Douglas J. Neilson
SIO-UCSD 0224, La Jolla, CA 92093-0224
dneilson@ucsd.edu

David K. Young, and John C. Kindle
Naval Research Laboratory, Code 7331/Bldg 1007, Stennis Space Center, MS 39529
kindle@nrlssc.navy.mil

Introduction

Bioluminescence intensity (BL) between the sea surface and 200m, was measured off the Omani coast in the Arabian Sea during the 1995 Southwest (SW) Monsoon, 1994 Northeast (NE) Monsoon, and 1995 Fall Intermonsoon (F) seasons. Subsequent analysis indicates the brightest BL, determined by integrating between the surface and 200m, occurred during the NE monsoon and the dimmest during the SW monsoon. This result is contrary to traditional methods used to predict BL.

Attempts to predict bioluminescence intensity over some spatial or temporal scale have traditionally relied on one of two methods: 1.) using historical records from random ship sightings or 2.) assuming BL intensity will be proportional to remotely sensed chlorophyll concentrations. In the first method, predictions are based on a limited number of observations that usually do not reflect the entire region of interest. In the second method, predictions assume a phytoplankton-based origin to the BL (i.e., if not from the phytoplankton then from associated grazers, etc.). This method ignores potential changes in trophic structure, which can occur over very small spatial scales, resulting in the presence or absence of additional, possibly BL, organisms. The first method suffers from insufficient data while the second assumes the trophic structure of the plankton is constant. Both of these methods would have predicted BL to be highest during the SW Monsoon.

This poster presents preliminary results of a modeling effort which uses spatially and temporally changing hydrodynamic and trophodynamic information to predict BL in the Arabian Sea. The SW monsoon creates intense upwelling off the Omani coast which, in turn, provides the nutrients to drive a bloom of diatoms. Numerically dominant, the diatoms result in a shortened food chain. As the SW monsoon fades, the diatoms, having used up available Nitrate or diminished Silicate to a level below that needed for growth, are replaced by smaller autotrophic species capable of prospering at the lower nutrient levels left by the bloom or by recycled nutrients. By the time of the NE monsoon, the plankton community is primarily supported by old production and consists of more links in the food chain. One possible explanation for this seasonal change in BL intensity lies in the succession from a diatom based, relatively non-BL, planktonic community during the SW monsoon to a flagellate based, predominantly BL, community during the NE monsoon. In order to address this question, we modified the existing Naval Research Laboratory (NRL)-developed, coupled NPZD/physical model to include additional compartments to represent separate plankton communities: one supported by upwelled nitrogen, the other by recycled old production.

The Model

An eight component biological model consisting of pools of nitrogen, ammonium, silicate, large autotrophs (diatoms), small autotrophs (dinoflagellates), diatom-based heterotrophs/predators dinoflagellate-based heterotrophs/predators, and detritus was embedded into a three-dimensional ocean circulation model. The physical and biological components of this model as well as the forcing function used in the simulation are discussed below.

Ocean Component

The ocean circulation model is the NRL reduced gravity, thermodynamic, layer-averaged formulation. The model is similar to that used by Young and Kindle (JGR, 1994) and Bruce et al. (JGR, 1994) except that the grid resolution has been increased to 0.25 degrees in latitude and 0.35 degrees in longitude. The model domain includes the entire Indian Ocean north of 30 degrees South. The present version assumes three active layers of 115, 200, and 235m thickness (initial values) within the upper 550m and includes the effects of thermodynamics and inter-layer mixing of momentum, mass and heat. Horizontal density gradients are permitted within each layer and are modified by entrainment, detrainment, advection, diffusion, surface heat flux, and a relaxation to mean density climatology base on Levitus (NOAA Prof. Pap., 1982). The finite difference equations are solved using an explicit scheme on a staggered C grid. Lateral boundaries are located at the 200m contour.

Biological Component

An eight component biological model is fully coupled to the circulation model described above. This model contains two nitrogen pools: upwelled (nitrate) and regenerated (ammonium), silicate, two phytoplankton pools: large (diatom) and small (dinoflagellate), two zooplankton pools: one grazing on each phytoplankton pool, and a detrital pool. Michaelis-Menten kinetics are utilized in all nutrient uptake parameterizations as well as zooplankton grazing. Parameterizations of senescence, detrital sinking, phytoplankton sinking, and detrital remineralization all use linear formulations. Zooplankton death uses a non-linear, Z2 term (Steele and Henderson, J. Plank. Res., 1992). Nitrate uptake is subject to a "silicate switch". Inhibition of nitrate uptake by ammonium is modeled, in both phytoplankton components by a reduction in the maximum uptake of nitrate proportional to the concentration of ammonia present.

Forcing

Nitrate and Silicate are initialized in the second layer only. They are assumed to be invariant and are set at 20mg-at/l N and 15mg-at/l Si. The entrainment of nitrate and silicate from the second to the first layer is activated when the upper layer reaches 65m in thickness. It is recognized that such a formulation neglects important processes operating in the mixed layer, such as entrainment due to wind stirring. The present formulation is intended as a simple parameterization of physical processes in regions of isopycnal outcropping such as the northern Arabian Sea during the SW monsoon. Model results using a more sophisticated mixed layer formulation will be utilized to study the annual cycle.

Sample Plots from Model Year 1992

	JUL 92	AUG 92	SEP 92
DIATOMS
DINOS
SILICATE

The NRL layer model successfully reproduces many features found in the Northern Arabian Sea. Most noticeably, the model reproduces the filaments moving offshore from Oman which are key to the distribution of nutrients and biology into the offshore waters during the SW monsoon. Represented here are snapshots of Silicate, Diatoms, and Dinoflagellates on (l to r) July 18, August 18, and September 18, 1992. Onset of the SW monsoon occurred about June 15. Dinoflagellates are distributed farther offshore than the Diatoms which are generally restricted, in any great numbers, to the coastal zones of upwelling. When the Diatoms do extend offshore, as in September, Dinoflagellates are distributed farther across the same region. Distribution offshore of much of the plankton distribution, can be seen to follow dominant offshore flows. Additionally, there is evidence which suggests the concentrations of Nitrate and Silicate within the filaments are reduced by uptake as the filaments move offshore. Low levels of Diatoms are found across the Northern Arabian Sea by September with the reduction in upwelling that occurs at the end of the SW monsoon.

Although the biological model is still in development and many of its parameters are currently being evaluated, it has been able to reproduce many of the spatial and temporal changes seen during the NRL/ONR cruises in both nutrients and plankton.

BL Decision Matrix

Initial efforts in modeling BL concentrated on identifying regions favorable to BL. A decision matrix was constructed based on levels of Nitrate, Ammonium, and Silicate. BL levels are directly proportional to the number of BL organisms present and, therefore, dinoflagellate (Psmall) and associated zooplankton (Zsmall) will provide a direct indication of the amounts of BL present. However, only the distribution of the three nutrients will provide an indication in the model of the future potential of a region. The decision table correlates patterns of nutrient abundance to BL. Generally, BL is Low only in the absence of Nitrate and Ammonium. Mid levels of BL occur with threshold levels of nutrients or when silicate is sufficiently high to allow diatoms (Plarge) to compete for available resources. Hi BL occurs when nutrients are high enough to allow dinoflagellates to prosper even in the presence of diatoms.

Small symbols represent threshold amounts while large symbols indicate bloom supporting levels. Color indicates relative level of BL and is the same as used in the BL prediction at left.

Model Components , and BL

Using the BL decision matrix, BL potential was determined for August 15, 1992. Since BL is produced solely by Dinoflagellates (Phyto 2) and their associated grazer/predator population (Zoo 2) and these components are primarily dependent on old production, the predicted BL closely matches the spatial distribution of ammonia. However, nitrate and silicate, were also used to determine the relative amounts of BL (Hi, Mid, Low) that could be expected. This prediction represents the potential for BL in the near future. Present amounts of BL can be estimated by comparing spatial distribution and concentrations directly from the Phyto 2 and Zoo 2 snapshots. In this respect, the model correctly indicates that BL would be higher inshore than offshore.

This model, still under development, shows promise in addressing problems associated with the spatial and temporal changes in dominance between an old production-driven trophic dynamic and a new production-driven dynamic. In terms of BL, the model represents the first attempt to relate hydrodynamic and trophodynamic patterns to expected levels of BL. Currently we are running the model for the years 1994 and 1995 which will allow direct comparison of model-generated BL to BL data collected in the Arabian Sea. Immediate additions to the model will include a ninth component, Predators, representing Myctophids with active modeling of diel vertical migration. The biological model is also currently being coupled to a sigma-coordinate level model for the same region.