Gayana 68(2): 381-384, 2004

S-CHLOROPHYLL SQUIRTS IN THE CHILEAN COAST: A SEAWIFS PERSPECTIVE

Víctor H. Marín & Luisa E. Delgado

Laboratorio de Modelación Ecológica, Departamento de Ciencias Ecológicas, Facultad de Ciencias, Universidad de Chile. Casilla 653, Santiago, Chile E-mail: vmarin@antar.uchile.cl

ABSTRACT

Squirts can be defined as one-way jets, transporting coastally upwelled water to the deep ocean and terminating in a counterrotating vortex pair. These mesoscale oceanographic structures can export coastal biological production up to 200 km offshore. We describe, based on the analysis of SeaWIFS images, a squirt located in the northern Chilean coast (30° S). Our results show that the spatial structure of the velocity field, and the hammerhead shape, agree with previous conceptual models. We used the circulation data, resulting from feature-tracking analysis, to estimate the offshore export of coastal carbon production. We conclude that squirts represent a net carbon exporting mechanism accounting for 5% to 12% of coastal carbon production.

INTRODUCTION

The biological production of the world oceans is especially high in their margins (Longhurst, 1998). One of these margins, the Humboldt Current Coastal Province (HUMB), is one of the most productive marine systems. Most of its production is sustained by coastal upwelling which pumps nutrients from waters deeper than 60 m into the euphotic zone (Mann & Lazier, 1991; Longhurst, 1998). Within the HUMB, coastal ecosystems are bounded, offshore, by the coastal transition zone (Hormazabal et al., 2004). This is a 600-800 km zone extending offshore from the coast and characterized by high meso-scale eddy activity. High-chlorophyll filaments are one of the most common coastal meso-scale structures (Thomas, 1999; Sobarzo and Figueroa, 2001; Marín et al., 2003a). These filaments have been proposed as playing a significant role in the spatial structure of the primary productivity of coastal waters, in the distribution of coastal zooplankton species and in the recruitment of coastal benthic invertebrates (Rothschild, 1988, Marín and Moreno, 2002, Marín et al., 2003a; Marín et al., submitted).

Strub et al. (1991) have proposed three conceptual models for coastal filaments: squirts, mesoscale eddies and meandering jets. Since the three types have different spatial structures, and generate surface patterns either in the temperature or the colour fields, they can be identified using remote sensing methods. Squirts are defined as: one-way jets, transporting coastally upwelled water to the deep ocean, terminating in a counterrotating vortex pair, with a shape in the sea surface temperature and pigments fields often referred to as mushroom, hammerhead or T. Since squirts tend to occur in specific places within the coastal region (e.g. between prominent coastal points), they can be important mechanisms for the recurrent offshore export of coastal pelagic organisms. Strub et al. (1991) propose that squirts may be detected through either SST or pigments fields. However, our analysis of SST images from the Chilean coast had never shown them. Our first record of such structures came from the analysis of images from the Sea­viewing Wide-Field-of-view Sensor (SeaWIFS), (Fig. 1). We then decided to study the dynamics of squirts analyzing SeaWIFS images from an area around 30°S off the Chilean coast. The main goal of this work is to describe our findings and to discuss the ecological effects of squirts for coastal marine ecosystems.

Figure 1: S-chlorophyll squirts off the Chilean coast. The picture, centred at 34°S corresponds to a L2 SeaWIFS image from January 7, 1998. The Black sector in the upper left corner represents clouds; the arrow points to the head of the main squirt.

METHODS

We looked for clear-sky SeaWIFS images from the Coquimbo area (30°S) using NASA's internet server (http://daac.gsfc.nasa.gov/data). Since our intention was to relate our findings to other field collected data (see http://antar.uchile.cl), we concentrated our efforts in two periods: January 1999 and 2002. We obtained four L1-A images (resolution = 1 km), all showing a squirt (Fig. 2). Chlorophyll-a was estimated using the OC4 algorithm (SEADAS 4.0 program, freely available from NASA). Since we did not calibrate images against chlorophyll-a samples, we utilize the term s-chlorophyll to refer to the Seadas-generated chlorophyll-a values. We utilized daily consecutive images to estimate the surface flow field, analysing the spatial displacement of s-chlorophyll features. This feature-tracking analysis has been used to study coastal circulation in many areas and it has been described in detail by Marín et al (2003b). Flow fields were generated using kriging for the objective mapping of u and v velocity components [Jongman et al., 1995]. Components were calculated from measured displacement's angles and distances (polar to rectangular transformation) for each pair of images. The data, including the spatial position of each pair, was exported into a geostatistical package (GS 5.0 for Windows) where both the semivariograms (ã(h)), for lag intervals (h) of 5.0 km, and the kriged surfaces were estimated. The selection of the semivariogram model (from the five models available) was done using the largest r² as criteria. In all cases a linear semivariogram generated the best fit,

(1)

where C0 is nugget variance (0 in this case), C is structural variance (0.01), and A0 is the range parameter (67 km). The interpolation was done using the adjusted semivariogram models on a 2 x 2 block kriging, for a 4-km cell size, an omnidirectional searching neighbourhood of 12 data points, and an irregular mask (polygon) to prevent interpolating outside the limits covered by the data.

RESULTS

The concentration that best delimited the squirt was 0.7 mg m^-3 for 1999 and 0.5 mg m^-3 for 2002. Nixon and Thomas [2001] have proposed that a chlorophyll threshold >1.0 mg m³ can be used to define the limits of productive habitats in coastal upwelling ecosystems. If we had used such a limit, for the observed squirt, most of its structure would have been lost. The 1999 squirt was longer (141 km along its main axis) that that from 2002 (125 km), although the spatial location was the same (Fig. 2).

Figure 2: S-chlorophyll squirt off the Chilean coast (30°S) recorded the 20th of January 1999.

Flow direction, as estimated by the feature-tracking analysis, was predominantly equatorward and onshore for 1999 [<u> = 0.02 m s^-1, <v> = 0.02 m s^-1], with a vortex at the tip of the squirt located closest to the coast. The flow direction for 2002 was mainly poleward and offshore [<u> = -0.09 m s^-1, <v> = -0.04 m s^-1], again showing a vortex-like structure at the tip of the squirt located closest to shore. In both cases the flow along the main axis of the squirt was directed offshore. Thus the structure of the velocity fields agrees with the conceptual model proposed by Strub et al. [1991].

DISCUSION

Squirts may have a strong influence on the ecological settings of coastal ecosystems. Strub et al. [1991] state that biological implications of this type of structure are mainly related to the nutrient enrichment of the deep ocean. Indeed, our data shows that squirts transport chlorophyll-reach water (1.0 mg m^-3) into low-chlorophyll areas (<0.1 mg m^-3). However, what percentage of the costal carbon production do squirts transport? We used our surface current estimates, from the feature-tracking analysis, and literature data on biological production in order to assess the offshore transport. We used an average coastal production of 1000 mg C m^-2 d^-1 and an average ocean production of 500 mg C m^-2 d^-1 (Montecino et al., 1996; Montecino & Quiroz, 2000 and Pizarro et al., 2002). The vertical extent of the squirts (10 ­ 20 m) was deduced from data on upwelling filaments available from Mejillones Bay, 23°S (Marín et al., 2001). If we use the maximum estimated offshore speed (-0.2 m s^-1) within a 20-km-wide squirt (Fig. 2), then the carbon export would amount to an 11.5% of the daily carbon production within the ecosystem or some 1.6 x 10³ metric tonnes C d^-1. Conversely, if we assume that the counterrotating vortex at the tip of the squirt together with the shift in the cross-shelf flows during spin-down of coastal upwelling (onshore flow) represent a likely carbon import mechanism for the coastal ecosystem, this would be of O (390 metric tonnes C d^-1) for an onshore speed of 0.1 m s^-1, which represents only 3% of the exporting capacity of the ecosystem. Thus squirts represent a net exporting carbon mechanism capable of exporting, under the analyzed conditions, between 5% to 12% of the coastal carbon production.

Squirts are generated by nearshore convergences such as those caused by local wind relaxations around capes (Strub et al., 1991). Our observations in the Humboldt Current agree insofar as their potential locations (Figs. 1 and 2). Given that capes and protected bays are characteristics of the Chilean coast, the likely presence of squirts along the coast is high. However, why there are no records about them in the literature? Our analysis shows that the main spatial features that characterize squirts (e.g. hammerhead shape) are not visible through sea surface temperature (i.e. NOAA) satellite images. And SST is, indeed, the most common remotely sensed variable for the Humboldt Current. Surface colour (e.g. SeaWIFS) images are not as ready available as SST images. First, the generation of secondary products such as s-chlorophyll (L2 images), has to be done through the SEADAS program. SEADAS does not run, yet, on Windows environments. It is only available for Linux and UNIX systems. Even if L2 images are available, its geographic rectification and subsequent export for more comprehensive analysis (e.g. export to a Geographic Information System) is rather cumbersome. Finally, since the SeaWIFS is a passive sensor and it covers a given sector once every three days (in order to have images at Nadir), the likelihood of clear-sky consecutive images is low. Thus, SeaWIFS images have been used mostly for large-scale analysis where many partial images are mixed loosing meso-scale details (e.g. Carr et al., 2002). We suggest, based on our results, that SeaWIFS images, when analyzed as individual scenes and/or consecutively, may provide important information about the dispersive characteristics of coastal ecosystems.

ACKNWLEDGMENTS

This work was financed by Proyecto FONDECYT-Chile 1040891 awarded to Víctor H. Marín.

REFERENCES

Carr, M.E., P.T. Strub, A. C. Thomas & J.L. Blanco (2002) Evolution of 1996-1999 El Niño condiciones off western coast of South America: A remote sensing perspective. Journal of Geophysical Research-Oceans 107 (C12): Art. No. 3236 DEC 31 2002. [1]

Hormazabal S, G Shaffer & O Leth (2004) Coastal transition zone off Chile. Journal of Geophysical Research. 109, C01021, doi:10.1029/2003JC001956,2004. [2]

Jongman, R. H. G., C. J. F. Ter Braak & O. F. R. Van Tongeren (1995). Data analysis in community and landscpate ecology. Cambridge University Press, London. [3]

Longhurst, A. (1998) Ecological geography of the sea. Academic Press, New York. [4]

Mann, K. H. & J. R. N. Lazier (1991) Biological-physical interactions in the oceans. Blackwell Sci., Malden, Mass. [5]

Marín, V., R. Escribano, L. Delgado, G. Olivares, & P. Hidalgo (2001). Nearshore circulation in a coastal upwelling site off the northern Humboldt Current system. Continental Shelf Research., 21, 1317­ 1329. [6] [7]

Marín, V. H., L. E. Delgado, G. Luna-Jorquera & V. Montecino (submitted). La surgencia costera y su efecto en la estructura espacial de los ecosistemas pelágicos en la Corriente de Humboldt. Parte I: Un modelo conceptual. Revista Chilena de Historia Natural. [8]

Marín, V. H.; L. E. Delgado & R. Escribano (2003a) Upwelling shadows at Mejillones Bay (Northern Chilean COSAT): a remote sensing in situ analysis. Investigaciones Marinas, Valparaíso 31:47-55. [9]

Marín, V. H.; L. Delgado & G. Luna-Jorquera (2003b) S-cholorophyll squirts at 30° S off the chilean coast (eastern South Pacific): Feature-tracking analysis. Journal of Geophysical Research, Vol. 108, N°. C12, 3378, doi:10.1029/2003JC001935, 2003. [10]

Montecino, V., & D. Quiroz, Specific primary production and phytoplankton cell size structure in an upwelling area off the coast of Chile (30°S). Aquatic Science, 62, 1 ­17, 2000. [11]

Montecino, V., G. Pizarro, & D. Quiroz (1996). Dinámica del fitoplancton en el sistema de surgencia frente a Coquimbo (30°S) a través de la relación funcional entre fotosíntesis e irradianza (P-I), Gayana Oceanol., 4, 139­151. [12]

Nixon, S., & A. Thomas (2001). On the size of the Peru upwelling ecosystem. Deep Sea Res., Part I, 48, 2521­ 2528. [13]

Pizarro, G., J. L. Iriarte, & V. Montecino (2002). Mesoscale primary production and bio-optical variability off Antofagasta (23­ 24_S) during the transition to El Niño 1997­ 1998, Rev. Chilena Hist. Nat., 75, 201­ 215. [14]

Rothschild, B. J. (1988) Toward a theory on biological-physical interactions in the world ocean. NATO ASI Series, Series C: Mathematical and Physical Sciences- Vol. 239, Kluwer Academic Press., Dordrecht. [15]

Sobarzo, M. & D. Figueroa (2001). The physical structure of a cold filament in a Chilean upwelling zone (Península de Mejillones, Chile, 23S). Deep-Sea Res. 48, 2699-2726, 2001. [17]

Strub, P.T., P. M. Kosro & A. Huyer (1991). The nature of the cold filaments in the California Current System. Journal of Geophysical Res. 96. 14743- 14768. [18]

Thomas, A. C. (1999) Seasonal distributions of satellite-measured phytoplankton pigment concentration along the Chilean coast. Journal of Geophysical Res. 104, 25877-25890. [19]