Environmental Monitoring and Assessment

, Volume 184, Issue 11, pp 6935–6956

Large and mesoscale meteo-oceanographic patterns in local responses of biogeochemical concentrations

Authors

    • Civil Engineering Postgraduate, Program-COPPE/UFRJ, Center of TechnologyFederal University of Rio de Janeiro
  • Gilberto C. Pereira
    • Civil Engineering Postgraduate, Program-COPPE/UFRJ, Center of TechnologyFederal University of Rio de Janeiro
  • Jorge Luiz F. de Oliveira
    • Geography Postgraduate Program, Geoscience InstituteFluminense Federal University (UFF)
  • Nelson Francisco F. Ebecken
    • Civil Engineering Postgraduate, Program-COPPE/UFRJ, Center of TechnologyFederal University of Rio de Janeiro
Article

DOI: 10.1007/s10661-011-2470-3

Cite this article as:
de Oliveira, M.M.F., Pereira, G.C., de Oliveira, J.L.F. et al. Environ Monit Assess (2012) 184: 6935. doi:10.1007/s10661-011-2470-3

Abstract

Investigations surrounding the variability of productivity in upwelling regions are necessary for a better understanding the physical–biological coupling in these regions by monitoring systems of environmental impacts according to the needs of the regional coastal management. Using a spatial and temporal database from National Centers for Environmental Prediction (NCEP) and National Center for Atmospheric (NCAR) Research reanalysis, Quick Scatterometer vector wind, and surface stations from the Southeast coast of Brazil, we investigate the meteorological influences due to the large-scale systems in the variability of the nutrient and larvae concentration, and chlorophyll a, describing statistically relationships between them in upwelling regions. In addition, we used multivariate analysis, such as PCA and clustering to verify spatial and temporal variances and describe more clear the structure and composition of the ecosystem. Correlation matrix analyses were applied for different water masses present in the study area to identify the relations between physical and biogeochemical parameters in a region, where frequently upwelling occur. Statistical approaches and seasonal variability show that the period of November to March is more sensitive to nutrients (1.20 mg/m3 for chlorophyll a, 2.20 μmol/l for total nitrogen and 5.5 ml/l for DO) and larvae concentrations (120 org/m3 for most of the larvae, except for cirripedia that presented values around 370 org/m3) relating to the influence of large and mesoescale meteorological patterns. The spatial and temporal variables analyzed with multivariate approach show meaningful seasonality variance of the physical and biological samples, characterizing the principal components responsible for this variance in spring and summer (upwelling period), emphasizing the monitoring of species as crustaceans and mussels that are present in the local economy. Then, the spring and summer season are characterized by high productivity due to the occurrence of upwelling in this period.

Keywords

LarvaeNutrient concentrationsCoastal watersBrazil upwelling

Introduction

Western boundary currents into the ocean basin are regions of intense air–sea interaction, where the ocean loses heat and moisture to the atmosphere and absorbs carbon dioxide. Large-scale air–sea interactions in western boundary currents can affect weather and climate both locally and remotely, on time scales of days to decades. They are characterized by intense, narrow, and well-defined flows that flow off the continental margins (Stewart 2007). The South Atlantic subtropical ocean near the Brazilian coastline is a biogeochemical local, with physical and biological characteristics extremely relevant with its mechanisms of ocean–atmosphere (OA) interactions (Piontkovski et al. 2003). The anticyclonic circulation (gyre), which extends from the equator to approximately 40° S, has the upper-layer circulation bounded by four major currents, the South Equatorial, Brazil (BC—limited from the south by the Malvinas–Falkland current), South Atlantic, Benguela, and West wind drift currents known as the Circumpolar current (Piontkovski et al. 2003). The Brazil Current (BC) is the western boundary current of the South Atlantic subtropical gyre and has been described as a shallow and weak current when compared with the Gulf Stream in North Atlantic in terms of mass transport (Silveira et al. 2000). The BC is separated from the continental shelf at approximately 38° S, where it forms an intense front with the cold waters of the northwards-flowing Malvinas Current (MC) called Brazil Malvinas Confluence (BMC) region in the extra-tropical Southwestern Atlantic Ocean. The BMC shows strong surface gradients of temperature, pressure, and salinity. This region is known as one of the most energetic zones of the World Ocean with two distinct nature of thermal, sea height, and chlorophyll a gradients between the BC, a warm, high, and oligotrophic current and the MC, a cold, low, and eutrophic one (Pezzi et al. 2005).

The high temperature and -salinity BC runs southwards carrying the Tropical Waters (TW) from Equator to approximately 38°S, where comes across the Subantartic Waters. Moving in the bottom on the opposite direction there is the cold and nutrient-rich South Atlantic Central Waters (SACW). Therefore, the oligotrophic TW is the prevailing water mass in the euphotic zone in this region of the South Atlantic Ocean (Andrade et al. 2004).

The upper ocean plays a fundamental role in building a structure of both wind-driven and thermohaline circulation. Many aspects of upper ocean dynamics are still unknown, especially how its variability is characterized by the interaction of different types of motion and scales (Griffa et al. 2008). Large-scale meteo-oceanography patterns are in accordance with the OA interactions through the South Atlantic subtropical high-pressure system, a predominant air mass above the central region of the South Atlantic Ocean basin, centered near 30° latitude that induces the currents of upper ocean due to the wind-driven forces (Stewart 2007). Divergence of the boundary layer is associated with wind forcing surface circulation and it is a well-known cause of coastal upwelling. Coastal upwelling is regionally more limited than open ocean upwelling but its stronger vertical motion is associated with a greater climatic and biological impact and could be defined as the vertical movement of water masses compensating for the offshore Ekman drift (Myrberg, et al. 2010).

Spatial variation in wind stress is another mechanism for the production of surface divergence. This divergence is reduced in surface transport that occurs between the mid-shelf and the coast. The wind stress curl strength is greatly increased by the atmospheric marine boundary layer, interacting with the coastal topography, resulting in hydraulic features such as expansion and compression bulges (Dever et al. 2006).

Investigations surrounding the variability of productivity in upwelling regions have raised the need for a better understanding of the mean distributions and variability of nutrients, as well as the physical–biological coupling, in these regions. Upwelling systems are characterized by the ascension of a cold and rich-in-nutrients water that disturb ecosystem dynamics and increases the environmental heterogeneity (Myrberg, et al. 2010).

This study is part of the research developed by the Federal University of Rio de Janeiro—Civil Engineering Program in remote monitoring systems of environmental impacts in coastal regions in order to develop trophic dynamic models to be used in the National Plan and Regional Coastal Management or in any other aquatic system. To aid in the continuing study of biogeochemical distribution and its variability in coastal upwelling zones we have been verified the relations between them and meteo-oceanographic patterns related to the seasonal climatology for the Arraial do Cabo region, northeast of Rio de Janeiro State. This place is known for its active wind induced upwelling (Campos et al. 1995) and is one of the most attractive sea and landscape for tourist and recreational activities, contributing to the local economy, but with a disorder urban increasing. Moore et al. (1997) asserts that coastal area is an environment, where there are conflicting interests as developmental, industrial and conservational, and management aims to reconcile these different viewpoints. Negative effects of anthropogenic contamination in the coastal zones are related not only to chemical pollutants, but also marine debris that enter the marine environment from any source (Santos et al. 2009). Marine conservation is increasingly on economics and the social sciences, and conservation practitioners are towards achieving ecosystem protection while maintaining economically viable activities that depend on marine resources. Nowadays, the acknowledgment that marine conservation is largely due to managing multiple human uses of the coastal zones is increasing (Salomon et al. 2011)

The seasonal variability of the South Atlantic high pressure is associated with the occurrence of upwelling in Arraial do Cabo. This condition is set up in the summer by large-scale, high-speed winds northeasterly blowing over the region off the coast (Pezzi 2006). This point divides Brazilian coast in environments with tropical and subtropical features in a small spatial scale (Guimaraens and Coutinho 1996).

The aim of this paper is to investigate the influences of large-scale meteorological patterns in the biogeochemical responses at Arraial do Cabo mesoescale system based on synoptic data from NCEP/NCAR reanalysis and the QuickScatterometer vector wind product over the South Atlantic Ocean near the Southeast Brazilian coastline, applying statistical tools.

Material and methods

Study area

We have widened the field of the study area in order to investigate the spatial and temporal meteorological influences due to the large-scale systems in the variability of the nutrient and larvae concentration, and chlorophyll a. According this criterion, we made a selection of remote grid points from NCP/NCAR reanalysis data in the oceanic area under the influence of high-pressure system. To verify local influences, we used a data set from surface meteorological stations and satellite winds. Thus, the study area is comprised between 21°S to 25°S and 39°W to 43°W, situated on the southeast coast of Brazil (Fig. 1).
https://static-content.springer.com/image/art%3A10.1007%2Fs10661-011-2470-3/MediaObjects/10661_2011_2470_Fig1_HTML.gif
Fig. 1

Study area with the points representing Sâo Pedro d’Aldeia (22°57 S/42°06′ W) and Arraial do Cabo (22°57′ S/42°14 W) meteorological stations, situated in Rio de Janeiro state; NCP/NCAR reanalysis data of mean sea level atmospheric pressure in grid point A (22°30′ S/40°00′ W), and B (25°00′ S/42°30′ W); zonal and meridional wind components in grid point 1 (21°53′ S/41°15′ W), 2 (21°53′ S/39°22′ W), and 3 (23°48′ S/41°15′ W); satellite QuickSCAT winds in 22°52′ S/41°52′ W and water harvest point in 23°00′ S/42°00′ W

Southwest Atlantic is characterized by the presence of the BC, a warm western boundary current that goes away inshore from the northern Brazil over a continental shelf toward the south, carrying the TW from the vicinity of the Equator (Gaeta et al. 1999). Due to the presence of this current, this region is known by its oligotrophy; therefore, in some areas, a seasonal wind-driven upwelling of cold, nutrient-rich SACW mass occurs, benefiting on biological productivity. In particular, in Arraial do Cabo the positioning of the Cabo Frio island (23°S, 42°W) forms the small (45 km2) and narrow (∼10 m deep) Anjos Bay (Fig. 1). This region can be considered yet a pristine area and the hydrologic conditions are strongly influenced by the wind pattern that influences the distribution of water masses. Winds that blow from northeastern and the Earth’s rotation result in a shunting of the nutrient-depleted surface TW of BC to offshore followed by the up-flow of the deeper (∼300 m) and nutrient-rich of SACW mass (Pereira et al. 2008). Upwelling events and inorganic nutrients are then supplied to euphotic zone by the exchange of water between nutrient-depleted surface water and nutrient-rich deeper water. The inverse wind pattern occurs due to the passage of cold fronts with winds blowing from the south and southwestern, bringing the oligotrophic TW back to the coast. These processes have a direct impact on the quantity and composition of the phytoplankton communities, modifying the trophic structure (Valentin 1989). On the other hand, elevated nutrient concentrations are the main origin of coastal eutrophication processes and their monitoring allows direct estimates of the degree of contamination, and obviously making possible their management (Topcu et al. 2010; Morgan and Kline 2010).

Large-scale climatology

Characterization of large-scale interactions between the ocean and lower atmosphere is essential for understanding the mechanisms of natural climate variability on different time scales. The atmospheric circulation over the oceanic region of the study area refers to the climatological pattern of the general circulation of the South Atlantic. This is the subtropical anticyclone with surface winds blowing counter-clockwise around this system and then we find between 40° and 60°S westerly winds and the prevailing easterlies to the north. The South Atlantic subtropical high-pressure is semi-stationary and presents a well-emphasized seasonal movement. During the summer (Fig. 2a), the subtropical high, over the continent, becomes weaker than winter (Fig. 2b), moving southerly and on its western side the winds blow northeasterly towards the southeastern coast of South America and they are more intense in the southeastern coast of Brazil. Therefore, in the north region of the BMC, at 38°–40°S, there is an intensification of the northerly winds and a weakening of the southwesterly winds which is difficult for the track of the cold fronts that reach the south and southeast coast of Brazil (Ahrens 2000). This seasonal variability is one of the most important factors related to the occurrence of upwelling in the region of Arraial do Cabo, where the winds blow along the coastline from north to south push the surface waters offshore on summer periods (Pezzi 2006). Matano et al. (2001) suggested that the transport of BC follows the curve of annual variation of wind shear over the subtropical basin with a maximum during the summer and minimum during the winter.
https://static-content.springer.com/image/art%3A10.1007%2Fs10661-011-2470-3/MediaObjects/10661_2011_2470_Fig2_HTML.gif
Fig. 2

Surface wind-flow pattern with a scheme of the South Atlantic high-pressure system for summer (a) and for winter (b). The circles represent the isobars and the positions of the high centers moving slightly during the year; the squared shows the winds blowing over the study area; BMC Brazil Malvinas Confluence region (source: NCEP/NCAR reanalysis data set)

Water masses in the southeastern Brazilian shelf

We present in this paper a brief description of the main water masses in the southeastern Brazilian shelf. Continental shelves are zones adjacent to a continent (or around an island) that extending from the surface water line to the depth, usually about 120 m, where there is a steep descent toward great depths. Traditionally, the classification of water masses is made through temperature/salinity diagrams according to their termohaline circulation. According to Stewart (2007), there are two factors that can identify the water masses: the geographic region where they arise, and the depth in which they reach the vertical equilibrium.

The TW is a warm and salty South Atlantic surface water mass, which at the western boundary is transported southward by the BC. This surface water is formed as a result of the intense radiation and excess of evaporation in respect to precipitation. On Brazilian Southeastern coast, the TW is characterized by temperatures above 20°C and salinities above 36% (Silveira et al. 2000). The Sub-Antarctic Water is cold and less saline high-latitude water mass and its western boundary layer reaches northward extensions due to advection by the Malvinas Current. From the mixing between these two water masses, the SACW is formed and takes place at the Subtropical Convergence Zone (confluence between the SACW and the Antarctic Circumpolar Current) that extend as far north as 35°S. This water mass is associated with sinking and northward transport. The SACW is found flowing into the region of pycnocline, with temperatures above 6°C and below 20°C and salinities between 34.6 and 36 PSS (Practical Salinity Scale). An index of the SACW thermohaline circulation is around 20°C and 36.2 PSS in the Brazilian Southeast (Silveira et al. 2000). The Coastal Water (CW) has the thermohaline characteristics varying according to the annual cycle of river runoff and mixture with offshore waters (Soares and Möller Jr. 2001).

The Environment National Council (CONAMA) Resolution 357/2005 that provides the classification of water bodies and environmental guidelines for its framework determines limits for the levels of several chemicals components, including nitrogenous nutrients (Table 1).
Table 1

Resolution CONAMA 237/2005 limits for oxygen levels and nitrogenous nutrients, according to the classification of water bodies

  

Salinity (g L−1)

Oxygen (mg L−1)

Nitrate (μmol L−1)

Nitrite (μmol L−1)

Ammonium (μmol L−1)

CONAMA 357

Class 1

≥30

≥6.0

28.57

5.00

28.57

Saline waters

Class 2

≥30

≥5.0

50.00

14.28

50.00

In the case of Arraial do Cabo, we have used temperature and salinity data provided by the Admiral Paulo Moreira Institute of Sea Studies—IEAPM and the water mass thermohaline indices are presented in Table 2.
Table 2

Southern Brazilian shelf water mass thermohaline indices in Arraial do Cabo

Water mass

Temperature (°C)

Salinity (g/L)

SACW

T < 18

S < 36

SACW/Coastal

18 < T < 20

35 < S < 36

Coastal

T > 20

S < 35.4

SACW/Tropical

18 < T < 20

S > 36

Coastal/Tropical

T > 20

35.4 < S < 36

Tropical

T > 20

S > 36

Source: Brazilian Navy Oceanography Department

Data

In situ measurements

Surface seawater medium-term of physical, chemical, and biological samples (0.5 m deep) were collected with a Nansen bottle with reverse thermometer outside, and in the bottom (water/sediment interface), by scuba diving using a 2-l polyethylene bottle (three samples). The salinity, DO, and nutrients were determined ashore as described in SCOR (1996)). To chlorophyll a was applied the method described in (Richard and Thompson 1952). An inversion thermometer fixed to the outside of a Nansen bottle was employed for temperature. The physicochemical parameters are then: Sea Surface Temperature (SST), salinity, dissolved oxygen (DO), nitrogen as ammonium cation (NH4+), nitrite (NO2) and nitrate (NO3), and ortho-phosphate (PO4). The biological variables are composed of chlorophyll a (milligrams per cubic meter) measurements as an estimation of microalgal biomass, but probably also contains all free-living autotrophic bacteria of the water column both influenced qualitatively and quantitatively by nutrient entrances that on the other hand, supplies itself as feeding material for meroplankton larvae which are expressed in numbers of organisms per cubic meter of water and were collected by means drag plankton net of 100 mesh and fixed in 10% formalin and then counted under a microscope (Table 3). These data were collected with weekly frequency from July 21, 1999 to June 28, 2007 in Anjos Bay, Arraial do Cabo city, Rio de Janeiro State and the nutrients are in accordance with the CONAMA resolution (Table 1).
Table 3

Statistical summary of the water measurements

Variables

Max

Min

Mean

S. Dev.

Biological

 Chlorophyll a (mg/m3)

11.9

0.0

1.0

1.19

 Ascidiacea (org/m3)

1115

0.0

11

59.5

 Bivalvia (org/m3)

1833

0.0

99

194.9

 Briozoa (org/m3)

101

0.0

2

5.7

 Cirripedia (org/m3)

3641

0.0

210

362.8

 Cypris (org/m3)

5192

0.0

22

255.5

 Decapoda (org/m3)

437

0.0

20

35.9

 Isognomon (org/m3)

2342

0.0

31

166.5

 Mytilidae (org/m3)

2636

0.0

93

173.9

 Polychaeta (org/m3)

1683

0.0

20

91.5

 Ostreidae (org/m3)

1132

0.0

27

76.9

Sample water

 Temperature (°C)

26.7

15.9

22.6

1.76

 Salinity(g/L)

36.6

33.5

35.7

0.46

 Oxygen (DO) (ml/L)

7.0

2.6

5.3

0.45

 Phosphate (PO4) (μmol/l)

3.7

0.0

0.3

0.21

 Nitrite (NO2) (μmol/l)

0.6

0.0

0.1

0.08

 Nitrate (NO3) (μmol/l)

10.2

0.0

0.7

0.97

 Ammonium (NH4) (μmol/l)

7.8

0.1

1.2

0.79

Through physical and chemical variables, it is possible to verify the variability of hydrological characteristics as a function of interchangeable periods of upwelling and downwelling events, which define the quality patterns of different water masses (Pereira et al. 2008).

Meteo-oceanographic data set

In order to analyze the remote and local influence of meteorological factors in the variability of these biogeochemical variables we used in this work an environment data set obtained from three sources from July 1999 to June 2007.
  1. (a)

    Station measurements and reanalysis data

     
Hourly tide gauge records and forecasting from the Arraial do Cabo station near Anjos Bay, 6-hourly (UTC) atmospheric pressure from São Pedro d’Aldeia (SPA) meteorological station, 6-hourly (UTC) atmospheric pressure (P_A and P_B), zonal (u) and meridional (v) wind components, at 10 m above the ground, from NCEP/NCAR reanalysis (Table 4). The reanalysis data were obtained from five grid points, four over the ocean region bounded at 21°S, 25°S, and 39°W towards the southeastern Brazilian coast. The wind data are considered to be from the most accurate class of data in the reanalysis data set (Kalnay et al. 1996). They are considered variables strongly influenced by the available observations, rendering them more reliable (Kistler et al. 2001).
Table 4

Statistical summary of the meteo-oceanographic data set

Variables

Max

Min

Mean

S. Dev.

Stations

 Pressure_SPA (hPA)

1028.0

1003.0

1016.0

4.65

 Wind stress_SPA (N/m2)

0.3256

0.0

0.0416

0.0512

 Tide (cm)

327

178

256.7

27.99

 Meteorological residual (cm)

51.5

−48.5

2.15

15.11

Reanalysis

 Pressure_A (hPA)

1027.8

1004.5

1017.1

4.21

 Pressure_B (hPA)

1029.3

1003.4

1016.7

4.71

 Wind stress_1 (N/m2)

0.1901

0.0

0.0458

0.0347

 Wind stress_2 (N/m2)

0.3297

0.0

0.0790

0.0568

 Wind stress_3 (N/m2)

0.3502

0.0

0.0716

0.0572

Satellite (QuickSCAT)

 Wind stress_Quick (N/m2)

0.4442

0.0

0.0963

0.0811

  1. (b)

    Satellite wind measurements

     

We used the Quick Scatterometer (QuikSCAT) vector wind product of Remote Sensing Systems (RSS) for the same period, available daily on a 0.25° grid. It was obtained from the National Aeronautics and Space Administration Jet Propulsion Laboratory Physical Oceanography Distributed Active Archive Center. We used Level 3 (L3) 25 km grid products from two passages per day (08 and 20 UTC) near the point of interest (Table 4). The QuikSCAT is the first satellite-borne scanning radar scatterometer which measures the surface roughness of the ocean, affected by the wind magnitude and direction, by transmitting microwave pulses (13.4 GHz) and receiving the backscatter. Multiple and simultaneous normalized radar cross-section values are obtained from the backscatter power at a single geographical location or wind vector cell and converted to wind speed and direction measurements (10 m neutral winds) using a Geophysical Model Function (Sharma and D’Sa 2008). High-resolution QuikSCAT vector wind fields suitable for coastal applications and studying of smaller oceanic processes have been produced by combining scatterometer measurements with a regional mesoscale model (Chao et al. 2003) or by use of “slices” (Tang et al. 2004).

Methodology

Direction and wind speed as well as the meteorological residual were calculated from the wind data set and the tide gauge records, respectively.

Surface wind stress provides the most important forcing of the ocean circulation, while the fluxes of heat, moisture, and momentum across the air–sea boundary are important factors in the formation, movement, and modification of water masses and the intensification of storms near coasts and over the open oceans (Atlas et al. 1987). Therefore, in this work, we have calculated the zonal (ZWS) and meridional wind stress (MWS) for each grid point using the following equations:
$$ {T_x} = \rho {C_d}\left| W \right|u $$
(1)
$$ {T_y} = \rho {C_d}\left| W \right|v $$
(2)
where:

\( \rho = 1.22\;{\text{kg}}\;{{\text{m}}^{{ - 3}}} \)(air density);

W = intensity of the wind (meters per second) calculated from zonal (u) and meridional (v) wind components Cd = 1.1 + 0.053 W (coefficient of drag for the southeast Brazil coast, (Stech and Lorenzzetti 1992).

The units used for wind stress are newtons per square meter, where 1 hPa is equal to 102 N m−2.

The meteorological time series are then measured weekly and for the same period as all the others.

Statistical analysis

Firstly, the basic statistical analysis was applied in the meteo-oceanographic time series to verify the maxima, minima, and mean values of the samples. Table 2 presents the values of this analysis for each variable with respective standard deviations.

Some outliers were identified in the SST and salinity data from the percentiles analyses. Then they were substituted with the average values between the previous and the following weekly data. Seasonality analysis and wind distributions from number of occurrences were carried out to verify the wind stress patterns in the periods of upwelling in the Arraial do Cabo coastline.

The seasonal patterns were investigated between meteorological forcing and sample waters to verify the potential characteristics related to the nutrient concentrations and chlorophyll a that occur in a context of seasonal and long-term meteorological changes (Anneville et al. 2004). Meteorology also plays a major role in the dynamics of aquatic systems. Environmental changes affect species in many different ways, altering their productivity and interactions with other species and understanding these effects of ecosystems is a challenge for research management (Breitburg et al. 1999; Harrington et al. 1999). Intentional changes for ecosystem management represent an investment to improve ecosystem services, such as fishing or aquaculture production, but can also pose risk interactions between native and non-native species and it may have unintentional consequences on the environment (Fulford and Breitburg 2011).

Therefore, anomalies and seasonal cycles in the samples were investigated in order to verify the variability of these parameters during the period. Cross-correlations between meteorological data, chlorophyll a and nutrients were also calculated to verify the respective time lag in the response of the meteorological variability that can cause variations in the samples waters data. Spectral and cross-spectral analyses were carried out, using fast Fourier transform to identify the stronger periodicity in the series at the respective frequency for each variable as well as the influence on the periodicity between them throughout the cyclical behavior. The purpose of cross-spectrum analysis is to uncover the correlations between two series at different frequencies. Thus, it is important to analyze the coherence between the peaks of the time series to verify the linear correlation between the components of the bivaried process. For research management is very important to understand the temporal coherence to identify the degree to which different locations behave similarly or dissimilarly through time.

The knowledge of trophic relationships in the different water masses seems to be crucial for the success of environmental management policies in coastal aquatic systems. Therefore, from the total data set were extracted those relating to each water masses, according to temperature and salinity. Thus, it was possible to define a typical data set concerning water masses. A correlation matrix was then applied to the meteo-oceanographic and biological variables of each water masses in order to identify the relationships between these variables in relation to ocean–atmosphere interaction for each marine environment of interest.

Multivariate analysis

The relationship between structural dynamics across scales requires a multivariate method that treats several dimensions simultaneously. Thus, this method can take into account the variability between the different locations (spatial scale) and the variability between successive samples (temporal scale). Principal component analysis (PCA) is used in all forms of analysis because it is a simple, non-parametric method of extracting relevant information from data sets. This method provides a very clear road map of how to achieve our goals in reducing a complex data set to a lower dimension and reveal a simplified structure that often belong it (Shlens 2005). Numerical analysis procedures, such as PCA, have been developed to interpret large space/time data sets, which can decompose total variance into spatial and temporal variances. In order to access the data structure, a matrix of 33 variables and 414 objects were used in two statistical approaches. The principal axis method was used to extract the components, and this was followed by a varimax (orthogonal) rotation. The data set was separated seasonally in spring–summer (upwelling period) and autumn–winter (downwelling period) to characterize which variables are more important in these distinct periods. Firstly, we applied the PCA, based on a correlation matrix, and so established a set of orthogonal factors providing information about similarities of the samples. As a second statistical approach, clustering analysis was performed to describe more clearly the structure and composition of ecosystem functional units for the two periods. For clustering variables, the Ward’s method and Pearson r coefficient were applied.

Results and discussion

Seasonal variability of the wind stress induced upwelling

Wind events induce different disturbances of the water mass structure depending on the season. In Arraial do Cabo coastline, when NE-wind events increase the intensity for a period, mixing with the water column, the response is the SACW upwelling with capacity of a redistribution of nutrients with effects on plankton dynamics.

The average wind stress from remote grid points present a well-defined seasonality between spring–summer and autumn–winter (Fig. 3a and b) with great values for the spring–summer months. For the points near the coastline, there are also higher values in the spring–summer months (Fig. 3b); the month of September being the most significant for all analyzed points, except for ws_2 point, presents the highest values. This is in accordance with the South Atlantic high-pressure system that is predominant in the region and has a well-emphasized seasonal variability with light zonal and meridional movement (Fig. 2).
https://static-content.springer.com/image/art%3A10.1007%2Fs10661-011-2470-3/MediaObjects/10661_2011_2470_Fig3_HTML.gif
Fig. 3

a Seasonality of the mean wind stress values for the period of 1999 to 2007 in the three-reanalysis grid points, b in the station and in the oceanic QuickSCAT satellite point

The northern wind direction is predominant during the occurrence of upwelling in the study area, and thus we verified the seasonality of monthly distribution of northeasterly wind stress on all respective points. Figure 4a and b show the patterns of these distributions for the remote grid points, and for the points near the region of interest (station and satellite), respectively. The remote grid points also show higher percentages for the period of spring–summer as verified to the patterns of wind stress values (Fig. 3a and b). Therefore, the higher values of wind stress are related to the higher occurrence of northeasterly wind stress for spring–summer period.
https://static-content.springer.com/image/art%3A10.1007%2Fs10661-011-2470-3/MediaObjects/10661_2011_2470_Fig4_HTML.gif
Fig. 4

a Seasonality of the occurrence of northeasterly wind stress for the period of 1999 to 2007 in the three-reanalysis grid points, b in the station and in the oceanic QuickSCAT satellite point

Seasonal variability of the temperature and nutrients

Nutrient concentrations ranged from 2.6–7.0 ml/l for DO, 0.0–3.7 μmol/l for phosphate, 0.0–0.6 μmol/l for nitrite, 0.0–10.2 μmol/l for nitrate and 0.1–7.8 μmol/l for ammonium; chlorophyll a concentrations ranged from 0.0–11.9 mg/m3. Considering the total nitrogen concentration the range is from 0.26–11.97 μmol/l with mean value around 2.04 μmol/l for the period (1999–2007). Spring and early summer months (SON–DJF) present values above the average; September presents the highest seasonal variability (2.46 μmol/l) and March the lowest (1.5 μmol/l). Oxygen concentrations present values below the average (5.29 mg/l) only in autumn with the lowest value in April (4.94 mg/l) and the highest in November (5.50 mg/l). For phosphate concentrations, winter months (July/August) and spring (September/October) present values above the average (0.27 μmol/l) with the highest values in September (0.42) and the lowest in March (0.18). The abundance of chlorophyll a shows a meaningful seasonal variability for the analyzed period. We verified three peaks in January, May, and August with the highest value in May (1.28 mg/m3) and the lowest in September and March (0.75 and 0.76 mg/m3, respectively). Average concentrations and standard deviations are presented in Table 1 for collected samples.

Seasonal variations were compared between SST and the nutrient concentration (DO, NO2, NO3, NH4, and PO4), and chlorophyll a in the Anjos Bay waters. The results are shown in Fig. 5a and e, where it is possible to verify that in autumn occurs an increase of the SST (maxima in April) and a decrease of the nutrients (minima in March). These patterns can be related to the oligotrophic TW mass. However, the opposite is verified in spring and summer periods (except for the PO4) which can be associated with the presence of a seasonal wind-driven upwelling of cold, nutrient-rich SACW mass. The minimum value of the SST and maxima of total nitrogen, and PO4 was verified in September. Verifying the seasonal variability of the wind stress (item 3.1), occurs in September with a strong wind pattern in all considered points with a frequency of 60% for northeasterly winds, approximately. The minimum of wind occurs in April, when the SST values are higher. According to Franchito et al. (2008), the SST values are quite low from September to February and they start to increase gradually, reaching the maximum values in April. These authors show that the SST seasonal variation is strongly dependent on the occurrence of upwelling in Arraial do Cabo region, with high SST values occurring during the austral autumn and winter and low SST values in the austral spring and summer months. They also verified that the wind speed is higher in the months when the upwelling is present with two maxima: in February and September and minima from April to July. Whereas the chlorophyll a minima values occur in September when the wind stress values are higher. Neves et al. (2008) comment that there are many studies that affirm the existence of a negative correlation between these two variables: when the wind is strong, primary production decreases. This fact can be explained due to the whitecapping breaking in the shallow water bodies lead to the production of short period waves under strong wind conditions that can prevent the transmission of light for phytoplankton (Jones and Monismith 2008).
https://static-content.springer.com/image/art%3A10.1007%2Fs10661-011-2470-3/MediaObjects/10661_2011_2470_Fig5_HTML.gif
Fig. 5

ad Seasonal variation between SST and nutrients and chlorophyll a for the period of 1999 to 2007

The seasonal variations relating to the abundance of larvae in Anjos Bay were analyzed to verify the intraseasonal variability of the physical–biological coupling. The vast majority of species have complex life cycles that include planktonic larval stages and the upwelling can additionally play a key role in this region, replenishing the euphotic zone with the nutritional components necessary for biological productivity. Then, the intraseasonal variability indicates that November, December, February, and March are of critical importance to the region for the analyzed period. Figure 6 shows that in the period of occurrence of upwelling (spring–summer) are verified the major peaks of abundance of larvae collected in the Bay with the highest value (387 org/m3) for Cirripedia on February, except for Polychaeta that presents a peak (68 org/m3) in July. These results are in accordance with the studies conducted by Andrade et al. (2004) in the southwest Atlantic Ocean off the Brazilian coast from October to December 1998. These studies showed that on the south of latitude 19° S, the input of nutrients are related to deep waters, where the remobilization of nutrients leads to a higher biological activity for the spring/summer periods.
https://static-content.springer.com/image/art%3A10.1007%2Fs10661-011-2470-3/MediaObjects/10661_2011_2470_Fig6_HTML.gif
Fig. 6

Peaks of monthly average values of abundance of larvae collected in the Anjos Bay for the period of 1999–2007

Cross-correlation and cross-spectrum of the physical and nutrient parameters

The convention we use for this correlation is that we arbitrarily consider the physical parameters to be the forcing function. Figure 7 shows the results of the best cross-correlation between the physical and nutrient parameters at 95% significance level. It shows the maximum absolute value of the cross-correlation at each analysis grid point, giving a direct measurement of the degree of local similarity of the features in the two data sets. It was verified that the wind stress remote grid points presented relatively higher cross-correlation with DO and PO4 nutrient parameters (Fig. 7a). The cross-correlation of ZWS on grid_2 and DO shows most notably correlation at around one week lag, as in others, having a negative peak of 0.246 and ZWS on grid_1 and PO4, a negative peak of 0.163. As shown formerly, there is a predominance of E-NE wind directions in the respective grid points. The more significant correlations related to zonal wind stress with negative peaks of correlation show the influence of warm waters of the Brazil Current moving towards the coast (onshore winds), inducing a decrease in the amount of nutrients in the region. It can be observed in Fig. 7b that the correlations of SST with nutrients and chlorophyll a also present negative peaks, being in agreement with the results obtained for the cross-correlations shown in Fig. 7a. The correlation between wind stress and the abundance of chlorophyll a is generally negative (Neves et al. 2008) and we verified a low negative correlation between chlorophyll a and the wind stress values only in grids 2 and 3 in September when the wind speed is stronger. The cross-correlation of SST and DO shows most correlation at around zero lag, as in others, having a negative peak of 0.386 and with NH4 a negative peak of 0.323. The meteorological residual presented more meaningful cross-correlations with SST and NO2 at around zero lag and one week lag, having positive peaks of 0.275 and 0.117, respectively (Fig. 7c). The positive correlation between residual and SST can be related to warming of sea water and therefore the water expansion. With respect to the positive correlation between residual and NH4, it can be associated with the influx of warmer coastal waters and an increase of this nutrient.
https://static-content.springer.com/image/art%3A10.1007%2Fs10661-011-2470-3/MediaObjects/10661_2011_2470_Fig7_HTML.gif
Fig. 7

Cross-correlation between physical parameters and nutrient, chlorophyll a and DO that presents more meaningful results. a Zonal wind stress on grid_1 × DO (−0.236), grid_1 × PO4 (−0,163), grid_2 × DO (−0.246), grid_2 × PO4 (−0.155), grid_3 × DO (−0.183), grid_3 × PO4 (−0.138), Quick × DO (−0.169), meridional wind stress on grid_3 × PO4 (−0.132); b SST × DO (−0.386), SST × PO4 (−0.249), SST × NO3 (−0.285), SST × NO4 (−0.323), SST × Chlorophyll a (−0.1835); c Residual × SST (0.275), Residual × NO2 (0.117–1 week lag)

The seasonal correlation shows that nitrogen concentration and phosphate present the major correlation (Fig. 8) with the wind stress obtained from grids 1, 3, and satellite (QuickSCAT) for the spring and summer months (upwelling). We verify a high positive correlation (R > 0.50) between ws (point 1) and PO4 in January (a) and a lower correlation between ws and nitrogen concentration in November (b). The correlations present low positive values (R < 0.50) between ws and nitrogen concentration for the points Q and 3 in January (c and d). Although during the analyzed period the correlation coefficients between these variables presented not very meaningful values, they are quite important for monitoring the region due to the lack of the information about these interactions.
https://static-content.springer.com/image/art%3A10.1007%2Fs10661-011-2470-3/MediaObjects/10661_2011_2470_Fig8_HTML.gif
Fig. 8

a to d: Scatter graphics between wind stress (point 1) and phosphate (a) and nitrogen concentration (b) with coefficients of correlation R = 0.56 and R = 0.44, respectively; scatter graphics between (c) wind stress (point Q) and (d) wind stress (point 3) between nitrogen concentration with correlation coefficients R = 0.37 and R = 0.30, respectively

A cross-spectrum analysis was applied for nutrients (dependent series) and physical parameters (independent series). A graphic interpretation of bilateral periodicity of both series was verified by plotting co-spectral density, cross-amplitude, and squared coherency graphs. For these, we calculated power spectral densities using bivariate Fourier analysis (cross-spectrum). The signals were divided into data windows for density estimates with a Hamming window. Coherence functions were then derived as the absolute components of the cross-spectral density squared divided by the product of the power spectral densities. Significance levels were derived using equivalent degrees of freedom that depended on the length of the series used and the width of the window. Figure 9 shows the best amplitude and coherence of these analyses between nutrients and SST, zonal wind stress on grid point 2, and pressure on grid point A, respectively. Significant correlation in the frequency intervals 0.014–0.024 (with maximum of 0.019 point-period of 51 weeks) was verified by analyzing cross-amplitude and squared coherency graphs for both independent time series. Significant correlation in the frequency intervals 0.0048–0.0121 (with maximum of 0.0072 point-period of 137 weeks) was also verified by analyzing cross-amplitude and squared coherency graphs for SST × DO (Fig. 9a) and Pres × DO (Fig. 9c). The probable periodicity of 51 weeks (approximately 1 year) that we ascertained in both series can be associated with the upwelling in this region.
https://static-content.springer.com/image/art%3A10.1007%2Fs10661-011-2470-3/MediaObjects/10661_2011_2470_Fig9_HTML.gif
Fig. 9

Cross-amplitude and squared coherence estimates of a SST, b ZWS, and c pressure between DO and PO4

Table 5 shows the results obtained in spectrum analyzing for both nutrients and physical parameters, where is possible to verify that the best results were obtained for DO and PO4 nutrients.
Table 5

Cross-spectrum and coherence between the nutrients and SST, zonal wind stress on grid point 2 and pressure on grid point A

Nutrients

SST

ZWS_2

Pres_A

Amp

Coherence

Amp

Coherence

Amp

Coherence

DO

23.92353

0.963312

0.288845

0.916220

68.72531

0.941535

PO4

9.415338

0.928837

0.115615

0.913798

27.01704

0.905796

NO2

2.818039

0.754988

0.032253

0.645277

7.910187

0.704542

NO3

17.23684

0.456085

0.203372

0.414253

45.64328

0.378766

NH4

23.80082

0.548535

0.297154

0.557877

69.15086

0.548407

Principal anomalies

Local anomalies in biogeochemical variables can help to improve weaknesses, increasing the validity of coastal ecosystem indicators. The principal anomalies found in this research are related to the ammoniacal nitrogen and chlorophyll a. The dissolved ammoniacal nitrogen is present in two forms: ammonia (NH3) and ammonium cation (NH4+) whose proportions depend on the pH, SST, and salinity. The ammonium arises from the decomposition of protein, chlorophyll a, and other organic nitrogen arrangements. It is predominant at sea and is directly related to the decomposition of organic matter and may be associated with human activities. The variation of chlorophyll a is important for understanding the development of primary production (based on the photosynthetic activity) and hence, understanding the development of the biological community that consist of the trophic dynamics. Moreover, the phytoplankton has a relatively short time response to environmental changes due to their high proliferation. Furthermore, variations in currents and physicochemical properties lead a consequence in the distribution of this primary production in the oceans.

The distributions of nutrients for the period demonstrate an evidence of chlorophyll a and ammonium (NH4+) positive anomaly (Fig. 10). We observed an increase of chlorophyll a from November 2004 and for ammonium between April and October 2006; from January 2007 is verified a decrease in the ammonium concentration. Although the low nutrient anomaly carried by the northeasterly flow of Tropical Water mass due to the presence of the Brazil current, characterizing this region as an oligotrophic one, we did not observe intense nutrient negative anomalies for this period.
https://static-content.springer.com/image/art%3A10.1007%2Fs10661-011-2470-3/MediaObjects/10661_2011_2470_Fig10_HTML.gif
Fig. 10

Distribution of chlorophyll a (in milligrams per cubic meter) and NH4 (micromoles per liter) anomalies for the 1999–2007 period

Correlation matrix analysis for different water masses present in the study area

  1. (a)

    South Atlantic Central Waters

     
The MWS presents positive correlation with PO4, NO2, and NO3 in remote points (grids 1, 2, and 3 as well as P_A and P_B), suggesting a relation to northerly winds. The larvae of crustaceans such as Decapoda (prawns, lobsters and crabs), Cypris (type of small crustacean related to shrimps and mussels), and Cirripedia (barnacles) show a positive correlation with all points of atmospheric pressure and meridional wind obtained with satellite, respectively. This result suggests a greater sensitivity of these larvae and nutrients associated with the position of the South Atlantic high pressure, blowing strong northeasterly winds over the region (Fig. 2), and leading to the resurgence of SACW. For salinity and the ZWS, the correlation is positive and more significant for remote grid points 1 and 3. The meteorological residual presents negative correlation with the salinity and positive with chlorophyll a, suggesting the occurrence of upwelling and the variability of the sea level (Table 6).
Table 6

Correlation matrix between the variables with 95% confidence interval (SACW)

Salinity

Residual (−89.2)

ZWS_1 (85.7)

ZWS_3 (91.9)

 

PO4

MWS_1 (94.4)

MWS_2 (90.0)

MWS_3 (95.6)

MWS_Q (92.1)

NO2

MWS_1 (92.1)

MWS_2 (88.9)

MWS_3 (93.3)

MWS_Q (85.9)

NO3

MWS_1 (96.8)

MWS_2 (93.4)

MWS_3 (97.5)

MWS_Q (87.3)

Chlor a

Residual (90.2)

   

Cirripedia

MWS_Q (88.6)

   

Decapoda

P_A (91.8)

P_B (82.1)

P_SPA (92.8)

 

Cypris

P_A (95.9)

P_B (89.9)

P_SPA (97.5)

 
  1. (b)

    Coastal Waters

     
SST presents a direct correlation with Ascidiacea class in the presence of CW, since this marine animal is native to tropical oceans and is usually found in shallow waters. PO4 and NO3 shows a positive correlation with pressure and ZWS obtained from satellite, respectively, suggesting a relation to the zonal transport of water in the E-NE (onshore) direction. This transport can also explain the concentration of larvae as Mytilidae (mussels Family), Ostreidae (oysters), and Bryozoa (corals, common in the intertidal zone in shallow tropical waters), which show a direct correlation with the ZWS, with grid 1 being the most meaningful. Whereas Cypris is inversely correlated with the pressure and the grid point A is the most significant. In the case of the Bivalvia (class of mussels) occurs a sensibility to the meteorological residual with an inverse correlation between them, that is, a rise in sea level due to meteorological forcing causes a reduction or dispersion of this larva in the region (Table 7).
Table 7

Correlation matrix between the variables with 95% confidence interval (CW)

SST

Asc (26.2)

  

PO4

P_SPA (21.6)

  

NO3

ZWS_Q (20.3)

  

Mytilidae

ZWS_1 (22.2)

ZWS_3 (21.9)

 

Bivalvia

Residual (−20.1)

  

Ostreidae

ZWS_1 (25.0)

ZWS_2 23.6)

ZWS_3 (21.8)

Cypris

P_A (−24.1)

P_B (−22.2)

P_SPA (−21.7)

Bryozoa

ZWS_1 (25.1)

ZWS_2 (21.1)

ZWS_3 (21.6)

  1. (c)

    Tropical Waters

     
Salinity shows a positive correlation (a characteristic of this body of water is to present high levels of salinity and temperature) with SST, ZWS, and MWS, which may be related to evaporation due to the increase of temperature and wind. PO4 and NO3 present positive correlations with atmospheric pressure at remote points and at the coast, characterizing the influence of large-scale atmospheric circulation. In the presence of TW, larvae of Mytilidae, Ostreidae, Cypris, and Bryozoa present correlations similar to those found for the CW, suggesting an interaction of these larvae with the predominant synoptic pattern in the region. The positive correlation with the ZWS characterizes the predominance of wind in the E-NE (onshore) direction. Whereas Cypris also presents negative correlation with the tide, showing sensibility to the variations in sea level associated with astronomical forcing (Table 8).
Table 8

Correlation matrix between the variables with 95% confidence interval (TW)

Salinity

SST (25.3)

MWS_2 (27.2)

ZWS_SPA (23.4)

MWS_SPA (22.2)

PO4

P_A (23.4)

MWS_2 (−20.8)

P_SPA (26.1)

 

NO2

ZWS_Q (20.5)

   

NO3

P_A (25.0)

P_B (26.2)

P_SPA (24.6)

 

Mytilidae

ZWS_1 (−21.2)

   

Ostreidae

P_A (−30.3)

P_B (−32.0)

P_SPA (−29.3)

 

Cypris

Tide (−21.6)

ZWS_3 (24.6)

ZWS_Q (34.1)

 

Bryozoa

ZWS_3 (24.5)

   

Multivariate approaches

In principal component and factors analysis, the most common criteria used for solving the number of components is the eigenvalue criterion, in which we retain and interpret any component with an eigenvalue greater than 1.00. In the present study, 33 variables were analyzed and 10 components were extracted with this criterion. The first component can be expected to have large amount of the total variance. Each succeeding component will account for progressively smaller amounts of variance and they are important enough to be retained for interpretation. Therefore, only the first ten components were retained for rotation. Combined, components 1 to 10 explained 73.43% of the total data variance for upwelling period (spring/summer), where the two most important factor loadings 1 and 2 are zonal wind, pressure, and meridional wind, respectively, accounting for 31.635% of the total variance (Table 9), showing the relationship with north and northeasterly strong wind velocity that occur in this period. The next factor loading 3 accounts for 11.973% and is represented by the larvae of Bivalvia, Briozoa, Decapoda, Isognomon, Mytilidae, Ostreidae, and Polycaeta; other factors are restricted to a few parameters with low variance, indicating that SST, salinity, nutrients, sea level variations, and some larvae do not account for great variations in this period, in spite of the presence of SACW (Table 9). For autumn–winter (not present here), the results suggest a relationship with south and southwesterly winds, characterizing the passage of frontal systems and the presence of cold waters due to Malvinas current. The factors composed by nutrients, some larvae and sea level variations presented also low variance. This analysis was important to show the environment characteristics related to the upwelling periods in the study area
Table 9

Loading variables factors for upwelling period (SON–DJF)

 

Factor 1

Factor 2

Factor 3

Factor 4

Factor 5

Factor 6

Factor 7

Factor 8

Factor 9

Factor 10

Eigenvalue

5.457

4.981

3.951

1.931

1.592

1.498

1.355

1.257

1.142

1.066

Variability (%)

16.539

15.096

11.973

5.852

4.824

4.540

4.108

3.810

3.461

3.230

Cumulative (%)

16.539

31.635

43.608

49.461

54.286

58.826

62.934

66.745

70.206

73.437

Factor loading

 Ascidiacea

0.110

−0.148

−0.033

0.108

0.081

−0.341

0.061

−0.370

−0.039

−0.523

 Bivalvia

0.261

−0.168

0.570

−0.016

0.150

−0.393

−0.210

0.241

−0.056

0.103

 Briozoa

0.314

0.133

0.701

−0.007

0.066

0.297

0.324

−0.107

0.088

−0.016

 Cirripedia

0.106

−0.294

0.323

−0.046

0.017

−0.468

−0.336

0.049

−0.392

−0.005

 Cypris

−0.071

−0.089

0.139

−0.055

−0.139

0.363

−0.645

−0.440

0.073

0.171

 Decapoda

0.147

−0.034

0.576

0.032

−0.190

0.159

−0.490

−0.322

0.083

0.020

 Isognomon

0.354

−0.015

0.802

0.048

0.118

0.066

0.107

0.070

0.009

0.041

 Mytilidae

0.400

−0.021

0.810

−0.035

0.146

0.125

0.177

0.036

0.036

−0.070

 Ostreidae

0.278

−0.063

0.796

−0.071

0.063

0.110

0.238

0.001

0.095

−0.047

 Polychaeta

0.196

−0.034

0.478

−0.089

−0.226

−0.275

−0.245

0.136

−0.051

−0.221

 SST

−0.123

−0.330

0.078

0.668

0.165

−0.239

−0.038

−0.008

0.085

−0.055

 Sal.

−0.003

−0.322

−0.025

0.114

−0.311

−0.390

0.192

−0.101

−0.079

0.126

 DO

0.020

0.138

0.029

−0.433

−0.487

0.128

0.246

−0.155

−0.358

−0.032

 PO4

0.011

0.290

−0.072

−0.248

0.316

0.124

−0.187

0.134

0.037

−0.487

 NO2

−0.037

0.215

−0.032

−0.671

0.391

−0.222

−0.025

−0.120

−0.021

0.110

 NO3

0.023

0.122

−0.084

−0.627

0.396

−0.209

0.014

−0.113

0.044

0.264

 NH4

0.044

0.108

−0.153

0.116

0.110

−0.024

0.087

−0.276

0.474

−0.028

 Chloro a

−0.137

0.125

0.046

−0.321

−0.528

0.124

−0.020

0.184

−0.071

−0.331

 TIDE

−0.364

−0.057

0.019

0.252

0.333

0.243

0.031

−0.284

−0.501

−0.103

 RESIDUAL

−0.465

−0.036

0.075

0.092

0.365

0.199

0.095

−0.159

−0.445

−0.091

 P_A

0.571

0.672

−0.225

0.060

0.060

0.030

−0.115

0.122

0.008

−0.146

 P_B

0.375

0.855

−0.135

0.118

0.044

0.006

−0.109

0.081

−0.019

−0.087

 P_SPA

0.413

0.796

−0.136

0.103

0.101

0.002

−0.121

0.086

−0.015

−0.134

 ZWS_1

−0.782

−0.375

0.121

−0.086

0.099

0.095

−0.102

0.258

0.122

−0.089

 MWS_1

−0.428

0.781

0.228

0.119

−0.011

−0.063

−0.043

0.017

−0.066

0.171

 ZWS_2

−0.807

−0.098

0.142

−0.019

0.094

0.126

−0.130

0.260

0.100

−0.074

 MWS_2

−0.204

0.854

0.158

0.169

−0.076

−0.049

−0.045

0.076

−0.096

0.088

 ZWS_3

−0.720

−0.382

0.119

−0.144

0.111

0.118

−0.096

0.243

0.123

−0.139

 MWS_3

−0.570

0.649

0.250

0.100

0.038

−0.047

−0.056

0.046

−0.033

0.189

 ZWS_SPA

−0.645

0.326

0.208

−0.076

−0.092

−0.236

0.076

−0.275

0.168

−0.211

 MWS_SPA

−0.613

0.320

0.214

−0.117

−0.057

−0.291

0.100

−0.343

0.205

−0.133

 ZWS_Quick

−0.648

0.137

0.224

0.007

−0.186

0.023

0.107

0.087

−0.028

0.058

 MWS_Quick

−0.524

0.551

0.228

0.165

−0.087

−0.138

0.105

0.032

−0.018

0.094

Data grouped according to the seasonality were analyzed for both spring–summer (upwelling period) and autumn–winter (downwelling period) clusters. The dendrogram for upwelling period showed two big clusters, one of physical variables as wind stress and sea level variations in one side and biogeochemical ones at the other, corresponding to the macrostructure of ecosystem (Fig. 11). The groups, ZWS (Q/SPA), MWS (1/2/3), ZWS (1/2/3), and sea level variations form one cluster, whereas, P(SPA/A/B), Ost/Myt/Iso/Bri, nitrite/ nitrate/phosphate, chlorophyll a/oxygen, Cyp/Dec, Pol/Cir/Biv, ammonium, and salinity/SST/Asc form the other cluster. The degree of refinement of similarities is expressed in 12 groups as a whole, which can be written as group 1 (physicochemical, except the class of Ascidiacea), group 2 (classes of mollusks, barnacles and oysters), group 3 (genus of crustaceans), group 4 (DO and Chlorophyll a, important elements for trophic structure), group 5 (nitrogen concentration and phosphorus), group 6 (genus and families of mussels, oysters and mosses), group 7 (pressures), group 8 (Tides), group 9 (zonal wind 1/2/3), group 10 (meridional wind 1/2/3), group 11 (zonal and meridional wind in SPA), and group 12 (zonal and meridional wind in Satellite point). The groups of larvae appear to have been separated according a relationship of taxonomy. This cluster analysis shows the importance of the zonal and meridional wind variability, including this variable in a single great cluster as noted in the first two factors of PCA.
https://static-content.springer.com/image/art%3A10.1007%2Fs10661-011-2470-3/MediaObjects/10661_2011_2470_Fig11_HTML.gif
Fig. 11

Ward dendrogram with Pearson r coefficient (SON–DJF)

Conclusion

The meteorological data set obtained from remote grid points present a well-defined seasonality for the mean wind stress with high values for the spring–summer months, with the month of September being the most meaningful. They also show higher percentages of northeasterly wind stress for the same period; mainly in the points more distant than Arraial do Cabo.

Seasonal variability of the temperature shows an increase of this variable in April and a decrease of the nutrients in March which can be related to the oligotrophic TW mass. The opposite is verified in spring and summer periods (except for the PO4) which can be associated with the presence of the SACW mass. September is also a critical month with minimum value of SST and maxima of total nitrogen, DO, and PO4, being in accordance with the seasonal variability of the northeasterly wind stress. Whereas the minima values of chlorophyll a occur in this month and it may be related to the high wind stress values.

Seasonal variations relating to the abundance of larvae in Anjos Bay show that November, December, February, and March are of critical importance and these months are related to the occurrence of upwelling in this region.

The meaningful cross-correlations are verified between zonal wind, DO, and PO4 and between SST, nutrients, and chlorophyll a with 1-week lags and zero negative lag, respectively. We suppose that the low values of correlations (below 40%) can be relating with the internal localization of the samples water harvest point or also with the sample dimension. Seasonal correlation shows that nitrogen concentration and phosphate present the major correlation with the wind stress obtained from remote grid points 1, 3, and satellite for the months of November and January (upwelling period). The cross-spectrum analysis presents significant coherence for peaks with period of 51 weeks, approximately 1 year, which can be associated with the upwelling in this region.

Chlorophyll a and ammonium (NH4+) present a positive anomaly with an increase of chlorophyll a from November 2004 and for ammonium between April and October 2006. Therefore, it is important to monitor these elements because they can be related to the decomposition of organic matter and may be associated with local human activities.

Correlation matrix analysis for each water masses was useful to verify the interaction between the physical parameters and the biochemical response in term of nutrients and larvae. For the SACW analysis, the larvae of crustaceans show a positive correlation with all points of pressure and meridional wind, suggesting a greater sensitivity of these larvae and nutrients associated with resurgence of this water mass. For the CW, the SST presents a direct correlation with Ascidiacea; nutrients, pressure and zonal wind stress. In the presence of TW, the predominant water mass, physical and biological parameters present correlations similar to those found for the CW, characterizing the influence of large-scale atmospheric circulation, showing the interaction with the synoptic pattern over the region.

Spatial and temporal variables analyzed with multivariate approach show meaningful seasonality variance of the physical and biological samples, characterizing the principal components responsible for this variance in spring and summer (upwelling period), emphasizing the monitoring of species as crustaceans and mussels that are present in the local economy. Therefore, the knowledge about the seasonality of physical and biological relationship is not only important for understanding the dynamics of the ocean–atmospheric interactions, but also in developing models to apply in coastal zone management.

Acknowledgments

The authors thank the Admiral Paulo Moreira Institute of Marine Studies—IEAPM of Brazilian Navy for data availability and logistical support. The authors also thank the financial support of the Coordination for the Improvement of Higher Level Personnel—Brazilian Research Agency (Capes). Appreciation and thanks are also given to anonymous reviewers for their constructive comments and suggestions to improve the manuscript.

Copyright information

© Springer Science+Business Media B.V. 2011