Seasonal and Interannual Variability of
the Particle Flux in the Humboldt
Current off Chile

Dierk Hebbeln

Geowissenschaften, Universität Bremen, Postfach
330440, 28344 Bremen, Germany,
E-Mail: dhebbeln@uni-bremen.de

Objectives

As being part of one of the largest Eastern Boundary Current systems, the waters off Chile belong to the most fertile regions in the world ocean. The upwelling of nutrient-rich subsurface waters sustains the high productivity in the surface waters and, finally, one of the worlds largest fisheries. However, the functioning of this system is still only poorly understood. Here we present the first long-term particle flux record from the waters off Chile, spanning a period of almost 10 years (although with some gaps).

Results

Between 1991 and 2001 up to three sediment traps have been deployed in different water depths at 30°S and 73°11´W off Coquimbo. The most complete time series record exists from ~2400 m depth. Throughout this time the flux pattern reveals some stable features. The total flux and the fluxes of the main constituents (carbonate, biogenic opal, organic and lithogenic matter) display very similar flux patterns, with a distinction between a high flux (August-October, >200 mg m^-2 d^-1) and a low flux period (April - June, <100 mg m^-2 d^-1). In general this pattern fits to the average pattern of surface water productivity and sea surface temperatures derived from satellite data, although the flux maximum is delayed versus the maximum pigment concentration by ~2 months.

Of special interest are the so-called stagnation periods, during which the traps collected no material at all, although they functioned perfectly. Their occurrence in two subsequent years (1994 and 1995) and in 2000 and the same pattern in the traps in 2400 m and 3800 m clearly points to the reliability of this observation. As these "events" always occur with the beginning of the spring bloom it is assumed that they are caused by the sinking of huge bio-aggregates (as e.g. diatom mats), which most likely clogged the traps. This hypothesis implies that the particle flux during the spring blooms in 1994 and 1995 must have been significantly higher than 600 mg m^-2 d^-1, an amount that has been sampled without any problems in 1993. Unfortunately, the particle flux rates during these events cannot be quantified more precisely.

From the five intensively sampled years only two were covered with >90% of the time (1993/1994 and 1997/1998). Comparing these two years (on an annual base from one winter mixing to the next, July to July, extrapolated to 365 days) reveals that the total flux in 1993/1994 (65 g m^-2 a^-1) was almost twice as high as in 1997/1998 (37 g m^-2 a^-1).

Carbonate is the most important constituent of the particle flux accounting for 48% to 57% on an annual base. These variations are counterbalanced mainly by changing biogenic opal contents, which vary between 7% and 17%. In contrast, the fluxes of organic carbon (4% to 7%) and lithogenic matter (26% to 29%) show relatively little variability. It is especially during the El Niño year 1997/1998 that very low biogenic opal and high carbonate contents occurred.

Discussion

The distinct seasonality of the particle flux in the Peru-Chile Current is in good agreement with the most important prevailing environmental parameters. The high flux period presents the annual spring bloom, which evolves when a shift in wind direction to dominating S-SW winds increases the strength of coastal upwelling resulting also in the lowest sea surface temperatures during the year. Pigment concentrations in the surface water reach their maximum together with the most westerly winds, almost two months before sea surface temperatures are lowest and the particle fluxes are highest. The two months offset between the maximum in surface ocean pigment concentration and the most prominent flux event is probably due to the development of a deep chlorophyll maximum in which a significant part of the spring bloom material is stored for some time. This assumption is based on the observation of very high sinking rates in the order of >300 m d^-1, thus, precluding a continuous sinking process after the vanishing of the surface water pigment concentration maximum.

The production system of the Peru-Chile Current is best described by the opal/carbonate ratio of the particle flux. The highest opal/carbonate ratio is linked to the spring bloom. There is a distinct peak in this ratio, probably reflecting the direct bloom sedimentation. This peak is significantly shorter than the whole high flux period, which probably includes a large part of the late bloom period, which is characterised by an enhanced content of the remains of calcareous organisms. Besides the dominant spring bloom peak the time series record of the opal/carbonate ratio reveals in most years a second, less distinct maximum in fall.

Based on this pattern the seasonal cycle of the productivity system in the Peru-Chile Current can be described as follows:

(1) Spring bloom (in the particle flux August to October), characterised by high total and opal fluxes by low sea surface temperatures.

(2) Late bloom (in the particle flux November to January), characterised by intermediate total, by high carbonate and low opal fluxes by rising sea surface temperatures.

(3) Fall bloom (in the particle flux February to April), characterised by intermediate to low total fluxes and a slightly increased opal/carbonate ratio by highest sea surface temperatures.

(4) Winter period (in the particle flux May to July), characterised by low total fluxes and a low opal/carbonate ratio associated with decreasing sea surface temperatures.

This pattern is quite robust through the whole sampling period (1991 - 2001) and prevails even through the 1997/1998 El Niño event.

In spite of this pattern there is also some interannual variability in the particle flux off Chile. While during the low flux period almost no interannual variability occurs, especially during the high flux period the flux rates can vary. Comparing the almost completely covered years 1993/1994 and 1997/1998 the total flux is as double as high in the earlier year. Assuming even higher fluxes during those years affected by the so-called stagnation periods 1994/1995 and 1995/1996 would even enhance the year to year differences in the particle flux rates.

In the SE-Pacific likely candidates to cause significant interannual variations are El Niño events. One of the strongest El Niño events of the last decades occurred in 1997/1998, when the particle flux off Chile differed quantitatively (the lowest annual flux observed, see above) and qualitatively from the other investigated years. The impact of an El Niño event is best reflected by the lithogenic matter to biogenic opal ratio, which increases significantly under El Niño conditions. Relative higher lithogenic input during the El Niño events of 1991/1993 and 1997/1998 most likely reflects enhanced continental precipitation, which in the semiarid hinterland of the study area probably causes a significant increase in the discharge.