Gayana 68(2) supl. t.I. Proc. : 161-166, 2004 ISSN 0717-652X

SPATIAL AND TEMPORAL ANALYSIS OF SEAWIFS CHLOROPHYLL IN THE SOUTH TROPICAL PACIFIC OCEAN

C. Dupouy¹, J. Neveux² & A. Le Bouteiller³

1. LODYC, IRD, BP A5, 98848 Noumea cedex, New Caledonia
2. LOOB, UMR CNRS, BP 44, 66651 Banyuls-sur-mer, France
3. IRD Centre de Nouméa, Nouvelle Calédonie PORSEC 2004- Conception (Chili), 29 nov-3 dec 2004

ABSTRACT

The South West Tropical Pacific (SWTP) is generally considered as an N-limiting oligotrophic area with parts characterized by permanent deep thermocline and nutricline as the Melanesian Archipelago. However, ocean color imagery shows in this region a strong seasonal chlorophyll cycle, with large blooms which have for long been attributed to nitrogen-fixing cyanobacteria as Trichodesmium. SeaWiFS surface chlorophyll analysis of the area between 5°S-25°S, 150°E-170°W has been conducted from 1997 to 2004. A strong seasonal and interannual variability is described, which relates well with the sea-truth measured chlorophyll concentrations. An empirical orthogonal function analysis helped in determining principal explicative factors of this variability, and in identifying different patterns, i.e., seasonality (austral winter and summer blooms), which may be driven by different fertilization processes, and a unexplained interannual variability. This work represents part of the French PROOF-IRD DIAPAZON (DIAzotrophy in a PAcific ZONe) program aiming at estimating nitrogen and carbon fixation by cyanobacteria at the SWTP scale.

Key words. SeaWiFS. Trichodesmium. South Western Tropical Pacific Ocean. Solomon, New Caledonia, Vanuatu and Fiji-Tonga archipelagos. Statistical EOF analysis. Oceanic nitrogen fixation.

INTRODUCTION

In the South Western Tropical Pacific ocean (SWTP), strong seasonal and inter-annual variabilities in the chlorophyll concentration were observed using CZCS (Dupouy et al., 1988, Dupouy, 1992), and lastly POLDER and SeaWIFS. Though biomass levels are generally low in the SWTP, the colonial Trichodesmium is suspected to play a major role in the intermittent increases of phytoplankton biomass, especially during summer (Dupouy et al., 2000). Estimating dinitrogen fixation from ocean color satellite data is a major challenge, especially for oligotrophic tropical oceans, where N₂ fixation is potentially a major source of « new N » (Capone et al., 2001). The origin of this archipelago's bloom, which is detected by the satellite in an oligotrophic ocean is still in question (Wilson and Adamec, 2001, Wilson, 2003) and the question arises to know whether these blooms are related to nitrogen fixing organisms or not. The French PROOF-DIAPAZON program was set to answer this question and cruises were organized in the region between 2001 and 2003, to measure nitrogen fixation, examine causes and consequences of Trichodesmium blooms, and determine its optical signature. In this paper, we examine the variability of large scale chlorophyll patterns detected by SeaWiFS between 1997 and 2004 in the South Western Tropical Pacific ocean in order to better understand the general enrichment patterns in the region.

MATERIAL AND METHODS

The region of study is defined between 150°E and 170°W, 5°S and 25°S, an area of 20 degrees of latitude and 40 degrees of longitude. It includes four major archipelagos (Solomon, New Caledonia, Vanuatu and Fiji-Tonga). Data were extracted using seadas subroutines from global 8-day level-3 SeaWiFS chlorophyll composites provided by the NASA/GSFC (version 4), so that each image is 360 pixels x 180 pixels size, with a spatial resolution of 11 km x 11 km (9 pixels for one degree). Cloud gaps were filled by an interpolation procedure over a maximum distance of 100 km (10 pixels, calculated from a variogram on some cloud-free images (Matheron, 1963)). The processing eliminated over-threshold pixels, and land mask. SeaWiFS pixels were binned (i.e. 1/3 degree) to match the coarse spatial resolution of usually available parameters. As a result of the developed softwares (smalltalk), Howmöller diagrams, or extraction of SeaWiFS chlorophyll values at any location is possible. Variance, and monthly/annual means were calculated. Empirical orthogonal function (EOF) analysis was used as a tool to decompose spatial and temporal variance of the SeaWiFS observations as in Murakami et al. (2000); Wilson and Adamec (2001); Yoder et al.(2002). The EOF analysis (MATLAB) was performed on SeaWiFS data on a final grid element size (x = y = 1°, t= 8 days).

RESULTS

The mean SeaWiFS chlorophyll map (Fig. 1c) shows that relatively rich chlorophyll waters (chla >0.1 mg m-3 ) are found around Solomon Islands, and between New Caledonia and Vanuatu I. Weaker enrichments are found around Fiji or Tonga. In the north-eastern corner of the area, the southern part of the chlorophyll-rich equatorial upwelling is visible. Chlorophyll is maximal in winter (June-July) at the south-west of New Caledonia (Fig.1a). Summer maxima reach their maximal intensity in February and are confined to the archipelagos of Vanuatu and New Caledonia (Fig. 1b). Chlorophyll increases are also intermittently observed around Fiji.

Figure 2 shows the SeaWiFS chlorophyll cycle in the center of the Loyalty Channel, 30 nautical miles from the coasts (DIAPAZON cruise's main station at 21.5°S, 167°E, see red circle on the map Fig 1d). The seasonal and inter-annual variability observed with the SeaWiFS chlorophyll series is high. Its maximum concentration (0.12 to 0.25 mg.m^-3) appears every winter (July). Summer peaks are irregularly observed in 1998, 1999, then they disappear in 2000, 2001 and 2002 and re-appear later in 2003 (December 2002) and 2004.

The first three EOF modes explained 80% of the total variance in chlorophyll (Fig. 3, Table 1). The first mode explained 43% of the total variance (Table 1). Its spatial structure is clearly oriented as a dipole on the north-west/south-east axis, and shows a sharp opposition between the south-west part (positive anomaly) and the north-east part (negative anomaly) of the South Western tropical Pacific. The EOF mode 1 shows a regular annual cycle with a positive anomaly of chlorophyll concentration in July-August. The EOF mode 2 (21%) shows two distinct patches of positive chlorophyll anomaly. One is between New Caledonia and Vanuatu (centered at 15°S) and the other is west of Solomon Is. (centered at 7°S). The EOF mode 2 shows a very irregular cycle with positive chlorophyll anomalies found in March 1998 and 1999, and in March 2003 and 2004. The third mode exhibits (16%, not shown) irregularly peaks in March and rich patterns are seen at the east of the Vanuatu and Fiji Is.

During the nine 2001-2003 diapalis (DIAPAZON program) cruises on the east coast of New Caledonia, sea-truth surface chlorophyll were in accordance with SeaWiFS data. The lowest values ( 0.09-0.12 mg.m^-3) were observed during the first 5 diapalis cruises (October to May 2002) and during the 2 last ones (diapalis 8 and 9 in June and October 2003). Maxima of chlorophyll biomass (0.3 mg.m^-3) were observed in winter (diapalis 6) and in summer (diapalis7). The summer biomass maximum observed in February 2003 was associated with a Trichodesmium contribution of 50%.

Metre a, b, c, et d sur la figure

Figure 1: Spatial patterns detected by SeaWiFS (NASA/GSFC) in the South Western Tropical Pacific Ocean in winter (a, JULY) and summer (b)FEBRUARY). (c) Mean chlorophyll from SeaWiFS (1997-2004). (d) Map indicating archipelagos and islands. The red circle indicates the site of the DIAPAZON cruises (Loyalty Channel).

Figure 2. Seasonal and inter-annual variability in surface chlorophyll extracted from SeaWiFS (NASA/GSFC) 8-D composites (dark blue, raw data; light blue, filtered data) at the DIAPAZON station, in the middle of the Loyalty Channel (see red circle on the map Fig.1d). Summer periods (November to March) with high temperatures are highlighted in brown. Diapalis cruise periods are indicated in red. Winter chlorophyll peaks appear regularly each year in July from 1998 to 2003. Summer peaks appear in January 1998, January 1999, are reduced in 2000, 2001 and reappear in 2003 and 2004.

Figure 3: Results of the EOF analysis of surface chlorophyll as seen by SeaWiFS in the box 5°S-25°S, 150°E-170°W, including archipelagos of (from north to south): Solomon Is. (5-10°S), New Caledonia and Vanuatu Is. (see map) Fig. 1d). Spatial patterns (top) and temporal evolution for each mode.

Table 1: Statistical results of the EOF analysis of the three parameters

DISCUSSION

Before the DIAPAZON program, biogeochemistry data in the region were scarce (Dandonneau and Gohin, 1984; Dandonneau and Lemasson, 1987; Blanchot et al., 1992; Radenac and Rodier, 1996). Between 15°S and 21°S, thermocline and nutricline have unimodal distributions with respective means at 130 and 120 m. The nutricline depth responds to the variability of the subsurface thermal structure of the SWTP (Delcroix and Hénin, 1989; Delcroix and Lenormand, 1997) which is related to the zonal displacements of the SPCZ and ITCZ (South Pacific and Inter Tropical Pacific convergence zones of clouds). Along the Revelle/98 cruise track, the deep chlororophyll maximum oscillated between 120m in the central Fijian Basin (175°E, 20°S) up to 50 m in the Vanuatu waters (170°E, 17°S) and reached the maximum level of 130m in Tonga waters. From 20 to17°S and along 165°E, during "normal" conditions, the nutrient reservoir resides in the lower thermocline, no nutrient at the surface were ever reported, and the mean chlorophyll concentration was 0.07 mg m^-3. At the opposite, in winter, surface chlorophyll can double at these latitudes (Radenac and Rodier, 1996). The DIAPAZON results at the surface of the Loyalty Channel station (where the deep chlorophyll maximum was at about 100 m) confirmed the SeaWiFS observations, that except in winter (August 2002, Diapalis 6) and in summer when Trichodesmium is dominant (February 2003, Diapalis 7), chlorophyll are below 0.12 mg.m^-3.

Different mechanisms may be at the origin of chlorophyll enrichments in the SWTP. In winter (June to August), the decrease in solar irradiance and daylight-length could enhance the convective mixing, i.e. the deepening of the mixed layer and the erosion of the nutricline over several tens of meters, bringing nutrients to the surface (Dandonneau and Gohin, 1984). This would happen preferentially in the southern part of the SWTP in August where Ekman pumping is favorable to upwelling (Delcroix and Hénin, 1989; Delcroix and Lenormand, 1997; Gouriou and Delcroix, 2002). This mechanism could be responsible of the winter maxima observed each year in the western part of SWTP (EOF mode 1). In summer (November to March), the wind stress can be favorable to downwelling (Dandonneau and Gohin, 1984), limiting nutrient inputs and lowering productivity at the surface, and this would happen in the whole SWTP in February or in the northern part of the area (north of 10°S). These severe oligotrophic conditions disadvantage the new production of phytoplankton based on NO₃ and favour new production by nitrogen-fixing organisms and Trichodesmium. Hansell and Feely (2000) concluded that elevated total organic nitrogen observed in the SWTP must originate from N₂ fixation, favoured by a high column stability under the SPCZ and ITCZ. This mechanism could be responsible of the summer maxima (EOF mode 2 and 3). At 10°S only, and during the 1987 El Nino event, a sea surface chlorophyll maximum (> 0.2 mg.m-3) related to a local input of nitrate (NO3 > 0.5 µM) was observed (Blanchot et al., 1992; Radenac and Rodier, 1996). This resulted of Ekman pumping as visualized by an upward of isotherms. This might be also the explanation for the 7°S bloom (EOF mode 2).

It is not known if the influence of the numerous island masses have an impact on surface productivity in the SWTP ocean as reported in the Pacific Ocean equatorial region (Signorini et al., 1999; MacClain et al., 2001). Topography in the studiedarea is such that the bottom is most often over 3000 meters deep. Islands have different geological origin and composition, the majority of which without continental shelves, except the large old plateau of Fiji Island. There are no large river inputs in the SWTP. The intensity of the geostrophic flow is weak (maximal value of 7 cm s^-1) and variable south of 15°S and will carry mostly warm and oligotrophic waters from the tropical gyre through the SWTP. Island mass effects could result from currents generated by reflection of main eastward currents coming from the east and inducing upwelling or downwelling eddies as detected by thermal satellite imagery at Hawaii (McGillicudy and Robinson, 1997) but these points have not been studied yet in the region. Hazardous submarine volcanic activity may cause transient nutrient enrichments, at least in iron or other limiting factors (Mackey et al., 2002) and could provoke chl a increases. The volcanic activity is related to the Tonga trench (Michel Lardy, pers. comm.) and abundant pummices may travel westward from Tonga to Vanuatu and New Caledonia. Iron concentration seems high in the northern and southern western Pacific ocean (Duce and Tindale, 1991) and sufficient to supply cyanobacterial nitrogenase activity. The computed mineral aerosol deposition rate measured from dust winds and rains during the SEAREX experiment (Prospero,1996) at the New Caledonia station is low but near the average of the whole Pacific network (mean of 0.3 µg m^-3 dust concentration). Unfortunately, no synoptic data are available yet in the South Pacific (Husar et al., 2001) though recent data have been collected during DIAPAZON.

CONCLUSIONS

SeaWiFS and pigment analyses collected during the DIAPAZON program have shown that the SWTP is a non typical N-limiting oligotrophic ocean. A strong variability has been observed with SeaWiFS, which can be decomposed into a classical annual cycle, i.e. a winter chlorophyll enrichment in response to the seasonal cooling (mode 1), and ­summer enrichments related to diazotroph cyanobacteria localized in the archipelagos of Vanuatu and New Caledonia or Fiji. This last hypothesis is reinforced by our recent observations that, even if Trichodesmium is present all the year in the region, it peaks only during the warm season. The strong inter-annual variability of the summer abundancess observed since 1997 with SeaWiFS explains the strong variability observed during the DIAPALIS cruises in the Loyalty Channel. It is probably related to variable inputs of macronutrients (iron?) brought to the SWTP from dust or changes in the general circulation.

AKNOWLEDGEMENTS

We thank LODYC-IPSL/IRD, the French-PROOF program and CNRS for their support to the DIAPAZON project at IRD Noumea. We thank the SeaWiFS Project and the Distributed Active Archive Center at NASA/GSFC for delivering the SeaWIFS data.

REFERENCES

Blanchot J., Le Bouteiller A. & M. Rodier (1992) Effect of El Nino Southern Oscillation events on the distribution and abundance of phytoplankton in the Western Pacific Tropical Ocean along 165°E. Journal of Plankton Research, 14, 137-156. [         [ Links ]1]

Capone D. (2001) Marine nitrogen fixation : what's the fuss? In Current Opinion in Microbiology, 4, 341-348. [         [ Links ]2]

Dandonneau Y. & F. Gohin (1984) Meridional and seasonal variations of the sea surface chlorophyll concentration in the southwestern tropical Pacific (14 to 32°S, 160 to 175°E). Deep Sea Research, 31, 1377-1393. [         [ Links ]3]

Dandonneau Y. & L. Lemasson (1987) Water-column chlorophyll in an oligotrophic environment : correction for the sampling depths and variations of the vertical structure of density, and observation of a growth period. Journal of Plankton Research, 9, 215-234. [         [ Links ]4]

Delcroix T. & C. Hénin (1989) Mechanisms of subsurface thermal structure and sea surface thermohaline variabilities in the southwestern tropical Pacific during 1975-85. Journal of Marine Research, 47, 777-812. [         [ Links ]5]

Delcroix T. & O. Lenormand (1997) ENSO signals in the vicinity of New Caledonia, South Western Pacific. Oceanologica Acta, 20, 3, 481-491. [         [ Links ]6]

Delcroix T., Gourdeau L. & C. Hénin (1998). Sea surface salinity changes along the Fiji-Japan shipping track during the 1996 La Nina and 1997 El Nino period. Geophysical Research Letter, 25, 3169-3172. [         [ Links ]7]

Duce R. A, & N. Tindale (1991) Atmospheric transport of Fe and its deposition in the ocean. Limnology and Oceanography 36: 1715-1726. [         [ Links ]8]

Dupouy C., Petit M., & Dandonneau Y., 1988. Satellite detected cyanobacteria bloom in the southwestern tropical Pacific. Implication for nitrogen fixation. International Journal of Remote Sensing, 8, 3, 389-396. [         [ Links ]9]

Dupouy C., 1992. Discoloured waters in the Melanesian archipelago (New Caledonia and Vanuatu). The value of the Nimbus-7 Coastal Zone Colour Scanner observations (14 p., 2 planches noir et blanc); in "Marine Pelagic Cyanobacteria: Trichodesmium and other diazotrophs", Carpenter, Capone and Rueter (eds.), Kluwer Academic Press, NATO ASI Series C: Mathematical and Physical Sciences - Vol. 362 : 177-191. [         [ Links ]10]

Dupouy C., Neveux J., Subramaniam A., Mulholland M., Campbell, L., Montoya J., Carpenter E. & D. Capone, 2000. Satellite captures Trichodesmium blooms in the Southwestern Tropical Pacific. EOS Transactions AGU, 81, pp. 13,14-16. [         [ Links ]11]

Hansell D. A. & R. A. Feely (2000) Atmospheric intertropical convergence impacts on surface ocean carbon and nitrogen biogeochemistry in the western Pacific. Geophysical Res. Letters, 27, 1013-1016. [         [ Links ]12]

Husar et al., 2001. Asian dust events of April 1998. Journal of Geophysical R.esearch 106, D16, 18317-18330. [         [ Links ]13]

McClain C., Christian J. R., Signorini, S. R., Lewis M. R., Asanuma I., Turk D., & C. Dupouy (2002). Satellite ocean color observations of the Pacific Ocean. Deep Sea Research 49, 13-14, 2533-2560. [         [ Links ]14]

McGillicudy & R. A. Robinson, (1997). Eddy-enhanced nutrient supply and new production in the Sargasso Sea. Deep-Sea Res., 44, 1427-1450. [         [ Links ]15]

Prospero, J.M. (1996) The atmospheric transport of particles to the Ocean. In Particle Flux in the Ocean, edited by V. Ittekkot, P. Schäfer, S. Honjo and P.J. Depetr SCOPE Report 57, John Wiley & Sons, Chichester, pp. 19-52. [         [ Links ]16]

Matheron G. (1963) Principles of statistics. Economic Geology, 58, 1246-1266. [         [ Links ]17]

Murakami,H., Ishizaka J. & H. Kawamura (2000). ADEOS observations of chlorophyll a concentration, sea surface temperature, and wind stress change in the equatorial Pacific during the 1997 El Niño onset. J. Geophys. Res. 105(C8): 19551 - 19559. [         [ Links ]18]

Radenac M. H. & M. Rodier, 1996. Nitrate & chlorophyll distributions in relation to thermohaline and current structures in the western tropical Pacific during 1985-1989. Deep Sea Res. 43, 4-6, 725-752. [         [ Links ]19]

Signorini, S., McClain C. R., & Y. Dandonneau, 1999. Mixing and phytoplankton bloom in the wake of the Marquesas Islands. Geophys. Research Letter, 26, 3121-3124. [         [ Links ]20]

Wilson C. & D. Adamec (2001). Correlations between surface chlorophyll and sea surface height in the tropical Pacific during the 1997-1999 El Nino-Southern Oscillation event. Journal of Geophysical Research, 106, 31175-31188. [         [ Links ]21]

Wilson, C., 2003. Late summer chlorophyll blooms in the oligotrophic North Pacific Subtropical Gyre. Geophys. Research Letter 30, 18, 1942, doi:10.1029/2003GL017770. [         [ Links ]22]

Yoder J. A., Schollaert S. E. & J. E. O'Reilly (2002) Climatological phytoplankton chlorophyll and sea surface temperature patterns in continental shelf and slope waters off the northeast U. S. Coast. Limnology and Oceanography, 47, 672-682. [         [ Links ]23]