Environmental Drivers of Phormidium Blooms in New Zealand Rivers

Abstract

Proliferations of toxic benthic cyanobacteria occur in rivers worldwide. During the last decade, there has been an increase in animal poisoning linked to toxic mat-forming cyanobacteria, and in New Zealand, there have been over 70 dog deaths reported since 2005. The predominant genus in New Zealand responsible for these proliferations and dog deaths is Phormidium, which produces a range of powerful neurotoxic compounds called anatoxins. In optimal conditions, Phormidium can form extensive green/black mats that span large expanses of river substrate. Phormidium proliferations have resulted in heightened concerns for drinking water safety and human health. Future climate change forecasts, including extended dry periods and rising nutrient concentrations in aquatics systems, are likely to increase the frequency and severity of Phormidium blooms. Despite the health risks and increased abundance, there is a limited understanding of the environmental drivers and processes leading to Phormidium proliferation and toxin production in New Zealand rivers. In this study, I examine a suite of physicochemical factors to help elucidate drivers of Phormidium proliferations and anatoxin production in New Zealand rivers.

River flow has been cited as the most important physical variable regulating Phormidium abundance. However, there have been no in-depth investigations exploring this relationship. In this thesis, over 650 measurements of Phormidium habitat preferences (velocity, depth and substrate type) were taken at seven Hutt River (Wellington) sites in stable, summer low flows, to create habitat preference curves. Quantile regression and forage ratio methodologies were used to fit habitat preference curves. Large substrates were the most suitable habitat for Phormidium growth, most likely due to increased substrate stability and surface heterogeneity which provides refuges for cells during flushing events. The optimal river velocity was between 0.7 and 1.1 m s⁻¹, however Phormidium mats were found in velocities over 2 m s⁻¹ demonstrating that Phormidium can grow in a wide range of physical conditions. The preference curves were used in a hydraulic habitat model, which predicts the amount of available habitat in different flows, to investigate the relationship between flow and Phormidium physical habitat availability. When flow reductions were modelled, only minor increases in the proportion of available habitat were observed. The results of these physical habitat analyses, in combination with other recent research, suggest that flow, and in particular the frequency between flushing flows, is important in regulating the extent of Phormidium blooms. However, other factors such as water chemistry determine whether blooms will occur at a specific site. During this study, higher Phormidium abundance was associated with increased nitrate concentrations, and this potentially explains the large variation observed in Phormidium cover throughout the Hutt River.

To investigate the response of Phormidium to variations in nitrogen and phosphorus concentration, two non-axenic Phormidium autumnale strains (CAWBG48 [non-anatoxin producing] and CAWBG557 [anatoxin producing]) were grown in batch monocultures. Cell concentrations of both strains were significantly (P<0.001) reduced under depleted nitrogen (<1.40 mg L⁻¹) and phosphorus conditions (<0.08 mg L⁻¹). Reduced growth in depleted nitrogen conditions supports field data that have shown Phormidium proliferations were positively associated with elevated dissolved inorganic nitrogen (DIN). Furthermore, molecular analysis of 13 New Zealand Phormidium autumnale strains revealed they were unable to fix atmospheric nitrogen; thus, elevated DIN may be a prerequisite for bloom events. The reduced growth observed under depleted phosphorus conditions is in contrast with observations in New Zealand rivers where Phormidium blooms occur in very low concentrations of dissolved reactive phosphorus (ca. <0.01 mg L⁻¹). Phormidium blooms might be able to utilise phosphorus from underlying sediment or from particles trapped within the mat, through biogeochemical processes. The ability of Phormidium to bloom in the low phosphorus conditions in New Zealand rivers may provide a competitive advantage over other periphyton species.

Phormidium autumnale strain CAWBG557 produces anatoxin-a (ATX), homoanatoxin-a (HTX) and their dihydrogen-derivatives dihydroanatoxin-a (dhATX) and dihydrohomoanatoxin-a (dhHTX). In the batch culture experiments described above, cellular anatoxin concentration (anatoxin quota) of all four variants increased up to four fold in the initial growth phase (days 0-9) indicating anatoxins may have a physiological benefit during this phase and could be involved in substrate colonisation. Dihydroanatoxin-a and dhHTX, accounted for over 60% of the total anatoxin quota and this suggests that dihydrogen variants (dhATX and dhHTX) are being internally synthesised and not just derived following cell lysis and environmental degradation of ATX and HTX, respectively. The four anatoxin variants differed in their response to varying nitrogen and phosphorus concentrations. Notably, dhATX quota significantly decreased (P<0.03) when nitrogen and phosphorus concentrations were elevated (nitrogen = 21.0 mg L⁻¹; phosphorus = 3.0 mg L⁻¹), while HTX quota significantly increased (P< 0.03) when the phosphorus concentrations were reduced (<0.08 mg L⁻¹). This latter finding is of considerable concern as HTX is 10 fold more toxic than dhATX and dhHTX.

Nutrient bioassay experiments were used in-situ to test nutrient limitation, micro-algal competition and anatoxin production at five sites along the Hutt River. Phosphorus and nitrogen availability were manipulated to provide four treatments: no nutrients control, nitrogen only, phosphorus only and nitrogen and phosphorus. Phormidium cell concentrations were significantly higher at the two downstream sites (P<0.001), but contributed less than 15% of the overall periphyton biomass after 12 days growth. Diatoms were dominant at all sites contributing over 50% of the total biomass. Phormidium cell concentrations did not differ significantly among nitrogen and phosphorus treatments. These data suggested that: (1) Phormidium is a mid/late succession species (and, that the experiment had not been left in place long enough for sufficient growth), (2) the growth rate of Phormidium is slower at upstream sites, and (3) unknown factors may be contributing to Phormidium growth downstream. Homoanatoxin-a and dhATX were the only anatoxins detected in this experiment. Notably, HTX was higher in the treatments without phosphorus which is consistent with the monoculture experiment. Bacterial diversity was assessed using Illumina™ sequencing at one of the five sites. Phormidium was the dominant taxon in the control, nitrogen only and phosphorus only treatments (44%, 50% and 35% of total sequence reads respectively) indicating it is the dominant prokaryote when either nitrogen and/or phosphorus is limited in the Hutt River.

In Summary, I elucidated key physical and chemical drivers leading to Phormidium proliferations in New Zealand rivers, as well as, the diversity of habitats suitable for Phormidium growth and proliferations. This study highlights the importance of nitrogen and phosphorus fluxes in regulating Phormidium growth. Nitrogen and phosphorus, as well as growth phase, were also responsible for the spatial and temporal variations in anatoxin concentration and the relative proportion of different variants. This provides new fundamentally important knowledge about Phormidium proliferations and their environmental regulation. Furthermore, it has laid the ground work for future studies that will be needed to accurately model Phormidium bloom events in New Zealand rivers.