Chiku Lagoon is a semi-enclosed coastal lagoon located on the south-western coast of Taiwan (23.16°N, 120.08°E; Figure 1). The lagoon is shallow (<2 m at low tide) and connected to Taiwan Strait through two narrow inlets on the sandbar. The lagoon receives freshwater mainly from the nutrient-rich Chiku River and Daliao Creek which drain agricultural soils, mangrove swamps and aquaculture ponds. The total lagoon area and volume are around 10 km² and 12×10⁶ m³, respectively.

The study area has a longer dry season (October–April) than wet season (May–September). Freshwater discharge ranges from 265–563x10³ m³ d^-1 in the wet season and 85–240x10³ m³ d^-1in the dry season. Because of the small system volume, the system salinity is subjected to seasonal variability, ranging during the study period from 20.6 psu in September (1997) to 33.8 psu in April (1997). Water temperature ranges from 16°C in winter to 32°C in summer.

The lagoon is highly productive, primarily due to inputs of nutrient-enriched freshwater. The gross production was estimated to be about 90 mol C m^-2 yr^-1 or 250 mmol C m^-2 d^-1 (Wong 1999). The lagoon is now intensively farmed with oyster-hanging culture and is also fished for finfish, oysters and shellfish (Lin et al. 1999).

Although the lagoon is apparently healthy and productive now, it may incur changes in the future from planned petrochemical and steel industries around the watershed. Detrimental changes in the lagoon may also endanger certain biological species, including black-faced spoonbills that use the lagoon as their winter shelter. In order to study current lagoon biogeochemistry and function, water samples were collected bimonthly from November 1996 to October 1998 for hydrochemical, nutrient and trace-metal distributions. Water, nutrient and carbon budgets were developed from collected data.

Mixing of lagoon water is generated by semidiurnal tidal currents moving back and forth, primarily through the southern inlet and secondarily through the northern inlet (Jan et al. 2000). The single box model was therefore applied for budget calculation under steady-state assumptions for water and salt balances (Gordon et al. 1996). Because of great temporal variability in freshwater and material delivery into the lagoon (see Table 1), water and nutrient budgets should first be constructed for each sampling period and then averaged for final annual budgets. Such annual budgets are different from those budgets simply derived from final means of various parameters.

The CNP budgets calculated by the two different averaging methods are expected to be different as well. This indicates that system budgets derived simply from mean values of parameters from dry and wet seasons may not be necessarily correct if temporal variability is also significant within a season. Table 2 summarizes temporal distributions of DIP, DOP, DIN and DON for Chiku Lagoon. Using such data, nonconservative fluxes and budgets of CNP were derived according to the guidelines of Gordon et al. (1996) and results were listed in Tables 3 – 5.

Apparently, Chiku Lagoon is an autotrophic system. The net carbon production is approximately 11 mol m^-2 yr^-1. This value is approximately equivalent to 12 % annual gross production (90 mol C m^-2 yr^-1) in the lagoon. The magnitude of net production is quite reasonably consistent with the fact that the system is highly productive.

(nfix-denit) = -3 mol m^-2 yr^-1 – (-0.1×16) mol m^-2 yr^-1 = -1 mol m^-2 yr^-1

The result indicates that Chiku Lagoon is denitrifying at rate of -1 mol m^-2 yr^-1. This value is within the range found previously in the Asian region (Dupra et al. 2000).

Meanwhile, Table 6 demonstrates the difference of system budgets ([p-r] and [nfix-denit]) derived from water and salt budgets using different methods. (p-r) is +58 mole m^-2 yr^-1 and (nfix-denit) is +2.6 mole m^-2 yr^-1 budgeting from temporal means of various parameters.

The implication is that system budgets derived merely from two contrasting seasons (dry and wet) may not be correct if temporal variation is highly significant throughout a year. Some system budgets previously reported were simply developed from mean values of various parameters averaged from dry and wet seasons. Justification and interpretation must be cautious to avoid uncertainty involved in temporal variations.

Figure 2. An example of water and salt budgets modelled from the first sampling event for the Chiku Lagoon. Water flux in 10³ m³ d^-1 and salt flux in 10³ psu-m³ d^-1.