Distribution of Drift Pumice as an Indicator for Pollutant Distribution and Sediment Transport in New Caledonia

Introduction

New Caledonia (35°S 165°W) (Figure 1) is a ~350 km long by 50 km wide island nation located in the Southwest Pacific, approximately 1500 km away from Australia and New Zealand. The main island is enclosed by a nearly continuous barrier reef creating a lagoon of over 20,000 km2 in area. The barrier reef generally grows 14 to 25 km from the coast, but can reach distances of 100 km along the axis of elongation. In 2008 the lagoon was designated a UNESCO World Heritage Site [1]. This honor brought attention to the plight of the lagoon, which has been suffering from the detrimental effects of nearly a century of open cast mining. Although nickel mining reached its peak in the 1960s to 1970s, the industry still accounts for a significant portion of the New Caledonian economy. The environmental impacts of open cast mining include acid mine drainage, destruction of littoral and coastal habitats, rapid shallowing of bays and reduced coral productivity in the presence of chronically elevated concentrations of suspended sediment [2, 3, 4, 5]. However, the full extent of environmental degradation is unknown, as there has only been one comprehensive catalog of the effects of mining on the rivers and coastline of New Caledonia [5] More recent studies of the lagoon ecosystem have focused on the southwestern part of the lagoon surrounding the population center of Nouméa [6, 7, 8, 9].

Click to enlarge

Click to enlarge

Figure 2: Map of Grande Terre, New Caledonia (after General Bathymetric Chart of the Oceans, 2010), with detailed insets of (A) the Baie de St. Vincent and (B) the Nouméan peninsula and Mt. Dore (after DTSI, 2009, 2010). Coral reefs (after Andréfouët et al., 2005) in dark pink shallow, pink (intermediate depth) , and light pink (deep). Circles represent sampling locations from 2008 (white) and 2010 (black). Legend shows elevation in meters above means sea level.

Black star marks the location of Taupo Volcanic Zone, New Zealand. Black box marks the) boundaries of Figure

2. Black and grey lines represent modern and relict tectonic margins, respectively (Colored lines represent generalized ocean currents, with arrows pointing toward the flow [10]) direction after [11, 12, 13, 14]. Ocean currents: EAC, East Australian Current; SEC, South Equatorial Current; STCC, South Tropical Counter Current. Smaller studies have also been carried out on the east coast and in the extreme north and south [6, 15, 16]. Our study used land based data collection methods to map the distribution of long shore and fluvial sediments around the island perimeter. The field study included two sample collection periods during which exotic drift pumice was collected from a total of 40 distinct locations. Samples were collected from 9 beaches in July 2008 and 33 beaches in July 2010 (Figure 2). Petrographic and geochemical compositions of the pumice samples were consistent with two recent eruptions in the Central Tonga arc. A basic drift trajectory was mapped from the proposed source, while assuming the pumice generally followed major ocean currents [17, 18]. The dominant ocean current from the equator to approximately 20°S is the SEC, which flows east to west and could potentially be enriched in pumice from Tonga, the Lau Basin, or Vanuatu prior to reaching New Caledonia [18, 19, 20]. At the eastern New Caledonian coast, the SEC bifurcates, with the North Jet recombining with the main body of the SEC and the South Jet wrapping around the southern terminus of New Caledonia to flow northward up the west coast before continuing west toward Australia [21, 14]. Upon reaching the Australian coast, the SEC bifurcates again, with the south, flowing arm transitioning into the Eastern Australian Current [22]. East flowing off shoots of the EAC combines to form the South Tropical Counter Current (STCC), which breaks against the northwestern coast of New Caledonia, but would likely be poor in pumice until passing the Kermadec is lands and New Zealand further to the east [22, 13]. Hence, ocean currents potentially rich in pumice were shown to have access to nearly the entire New Caledonian coastline except for the deeper bays and a stretch of the coastline in the northwest, which was serviced by pumice poor waters coming from Australia. The intensity of longshore waves imported from the open ocean directly corresponded to pumice abundance, allowing the energy distribution of these waves, which are critical to flushing sediment from nearshore areas, to be mapped using land based data collection methods [23]. A simplified model comparing pumice abundance to other beach parameters (width, composition, etc.) can then be used to determine whether fluvial sediments are polluting nearshore areas or being exported to the lagoon. The results were compared to previously published energy distribution maps created from a highly. Complex three dimensional hydrodynamic box model of the southwestern lagoon developed by the French- government [7, 24, 9].

Materials and Methods

Detailed field observations were recorded at all of the beaches centered around descriptions of the beach in terms of the width of the beach along its profile, the size distribution of beach sediments, the amount of subaerial sediments and pumice in the beach backshore, and the landform acting as the landward boundary. Beach widths were measured from the low tide line, including exposed bedrock but not (ephemeral) sand bars, to the upper limit of wave activity at the base of vegetation, cliffs, or a seawall. The width of the beach was initially described as narrow (<5 m), intermediate (5 to 40 m), or wide (>40 m), and later quantified using a combination of site photographs and satellite images. Satellite images were al so used to quantify the width of the lagoon from the beach to the outer barrier reef, as well as to verify the presence of fluvial sediments, as indicated by brown discoloration of nearshore waters traceable to a nearby fluvial source. Beach compositions encountered in the field included clay to silt sized mud, fine to very coarse sand, very fine pebble gravel to coarse cobble gravel, and bedrock or boulder gravel. The last two size categories were combined due to similar rarity of pumice. Pumice samples were collected with a bias toward large clasts, so only the sieve diameter, in φ, of the largest sample was recorded for each location. Pumice abundances were rated as follows: (1) very abundant meant there was a pumiceous ridge deposit, (2) abundant meant there was a pumiceous strand line or the volumetric equivalent, (3) rare meant >10 minutes were required to fill the sample bag to an acceptable level, and (4) very rare meant only one sample was encountered. Beaches without any pumice were not encountered.

Discussion

On the macro scale, pumice abundance was directly related to the width of the lagoon adjacent to the beach. However, this relationship did not continue into the meso scale. For example, there were localized sandy beaches in the deeper bays with anomalously abundant pumice, and there were also narrow, cliffed beaches on the open coast with anomalously rare pumice. Furthermore, pumice was always rare to very rare in the northwest, regardless of other factors. A simple explanation is that on the macro scale, lagoon width was directly proportional to longshore wave energy, which in turn was inversely proportional to the distance from the barrier reef [7, 24]. However, on the meso scale, bathymetric and topographic irregularities created fluctuations in the wave energy distribution that were reflected in the pumice distribution [11, 8, 24]. Modeled the energy distribution of surface currents in the New Caledonian lagoon from the Baie de St. Vincent south. Wave energies were reported in terms of the surface water export time, θ, which is the number of days needed to replace an initial volume of surface water in the lagoon with new water from the open ocean. Pumice was abundant wherever θ was below a threshold value that varied by region. Specifically, in the Baie de St. Vincent, pumice was abundant wherever θ was less than 8 days, which increased to 15 days in the bays surrounding the Nouméa peninsula. Only two beaches in the modeled area did not conform to this model (10-06, 10-29), possibly due to unmodeled complexities in water flow at the micro scale in structurally controlled bays [30]. In Boulari Bay and the Baie de St. Vincent, two sandy beaches that were previously described as having anomalously abundant pumice (10-08, 10-30) actually corresponded to localized energy highs where θ dropped below the regional threshold values. Additionally, in one of these high energy areas, there was a narrow, cliffed beach with rare pumice (10-07), similar to ones on the open coast, supporting the idea that pumice was rare at these beaches not because of a low θ but because there was no supratidal depositional area. Extrapolating to the rest of the island perimeter, pumice should have been abundant wherever θ was below the regional threshold value, provided there were no topographical restriction ns to pumice deposition. However, since the beach characteristics observed in the northwest did not deviate significantly from other beaches around the island, there must have been some other, unidentified factor involved in pumice distribution.

Since most of the beaches around the island were supplied pumice by long shore waves derived from the SEC, it could be that pumice was universally rare in the northwest because of a lack of pumice in long shore waves derived from the STCC. In addition to supplying pumice, long shore waves provided the dominant means of removing sediment from a beach, as well as supplying sand produced by sorting during long shore drift [28]. Sandy beaches were typically of intermediate width (5 to 40 m) and had abundant pumice, suggesting that long shore sediments only accumulated where θ was below the regional threshold (e.g on the open coast). Where long shore input lagged behind removal, either due to low θ (e.g. deep in bays) or headland bypass, beaches were composed of bare bedrock or locally derived gravels. Conversely, beaches near fluvial point- sources became dominated by fluvial mud or gravel that could potentially overwhelm long shore sediment removal, increasing beach widths to well above the 40m attained by sand beaches. Taking narrow (0 to 5 m) to intermediate (5 to 40 m) width beaches as “healthy”, it becomes apparent that of the beaches with fluvial sediments, “healthy” beaches tended to be located on the open coast (i.e. θ < threshold value, pumice abundant unless restricted by topography; 10-12, 10-17, 10-19, 10-23, 10-27 to 28). If embayed, “healthy” beaches were either somewhat removed from the dominant fluvial source (i.e. θ > threshold value, pumice rare; 10-11, 10-32 to 33), or situated within a localized energy high (10-07, 10-24). Conversely, “unhealthy” beaches (i.e. those wider than 40 m) tended to be situated very near the river mouth (10-21) or in secondary depositional areas downstream, such as mangrove swamps (not sampled), tidal lagoons (10-29), or mudflats (10-05). However, for “healthy” beaches with low  values and either abundant pumice or rare pumice with cliffs in the backshore, further research would be needed to determine the volume of sediment bypassing each beach as well as to determine the ultimate destination of the sediment, such as a secondary depositional area downstream, or out to the lagoon, where elevated turbidity could affect corals.

Conclusion

The pumice samples collected in 2008 and 2010 from beaches around New Caledonia were of fairly uniform geochemistry and were shown to be from two recent eruptions in Central Tonga. The 2008 samples were dominantly from a 2006 eruption of Metis Shoal, but the

2010 samples were dominantly from a 2001eruption of an unnamed volcano. From 2008 to 2010, the Metis Shoal samples were likely removed from beaches by storms, exposing older pumice at higher elevations as the dominant pumice deposits. This natural attenuation of pumice by high energy waves caused pumice abundances to naturally mimic wave energies, providing a land based method of measuring wave intensities. On a macro scale, pumice abundance corresponded inversely with the width of the lagoon from the barrier reef to the beach. However, this relationship broke down on the meso scale, where bathymetric and topographic irregularities altered the wave energy distribution. On the meso scale, pumice abundance no New Caledonia drift pumice: implications for sediment transport 10 longer corresponded to lagoon width, but continued to correspond directly to wave energy. Furthermore, this relationship continued into the micro scale, to a resolution not attained by a published energy distribution map by Jouon et al. (2006) [24]. By using beach width as an indicator of the presence and extent of hyper sedimentation in the near shore area, pumice abundance can be used to determine whether sedimentation is restricted to the area immediately surrounding the beach, or whether sediment bypass might be contributing to hyper sedimentation downstream. This application could be useful for determining where to search for the places worst hit by mining or urban waste.

Acknowledgements

The authors would like to thank the National Science Foundation (OISE IRES Grant 0727323 to Dr. K.N. Nicholson), and Ball State University (ASPiRE Research Grant I215 10 to Ariel Stewart).