Capabilities and considerations for the implementation of a phycocyanin sensor to enhance cyanobacteria monitoring in lakes

Abstract

Bloom-forming cyanobacteria are problematic in recreational waters as humans may be exposed to their toxins via primary contact or ingestion. Cyanobacteria can be monitored using microscopy but this is time consuming, costly and requires taxonomic expertise. Phycocyanin (an accessory pigment specific to cyanobacteria) can be used as a proxy for cyanobacteria biomass. Phycocyanin sensors are thus increasingly being used to monitor cyanobacteria, even though many limitations to their use still exist. This research investigated the opportunities and challenges of use a phycocyanin sensor for cyanobacteria monitoring. It tested three hypotheses that: 1) there would be a strong relationship between phycocyanin and biovolumes in samples collected from the Te Arawa/Rotorua Lakes, North Island, New Zealand, 2) colony morphology and cell size affect phycocyanin readings, and 3) nutrient and light exposure would affect phycocyanin quotas independently of growth in Microcystis aeruginosa.

The relationship between phycocyanin and biovolume was investigated using data collected in the field from over two summers (2016 and 2017). A phycocyanin sensor was used to measure phycocyanin in situ, and biovolume was enumerated by microscopy. Eutrophic lakes with high biovolumes (>1.8 Mm³ L⁻¹) and single species dominance had stronger relationships between phycocyanin. Phycocyanin concentration >40 µg L⁻¹ derived from the sensor approximated a biovolume of 1.8 Mm³ L⁻¹, which is the health warning level for potentially toxic cyanobacteria species under the New Zealand guidelines for recreational monitoring of cyanobacteria in fresh waters.

The effect of colony morphology and cell size on phycocyanin detection was tested with serial dilutions of cultures of four cyanobacterial species. Large colonial Microcystis wesenbergii had the highest variability in phycocyanin readings from the sensor. Non-linear relationships in all four species resulted in low confidence for predicting low biovolumes <1.8 Mm³ L⁻¹ from phycocyanin.

The effects of nutrients and light intensity on growth and phycocyanin quota in M. aeruginosa were assessed by a laboratory experiment using a Central Composite design and Response Surface Methodology (RSM). RSM models tested for significant interactions and effects of 20 different combinations of nitrogen (26-84 mgL⁻¹), phosphorus (0.05-5.47 mg L⁻¹) and light intensity (10-300 µmol M⁻² s⁻¹) on the growth and phycocyanin quota. Phycocyanin from the sensor and cell concentrations from microscopy were measured over 26 days at five-day intervals. Phycocyanin quota was significantly (P<0.05) higher in four of the 20 treatments at day 18 compared to the day 22. RSM demonstrated that light and nutrient concentrations affected both growth rate and phycocyanin quota differently. Importantly phycocyanin quotas at day 18 and 22 were affected differently by light and nutrients. This experiment suggests phycocyanin quota changes in cyanobacteria, and this may result in over or underestimates of biomass by a sensor. Regardless of the challenges of using phycocyanin sensors with changing species compositions, morphology, density and with the effects of nutrients and light on phycocyanin and growth. Phycocyanin sensors offer an opportunity to increase current sampling capability and the prioritisation of high-risk samples for counting which may lead to improved protection of human health from the toxicity associated with cyanobacteria blooms.