Page 355 - Soil and water contamination, 2nd edition

342 Soil and Water Contamination

below the soil root zone and that land use changes tend to reduce both the length of the flow
paths and the nitrate removal capacity along the flow paths.
During transport through the river network, contaminant loads are attenuated by
physical and biogeochemical processes. Seitzinger et al. (2002) related N removal in sixteen
eastern US river networks to the Strahler stream order (Figure 18.4). In the Strahler stream
classification system, a first-order stream has no tributaries; where two 1st order channels
nd
nd
join, a stream of 2 order is formed and where a 2 order stream confluences with another
rd
2nd order stream, a 3 order stream is formed, and so on (Strahler, 1952). In general, the
contribution of each stream order to total stream length in a river network decreases with
increasing stream order. Seitzinger et al. (2002) found that reach-specific N removal (and
thus the first-order denitrification constant: see Section 13.3) is greatest in the smaller,
lower-order streams (Figure 18.4a). This can be explained by the increase of N removal with
decreasing water depth (Alexander et al. 2000; see Table 13.3) and by the fact that lower-
order streams are shallower than higher-order streams. In contrast to the reach-specific N
removal, the total mass of nitrogen removed per unit stream length increases with increasing
stream order, since the discharge and N load increase with increasing stream order (Figure
18.4b). Overall, 37% to 76% of the total N input to the rivers studied by Seitzinger et al.
(2002) is removed in the river network, and approximately half of this is removed in 1st
through 4th order streams, which account for 90% of the total stream length (Figure 18.4c).
The other half is removed in higher-order rivers, which account for only about 10% of the
total stream length (Seitzinger et al., 2002). This means that the proportion of the N inputs
that is removed to the catchment outlet is smaller for the higher-order streams, because in

Figure 18.4 Relationship between Strahler stream order for the Penobscot, Connecticut and Kennebec river
networks and a) the proportion of N input to a reach that is removed within that reach (average ± s.d.), b) the
average amount of N (kg) removed per meter of stream/river with uniform direct catchment N loading to river
-2
-1
(100 kg N km y ), and c) the proportion of N input into a particular Strahler order that is removed within the
river network. Adapted from Seitzinger et al. (2002).

10/1/2013 6:47:11 PM
Soil and Water.indd 354 10/1/2013 6:47:11 PM
Soil and Water.indd 354