¹Department of Ecology & Biodiversity, The University of Hong Kong

Hong Kong, CHINA

lamhyi@hkusua.hku.nk

　

²School of Science and Technology, The Open University of Hong Kong

Hong Kong, CHINA

　

　

Nutrients, particularly total inorganic nitrogen (TIN) and dissolved inorganic phosphates (DIP), are generally considered to be the major triggering and limiting factors for harmful algal blooms. While some models have been developed on the basis of TIN:DIP ratios to interpret red tide occurrences in Hong Kong, collection of real time data by in situ measurement of nutrients has been constrained by limited research funds and available technology. A pilot test incorporating an auto-nutrient analyzer with a telemetry system was conducted to study the possibility of providing in situ and real time nutrient data for environmental models. A diurnal study of phytoplankton dynamics at Crooked Island, Hong Kong evaluating the effectiveness of the integrated analyzer-telemetry system is reported.

　

　

Since the first scientific report of red tide in Hong Kong (Morton and Twentyman, 1971), there has been more three decades of red tide research in Hong Kong. However, it appears that no significant progress has been made in understanding the formation mechanisms of red tides. Some studies have indicated that eutrophication (Lam and Ho, 1989) and an increase in nutrient (N and P) loading (Hodgkiss and Ho, 1992) may be considered as the most important factors related to the frequent occurrence of red tides in Tolo Harbour, Hong Kong. Ho and Hodgkiss (1993), using bottle bioassay tests, showed that the growth of four dinoflagellates isolated from Tolo Harbour was optimum at N:P (atomic) ratios from 4-22, and further studies on Prorocentrum micans, P. sigmoides and P. triestinum suggested that increased red tide due to all three species in the late eighties might have been related to a reduced N:P ratio in the water of inner Tolo Harbour (HO and Hodgkiss, 1995). Thereafter, nutrients, and total inorganic nitrogen (TIN) and dissolved inorganic phosphates (DIP) in particular, have been considered as the major triggering and limiting factors for harmful algal blooms in Hong Kong.

　

Hodgkiss and Ho (1997) reviewed evidence collected from Tolo Harbour together with data from Japan and North European coastal waters and indicated that both long term and relatively short-term changes in the N:P ratio are accompanied by increased blooms of non-siliceous phytoplankton groups and showed that nutrient ratios could be more informative than nutrient concentration per se. Some models on the basis of TIN:DIP ratios have been developed to interpret red tide occurrences in Hong Kong and such a N:P ratio hypothesis seems to have been established validly both in Tolo Harbour and other areas of Hong Kong waters in recent studies (Yang, 2000; Lu, 2001).

　

However, it is unknown whether such N:P ratios are applicable at Crooked Island, where it is believed that the 1998 episode of large scale red tides was initiated (Anderson et al., 1998). Indeed, little is known concerning the effects of nutrient dynamics on the formation of red tides there. To better understand the relationship between N:P ratios and red tides, the collection of real time data by continuous in situ measurement of nutrients is important but has been constrained by limited research funding and the available technology. The traditional approach in red tide studies is based on periodic and discrete sampling, and such an approach can cause loss of data between sampling intervals. In view of this limitation and in order to provide more background information about nutrient dynamics during red tide formation, telemetry has been adopted to study and monitor red tides (Lee et al 2000). However, the lack of real time in situ as well as continuous nutrient data prevented insight being gained into the role of nutrients in the period preceding red tides. A pilot test involving diurnal observations of marine microalgal dynamics, was conducted to test the feasibility of integrating an auto-nutrient analyzer into this telemetry system, providing in situ and real time nutrient data for environmental models, as well as to investigate the limitations and technical problems, if any. The telemetry system itself consists of a series of probes to monitor chlorophyll a, water temperature, dissolved oxygen, solar radiation, wind, tidal level and current. The data are logged in a micrologger and transmitted via modems to The University of Hong Kong (for data retrieval, analysis and storage) from the fish raft mounted system at Crooked Island (Lam and Hodgkiss, 2000, 2001).

　

　

　

A fixed field research station at Crooked Island (latitude: 22°33'N, longitude 114°18'E), which is situated in the north-eastern waters of Hong Kong, located between latitude 22°9' and 22°37'N and longitude 113°52' and 114°30'E, in Mirs Bay (Fig.1), was chosen for this pilot study. At the station, a telemetry system was deployed on a fish raft. This system consists of a membrane type dissolved oxygen probe, fluorometer, Acoustic Doppler Current meter, a peristaltic pumping system, a micrologger, a 12V car battery, thermistor, anemometer and pyranometer, telecommunication software through modems for data retrieval. The more detailed configuration, operation and functioning of the telemetry have been given in the literature (Lee et al., 2000).

　

　

　

An auto-nutrient analyzer, AutoLAB Model ANA-1 developed and manufactured by W.S. Ocean System Ltd, was employed for this pilot testing. The analyzer measures the concentrations of nutrients in water using conventional wet chemistry and colourimetric analysis techniques and consists of several parts, including a nutrient analysis module, an electronics enclosure and bottles for the storage of chemicals. Each analysis module is comprised of three major components: an eight-way rotary valve; a motor driven syringe; and a colorimeter. Water samples and chemical reagents are taken by syringe plunger and mixed for reaction and then injected into the colourimeter for further development and measurement. Results of measurements can be stored in memory for data extraction, and the system can be used on-line or in telemetry applications (W.S. Ocean Systems, 2000).

　

The two analysis modules employed in the pilot test were nitrate and phosphate and the standard seawater methods of nitrate and phosphate analysis were based on the literature (Armstrong et al. , 1967; Murphy and Riley, 1962 respectively) (W.S. Ocean Systems, 2000).

　

　

Water sampling for diurnally based nutrient analysis was carried out from 16:00h on 30 January 2002 to 13:00h on 31 January 2002. Two 300 ml water samples were taken simultaneously every four hours near the water surface (depth 1 m) and filtered on site through a membrane of 0.45μm pore size. One set of the filtrates was collected in a 100 ml polyethylene vial, stored in an ice container and then sent to the Agriculture, Fisheries and Conservation Department (AFCD) laboratory, Hong Kong Special Administrative Region (HKSAR) Government for analysis with a ChemLab Continuous Flow Analyser. The other set of filtrates was analyzed in situ with the auto-nutrient analyzer.

　

In addition, another set of water samples was taken hourly from the same water depth and analyzed in situ with the auto-nutrient analyzer in order to obtain finer resolution and to test the reliability of the analyzer's response to the changes of nutrient concentration within a short period.

　

　

Water samples, in addition to those from 1 m water depth referred to above, were collected from 3 m and 5 m every four hours, and sent to AFCD for analyzing dissolved inorganic nitrogen as ammonia, nitrate, nitrite; and dissolved inorganic phosphorus (DIP) as ortho-phosphate. The filtration procedure and storage of water samples for analysis was the same as above.

　

　

N:P ratios (atomic weight) were calculated from the AFCD results by dividing the total inorganic nitrogen (TIN) value, which was the summation of ammonia, nitrate and nitrite, by the relevant ortho-phosphate value.

　

　

2L water samples were collected simultaneously with the samples for nutrient analysis from the same three water depths, that is, 1 m (surface: s), 3 m (middle: m) and 5 m (bottom: b) and concentrated by a special hand-made tube-shaped sieve of mesh size 20 μm, and then collected in a 100 ml polyethylene vial; preserved with 3 ml Lugol's solution immediately; stored in an ice container; and taken back to the laboratory for overnight settling in a refrigerator at 4℃ to a final volume of 5 ml - 10 ml, after the supernatant was siphoned off. A thoroughly mixed aliquot of 1 ml from each sample was pipetted into a Sedgwick Rafter counting chamber and 10-20 minutes allowed for settling of the microalgae in the counting chamber. Under an Olympus IX50 inverted microscope with magnifying power ranging from 100 to 400 times, the algal species present were identified and the cell number of each enumerated. The total number of cells of each individual species in the original sample was finally calculated and the total microalgal "biomass" was expressed as the number of cells per litre.

　

Identification of microalgae was carried out using Jin et al (1965). Yamaji (1984). Larsen & Moestrup (1989), Fukuyo et al. (1990), Hallegraeff (1991), Hallegraeff et al (1995), Balech (1995) and Tomas (1997). Diatoms and dinoflagellates were the main focus of this research, but other microalgae were identified as far as possible.