameters and behavior patterns of HABs species. MOHAB proposed to improve the knowledge on the population dynamics and behavior of HABs species and integrate it within a variety of realistic simulation models. MOHAB will use these models to understand the main mechanism and drivers of the bloom end, and to explore management scenarios and develop bloom indicators to assist in management decisions.

During the outgoing phase we have analyzed the effect of parasites on the development of the 2012 spring bloom of A. fundyense at Salt Pond (Eastham, MA, USA). Results from this work show the importance of parasites to understand the end of A. fundyense blooms and provide one of the first complete series of free-living parasites in the area. Laboratory experiments were carried out to estimate important parameters to model the relationship between A. fundyense and its parasite Amoebophrya spp. These parameters have been estimated in the bibliography for Amoebophrya spp. strains infecting other dinoflagellates. However, we decided to estimate our own parameters due to the high variability of host-parasite interactions and behavior between different strains-species.

Amoebophrya sp.-A. fundyense dynamics were analyzed first using the Rosenzweig–MacArthur modification of the traditional Lotka–Volterra predation model. The model was parameterized using laboratory experimental data. Local and global parameter sensitivity for the model was carried out showing that parasitoid search rate is key to model parasite-host interactions.

During the return phase, parasite-host parameters were also used to run an Individual Based Model (IBM) developed in collaboration with Ifremer colleagues (Arancio et al., 2014, Ecological Modeling). IBM simulations respected the parasite maturation time; no parasites were produced prematurely and different cohorts of parasites and infected hosts were clearly separated. Second waves of infection at a low and parasite-host ratio (1:1; 40:1) were correctly simulated. Host and dinospore concentrations had similar orders of magnitude. Weak and strong but incomplete infection with low and high initial parasite concentrations were also represented in the simulations.
Comparison with the Rosenzweig–MacArthur simulations show that IBM results fitted laboratory data better than previously eulerian models proposed in the bibliography (Fig 1).
IBM simulating the infection dynamics Alexandrium fundyense- Amoebophrya was further used to evaluate different processes that could lead to long-term persistence of host-parasitoid couple in Salt Pond (see pdf presentation from ICES Coruña meeting). Simulations were calibrated with observed data from 2012 (Velo-Suarez et al. 2013). Field infection dynamics started on 20 March 2012 and used as initiation values host and dinospores concentrations from field results. 3 different parasite-host “life-cycle” approaches were used:

1. Vegetative division for host- no co-encystment-excystment between parasite and host.
2. Vegetative division for host and co-encystment-excystment between parasite and host.
3. Complete life cycle (vegetative division + sexual processes) for host and co-encystment-excystment between parasite and host.

Approach 1: Amoebophrya dinospores could not survive in the water column and no infections were observed in 90% of all simulations.

Approach 2: Amoebophrya controlled A. fundyense population in 100% of all simulations. Comparison of observed and simulated host-parasite dynamics in the field show that at low host concentrations, dinospore abundances in the water column depends on Amoebophrya released from infected A. fundyense cysts; during this pre-bloom period, Amoebophrya dinospores are not able to infect their host due to its low abundance (< 500 cells L-1). A. fundyense cells can develop and thrive in Salt Pond because of the low encounter rate with Amoebophrya dinospores (proportional refuge ; Hochberg and Holt (1995)). A decrease in the excystment rate or cyst prevalence in the model modifies the dinospore concentration in the water column and affect the further development of the bloom. However, once A. fundyense reach a concentration above Amoebophrya’s infection threshold, host cells become highly infected resulting in a sharp decline of host population. Alexandrium bloom ended with 100% of infection in 2 waves. Differences between model outputs and observed field data are represented by the over -estimation of infection rates during the bloom. These differences suggest that the host population somehow prevents Amoebophrya infection in the field.

Approach 3: Sexual processes and life-history shifts (e.g. the Cheshire cat strategy) have been pointed out as host refuge strategies to avoid parasite and viral infections. In Salt Pond 2012, planozygote percentages reached very high proportions during the final stages of bloom termination. An apparent life-cycle shift occurred on April 15. Sexual reproduction simulations show a very sensitive response to gametogenesis timing. While early gametogenesis resulted in low –or none- planozygote production due to very low gamete encounters, late gamete production did also prevent cyst production due to Amoebophrya high mortality. Yet, if gametogenesis occurred on 13-14 Apr 2012 (when infected cells reached it maximum), A. fundyense densities were high enough for gametes to meet and Amoebophrya infections still low. In these simulations, infected cysts are produced (18 % of total cyst). Alexandrium bloom ended due to sexual processes and no second wave of infection is produced.

Final results from MOHAB highlight the importance of biological factors affecting natural population dynamics of HABs species. Understanding these dynamics is needed before proposing any mitigation and/or management action: this understanding requires an integrated strategy that includes not only observation but also modeling. MOHAB have contributed to enhance our knowledge on parasite-host couples and characterized biological key processes involved in HABs decline.