Forecasts of possible phytoplankton responses to elevated riverine nitrogen delivery into the southern firth of Thames
Author: Niall Broekhuizen, John Zeldis (NIWA)
Land use change in the catchment of the upper catchments of the four large rivers draining through the Hauraki plains is expected to lead to there being a greater delivery of nitrogen into the south-eastern Firth of Thames. Environment Waikato contracted NIWA to make forecasts of how increased nitrogen concentrations in the rivers feeding into the southern Firth of Thames might influence the likelihood that high standing stocks of phytoplankton would develop.
After preliminary discussions, it was agreed that NIWA would undertake two pieces of work. Firstly, existing measurements of depth-specific light intensity (for the northern and central Firth of Thames) would be analysed with a view to developing an empirical predictor of the attenuation coefficient for photosynthetically active radiation (PAR), for example, as a function of water-column depth and salinity. Secondly, NIWA's existing spatially explicit biophysical model (which simulates the dynamics of nutrients, organic detritus and three phytoplankton taxa) would be modified to incorporate the results of the attenuation analysis. The model would then be used to make a total of nine simulations (three nutrient-loading scenarios in each of three different months).
The indicative budget for the project precluded any sensitivity analyses - though this report includes a very limited one. The simulation results were to be used to infer the consequences (for phytoplankton in the southern Firth) of increased nutrient loading. A nuisance-phytoplankton concentration threshold of 10 mg chl a m-3 was defined as one standard against which to assess the influence of increased nutrient loading.
The three nutrient-loading scenarios were:
- baseline (present situation)
- two-fold and five-fold' - being respectively two fold or five-fold increases in the concentrations of dissolved inorganic and total organic nitrogen passing down the Waihou, Waitoa, Piako and Ohinemuri rivers.
Simulations were made under physical conditions (irradiance, riverflow, wind-conditions and hence current vectors, water temperatures and salinities) corresponding to September 1999, March 2000 and May 2003.
In simple regressions, the inferred light attenuation coefficient was found to be significantly correlated with water-depth (declining with increasing water-depth) and salinity (declining with increasing salinity). There was also a much weaker, positive relationship with chlorophyll concentration. There are significant cross-correlations between these explanatory variables, but in a multiple regression, both depth and salinity proved to be significant (P<<0.05).
The resulting regression-model predictor of light attenuation was incorporated into the biophysical model. It proved necessary to perform an informal recalibration of the model such that it would produce plausible phytoplankton concentrations in the northern Firth (for which we have observational data).
The recalibrated model predicts high phytoplankton concentrations in the southern Firth (ranging from a space-time average of ~ 6 mg chl a m-3 in May 2003 to in excess of 40 mg chl a m-3 in March). These concentrations are several times greater than are usually observed in the northern (cf southern) Firth of Thames.
In the turbid, shallow, nutrient-rich, north-eastern Manukau harbour, chlorophyll concentrations are within the range 2-10 mg chl a m-3 for most of the year, but sporadically rise to more than 60 mg chl a m-3 during the summer months (Williamson et al. 2003). In the less enriched south-eastern Manukau, chlorophyll concentrations rarely exceed 10 mg chl a m-3.iven
- the relatively large area over which we are calculating average chlorophyll concentrations and,
- the prolonged duration of the periods of high simulated chlorophyll
we suspect that, despite yielding plausible phytoplankton stocks in the northern Firth of Thames, our biophysical model is over-predicting 'basal' stocks (cf. short-lived 'blooms') in the very shallow, southern Firth. We describe several shallow-water-processes (absent in the present model) that would serve to reduce shallow-water standing stocks.
The model suggests that the system was not nitrogen limited in either September 1999 or May 2003. Thus, increased riverine nitrogen loads had little influence upon phytoplankton concentrations in those months. In contrast, for the March 2000 simulations, chlorophyll concentrations in the vicinity (to approximately 5 - 10 km) of the Waihou river-mouth are predicted to increase by 50 - 100 per cent in the two-fold scenario and approximately 300 per cent in the five-fold scenario. Diatoms and phytoflagellates increase much more dramatically than do dinoflagellates.
Whilst we are doubtful that phytoplankton abundances as high as those forecast would persist for long periods, we believe that the forecasts provide an indication of the 'worst-case' conditions that might arise sporadically. In the final section of this report, we identify several items of additional field- and simulation-work that would help to refine the forecasts that are presented within this report.
Download a copy of the report below:
|2.1 Estimation of light attenuation in the Firth of Thames||4|
|2.2 Phytoplankton modelling||5|
|2.2.1 Grid resolution||6|
|2.2.2 Hydrodynamic scenarios||6|
|2.2.3 Implementation of PAR attenuation||7|
|2.3.3 Matamata-Piako district||6|
|2.2.4 Boundary conditions, nutrient loading scenarios and initial conditions||10|
|2.3 Analysis of model results||11|
|3.1 PAR Attenuation||14|
|3.2 Phytoplankton modelling||19|
|3.2.1 Model recalibration||19|
|3.2.2 Simulation results: Effects of increased nutrient loading to the southern Firth||21|
|3.2.3 Robustness analysis||29|
|4.1 Plausibility of results: Shallow water processes absent from the model||32|
|4.2 Caveats: Other simplifying assumptions||35|
|4.3 Seasonality of response to loading||36|
|4.4 Robustness of conclusions||37|
|4.5 Further work||38|
|Appendix A: Point source nutrient loads (baseline scenario)||43|
|Appendix B: Summary illustrations of the results from the individual cruises from which light-attenuation was inferred||47|