Tripton

The inanimate particulate matter, or tripton, of natural waters is the fraction whose light absorption properties have received the least attention because they are so difficult to measure. At typical concentrations the material does not absorb light strongly but scatters quite intensely and so its absorption properties cannot be characterized by normal spectropho-tometry with long-pathlength cells. One approach to overcoming these problems is to collect the particulate matter from a natural water sample on a filter and resuspend it in a much smaller volume: the absorption spectrum of the concentrated material is then measured in a short-pathlength cell, placed at the entrance port of an integrating sphere, and from the absorbance values the absorption coefficients due to the particu-late matter, in the original water body, may be calculated.701 Such measurements still have to be corrected for the attenuation due to backscattering. To improve the procedure, Babin and Stramski (2002) placed the sample cuvette at the centre of a large (15 cm diameter) integrating sphere: the scattering error became negligible. An alternative approach is to measure the total absorption spectrum of the water sample with an ICAM or PSICAM, taking advantage of the long equivalent pathlengths and immunity to scattering error of these instruments (§3.2, above), repeat the measurement on a filtrate and obtain the spectrum of the particulate fraction by difference.

The total particulate fraction (seston, in limnological parlance) will of course include phytoplankton as well as tripton. If, however, the phyto-plankton concentration is low then the particulate fraction spectrum can be attributed to tripton. Figure 3.6 shows the absorption spectra of the particulate and soluble fractions of several Australian inland waters. In all cases except (c) and (f) the absorption by the particulate fraction is almost entirely due to tripton, although in two of these ((d) and (g)) a small 'shoulder' at about 670 nm due to phytoplankton chlorophyll may be seen.

To obtain a tripton spectrum uncontaminated by phytoplankton absorption, Kishino et al. (1985) extracted the filter on which the

{d) L. Burley Griffin

{d) L. Burley Griffin

350 450 550 650 750

Fig. 3.6 Comparison of the spectral absorption properties of the particulate and soluble fractions of various Australian inland waters in the southern tablelands of New South Wales/Australian Capital Territory (from Kirk, 1980b). The absorption spectrum of pure water is included in (a) for comparative purposes. The phytoplankton chlorophyll a contents (Ca) and nephelometric turbidities (Tn) of the different water bodies at the time of sampling were as follows: (a) Corin Dam, 8 June 1979, Ca = 2.0 mg m~3, Tn = 0.51 NTU; (b) Lake Ginninderra, 6 June 1979, Ca= 1.5 mg m"3, Tn= 1.1 NTU; (c) Cotter Dam, 8 June 1979, Ca = 9.0 mg m"3, Tn= 1.6 NTU; (d) Lake Burley Griffin, 6 June 1979, Ca = 6.3 mg m"3, Tn = 17.0 NTU; (e) Googong Dam, 21 June 1979, Ca = 1.7 mg m"3, Tn = 5.8 NTU; if) Burrinjuck Dam, 7 June 1979, Ca = 16.1 mg m"3, Tn= 1.8 NTU; (g) Lake George, 28 November 1979, Ca = 10.9 mg m"3, Tn = 49 NTU. The ordinate scale corresponds to the true in

350 450 550 650 750

350 450 550 650 750 350 450 550 650 750 850 Wavelength (nm)

Fig. 3.6 Comparison of the spectral absorption properties of the particulate and soluble fractions of various Australian inland waters in the southern tablelands of New South Wales/Australian Capital Territory (from Kirk, 1980b). The absorption spectrum of pure water is included in (a) for comparative purposes. The phytoplankton chlorophyll a contents (Ca) and nephelometric turbidities (Tn) of the different water bodies at the time of sampling were as follows: (a) Corin Dam, 8 June 1979, Ca = 2.0 mg m~3, Tn = 0.51 NTU; (b) Lake Ginninderra, 6 June 1979, Ca= 1.5 mg m"3, Tn= 1.1 NTU; (c) Cotter Dam, 8 June 1979, Ca = 9.0 mg m"3, Tn= 1.6 NTU; (d) Lake Burley Griffin, 6 June 1979, Ca = 6.3 mg m"3, Tn = 17.0 NTU; (e) Googong Dam, 21 June 1979, Ca = 1.7 mg m"3, Tn = 5.8 NTU; if) Burrinjuck Dam, 7 June 1979, Ca = 16.1 mg m"3, Tn= 1.8 NTU; (g) Lake George, 28 November 1979, Ca = 10.9 mg m"3, Tn = 49 NTU. The ordinate scale corresponds to the true in particulate fraction had been collected, with methanol. This removes the chlorophylls and carotenoids but leaves the humic colour. Methanol does not, however, remove the biliprotein pigments that will be present in any cyanobacteria or cryptophytes algae in the sample. As an alternative, Tassan and Ferrari (1995) have developed a procedure in which the filter is treated with sodium hypochlorite. This bleaches the algal pigments, including the biliproteins.1345,392 It is assumed in this procedure that the short exposure to hypochlorite does not affect the non-algal pigments.

The tripton absorption spectra all have much the same shape: absorption is low or absent at the red end of the spectrum and rises steadily as wavelength decreases into the blue and ultraviolet. These are typical humic substance absorption spectra and indeed have much the same shape as those of the dissolved yellow materials. Furthermore, typical tripton samples collected on a filter are brown in colour. The most plausible supposition is that the yellow-brown tripton colour is largely due to particulate humic material existing either bound to mineral particles or as free particles of humus. It seems likely that in inland waters it arises, together with the soluble humic material (gilvin), from the soils in the catchment. In productive waters or in oceanic waters well away from land drainage, some of the light-absorbing inanimate matter arises by decomposition of the phytoplankton. The detrital (non-living, particu-late) fraction in sea water also has an absorption spectrum of the humic type (Fig. 3.7) but sometimes with shoulders due to the breakdown products of photosynthetic pigments.164,962,1146,615

The spectral slope of the spectrum of non-algal particulate matter is typically less than that observed for CDOM. For coastal waters around Europe, Babin et al. (2003b) found the spectral slope for non-algal particles to have an average value of 0.0123 ± 0.0013 nm"1. In coastal waters

Wavelength (nm)

Fig. 3.7 Absorption spectra of detrital and phytoplankton particles from a mesotrophic station in the Sargasso Sea (after Iturriaga and Siegel, 1989).

Wavelength (nm)

Fig. 3.7 Absorption spectra of detrital and phytoplankton particles from a mesotrophic station in the Sargasso Sea (after Iturriaga and Siegel, 1989).

in the Northern Gulf of Mexico, influenced by Mississippi River outflow, D'Sa et al. (2006) found an average value of 0.011 nm-1 for the spectral exponential slope of this fraction.

It is commonly assumed that absorption by the living and non-living particulate matter of natural waters in the near-infrared, ~ 720 to 750 nm, is negligible, and on the basis of this assumption, measurements in the near-IR are commonly used to provide the scattering correction for absorbance measurements. Using their improved integrating sphere method (above), Babin and Stramski (2002) have shown that this assumption is indeed valid for a variety of particles - phytoplankton species, phytoplankton detritus and coloured mineral particles - of the type likely to be encountered in marine or inland waters.

The in situ absorption coefficient due to particulate matter at 440 nm is a convenient general measure of particulate colour in any water. A suitable symbol would be p440, which is analogous to g440 previously defined as a measure of soluble colour, Table 3.2 lists some observed values ofp440 for various natural waters. In turbid waters containing large amounts of suspended material derived from soil erosion in the catchment or (in shallow waters) wind resuspension of sediments, non-living particu-late absorption can exceed absorption due to dissolved colour. The waters of the Murrumbidgee Irrigation Area (Table 3.2) are examples of the first situation (see also Fig. 3.6d) and Lake George, NSW, is an example of the second (Fig. 3.6g; Table 3.2). Rivers frequently become more turbid and coloured with increasing distance downstream, from the headwaters to the estuary. A case in point is New Zealand's longest river, the Waikato. Davies-Colley (1987) found a progressive increase in both particulate (p440) and soluble (g440) colour in this river down its 330 km path from Lake Taupo to the sea (Table 3.2): the water is clear and blue-green as it leaves the oligotrophic lake, but is yellow and turbid when it finally enters the Tasman Sea.

In some coastal waters, wind-blown dust arising from dust storms in the dry interiors of nearby continents can constitute a significant proportion of the suspended particulate matter. 'Red-rain' events, loaded with Saharan dust are, for example, common along the Mediterranean coast. Samples of Asian mineral dust were found to have mass-specific absorption coefficient values at 440 nm ranging from ~0.028m2g-1 from a Chinese desert to 0.15 m2g-1 for volcanic soil dust from Korea.1309 In dust, or soil-derived, particles with low organic matter content, a yellow or reddish colour where present is typically due to the presence of iron oxide. Samples with high concentrations of iron, instead of smooth, exponentially rising, spectra, have substantial shoulders in the blue-green and near-UV regions.49

It should be realized that p440 is not as good a general guide to particulate colour as g440 is to soluble colour. Whereas all gilvin spectra have approximately the same shape, the shape of the particulate fraction spectrum can change significantly in accordance with the proportion of humic material or phytoplankton (see below) in it. Nevertheless, in those waters (of common occurrence) with a substantial particulate humic component, it is a useful parameter.

Was this article helpful?

0 0

Post a comment