Whole Ecosystem Evidence of Eutrophication Enhancement by Wetland Dredging in a Shallow Tropical Lake
- First Online:
- Cite this article as:
- Marotta, H., Bento, L., Esteves, F.A. et al. Estuaries and Coasts (2009) 32: 654. doi:10.1007/s12237-009-9152-1
The purpose of this research was to assess the effects of dredging performed in a marginal wetland colonized by aquatic macrophytes on eutrophication of the adjacent shallow tropical lake (Imboassica Lake, Brazil). The river mouth of the Imboassica River that drains into Imboassica Lake had been densely colonized by aquatic vegetation dominated by Typha domingensis (Pers.) when it was dredged. Total and dissolved nitrogen and phosphorus concentrations were measured monthly over 13 years at four stations in the Imboassica river-lake system. Dredging activities reduced phosphorus and nitrogen retention at the river mouth and subsequently increased these nutrient stocks in the lake waters. Nutrient retention by non-dredged wetland was estimated to be ca. 1,200 kg year−1 (87.3 g m−2 year−1) for nitrogen and 60 kg year−1 (4.5 g m−2 year−1) for phosphorus. Our whole-lake approach suggested that dredging might intensify rather than mitigate eutrophication in shallow tropical lakes when the removal of aquatic macrophytes is coupled to the persistence of anthropogenic nutrient inputs from the watershed.
KeywordsDredging Nutrient retention Eutrophication Shallow tropical lakes Typha domingensis (Pers.) Wetlands
Lakes are aquatic ecosystems distributed broadly in the terrestrial surface (Downing et al. 2006), constituting an important destination for allochthonous (Lennon 2004; Sobek et al. 2005) and autochthonous (Brevik and Homburg 2004) matter in the watershed. Organic and inorganic inputs into lakes may originate from anthropogenic point and non-point sources (Arbuckle and Downing 2001). One recognizable consequence of pollution by nutrient-enriched effluents is the growth of submerged (Rooney et al. 2003), floating (Kadlec and Knight 1996) and emergent (Gophen 2000) aquatic macrophytes.
Emergent aquatic macrophytes are often abundant in wetlands—transitional habitats between aquatic and terrestrial ecosystems that are inundated or saturated by water during part or all of the year (Neue et al. 1997). Wetlands may mitigate eutrophication at low cost (Verhoeven and Meuleman 1999), showing great potential as complementary sewage treatment systems, even for big cities (Costa-Pierce 1998). Aquatic macrophytes reduce organic matter inputs (Wentz 1987) and contribute to retention of nutrients, mainly nitrogen (Fleischer et al. 1991) and phosphorus (Kadlec 1997), in aquatic ecosystems. This nutrient retention, or the difference between nutrient inputs and outputs, is caused by biological activity, sedimentation and chemical reduction to gases (Saunders and Kalff 2001). The role of emergent aquatic macrophytes in nutrient retention might be especially important in shallow lakes at low latitudes, where tropical conditions favor the colonization of these plants in marginal wetlands (Wetzel 1990).
Eutrophication contributes to silting in lakes because it increases autochthonous primary production and organic matter sedimentation (Brevik and Homburg 2004). The removal of peat sediments through dredging is considered an important intervention to restore the nutrient assimilative capacity of enriched lakes by reducing the internal loading from the sediment to the water column (Ruley and Rusch 2002). Hence, dredging may restore eutrophic aquatic ecosystems, increasing depth and reducing nutrient stocks (Sagehashi et al. 2001). However, this intervention may also remove the aquatic macrophytes that play an important role in nutrient retention in lakes (Rooney et al. 2003).
Here, we assessed the effects of dredging in a marginal wetland densely colonized by aquatic macrophytes on the eutrophication of the adjacent shallow tropical lake. This wetland is situated in a small river mouth that drains into the lake. Dredging restored fluvial depth and water flow into the lake by removing a large amount of sediment and aquatic vegetation, growth of which had been encouraged by the shallow depth and input of anthropogenic-generated nutrients. However, nutrient input from sewage into the river remained unchanged, allowing us to assess the role of dredging and removal of aquatic macrophytes on lake eutrophication at an ecosystem scale. We compared nitrogen (N) and phosphorus (P) concentrations in the water column of both river and lake for three periods over 13 years: two before and one after the dredging. Our hypothesis was that dredging with removal of aquatic macrophytes that was not coupled with a reduction in nutrient input would increase eutrophication in a shallow tropical lake.
Materials and Methods
Sampling Stations and Periods
More Pristine (PRIST)—from May 1992 to December 1994 (32 months): Imboassica Lake and Imboassica River showed more oligotrophic characteristics. The river mouth of the Imboassica river-lake system had a permanent water flow (ca. 5 m3 s−1) and aquatic macrophytes were absent along the river shore.
Before Dredging (BD)—from January 2000 to June 2002 (30 months): the mouth of Imboassica River became densely colonized by T. domingensis due to eutrophication. In this area, a permanent superficial water column persisted only at the RIV station, as the river section from this point to Imboassica Lake no longer flowed superficially because of silting, showing only a groundwater flow. This lake had also been eutrophicated by sewage inputs directly from the margins; the average flow in the main discharge channel was estimated at 82.5 L s−1 (Lopes-Ferreira 1998).
After Dredging (AD)—from May 2003 to October 2005 (28 months): dredging of the river channel restored the water flow between Imboassica River and Imboassica Lake to values similar to those of the PRIST period (ca. 4 m3 s−1). However, comparing the periods before and after dredging, nutrient discharges in the Imboassica River were not enhanced. A marked reduction in the sewage inflow directly from the margins of Imboassica Lake with respect to the BD period was measured, showing mean values of 16 L s−1 in the main discharge channel (Silva 2006).
Assessment of Dredging Effects on Eutrophication in the Water Column of Imboassica Lake
Water samples were collected with a Van Dorn bottle, transported to a field laboratory, filtered through 1.2 µm glass fiber filters (GF/C Whatman) for dissolved forms analyses and frozen. Total and dissolved N concentrations were obtained by the sum of Kjeldahl nitrogen (Mackereth et al. 1978) and nitrate (Golterman et al. 1978) concentrations. Total and dissolved P concentrations were determined by the blue molybdenum method (Golterman et al. 1978).
Transformed data did not meet the assumptions of parametric tests (Zar 1996); the data were not normally distributed (Kolmogorov-Smirnov, P < 0.05) and variances were heterogeneous (Bartlett, P < 0.05). Medians and a non-parametric test were used to compare sampling groups (Kruskal-Wallis, P < 0.05; Zar, 1996). Statistical tests were carried out using GraphPad Prism 4.0. After this step, pair-to-pair differences were assessed with False Discovery Rate (FDR; P < 0.05), considering P-values from Mann-Whitney tests and a specific lambda equal to zero (Benjamini and Hochberg 1995). FDR was calculated using the software Q-VALUE, as this approach may be more powerful than traditional Bonferroni corrections and deserves more frequent use in ecological field studies, especially in those involving a relatively large number of multiple comparisons (Garcia 2004).
Dredging effects on standing stock of total nitrogen (N) and phosphorus (P) in the waters of Imboassica Lake
Standing stock of total nutrients
Absolute increase (kg)
Increase rate (kg year−1)
Dredging effects in water (kg year−1)
Between BD and PRIST
Between AD and BD
Between BD and PRIST
Between AD and BD
Nutrient inputs from the watershed over 13 years resulted in a general enrichment of N and P at all sampling stations. The highest N and P concentrations observed at the river station when the river mouth was silted up (before dredging) might be attributed to factors other than just increases in sewage discharges. The very low water flow also influenced this result, as this would reduce nutrient dilution and contribute to eutrophication in natural waters (Schindler 2006).
Despite the strong eutrophication in Imboassica River, the lake station closest to this river (MAC) did not show a significant nutrient enrichment before dredging, probably due to dense colonization by aquatic macrophytes in the Imboassica river–lake system during this sampling period. Wetlands colonized by aquatic macrophytes can remove significant amounts of P (Kadlec 1997) and N (Saunders and Kalff 2001) from aquatic ecosystems. In our study, the removal of aquatic macrophytes with decreases in water residence time by dredging was related to a strong increase in N and P standing stocks in Imboassica Lake.
Retention of N and P per area in this tropical wetland, which is densely colonized by aquatic macrophytes, was estimated as about two orders of magnitude lower than often described for N (Arheimer and Wittgren 2002; Jiang et al. 2007; Blankenberg et al. 2008) and P (Uusi-Kamppa et al. 2000; Jiang et al. 2007) in highly nutrient-enriched constructed wetlands (effluent treatment systems). However, nutrient retention in the Imboassica river–lake system before dredging was comparable or higher than one currently described in natural or more oligotrophic constructed wetlands for N (Jansson et al. 1998; Fisher and Acreman 2004; Silvan et al. 2004) and P (Mitsch and Reeder 1991; Fisher and Acreman 2004; Silvan et al. 2004). Therefore, our results confirm the potential role of aquatic macrophytes in retaining excess nutrients (Hutchinson 1970; Demars and Harper 1998), and in reducing nutrient pollution of eutrophic lakes (Steinmann et al. 2003).
The removal of internal nutrient loading in the water–sediment interface by dredging with increases in the water flow can contribute to the restoration of aquatic ecosystems (Lohrer and Wetz 2003; Jiang and Shen 2006), but this intervention was related positively to nutrient enrichment in the studied lake. Many studies have independently assessed the role of dredging (Sagehashi et al. 2001) and aquatic vegetation (Steinmann et al. 2003) on the mitigation of aquatic eutrophication. On the other hand, the effects of drastic removal of large stands of aquatic macrophytes by dredging are still not well known. Our results suggest that eutrophication in a shallow tropical lake was significantly enhanced after dredging, confirming the proposed hypothesis. In conclusion, dredging may be considered an effective engineering technology for restoration of lakes (Jiang and Shen 2006), but might intensify rather than mitigate eutrophication when the removal of aquatic macrophytes is coupled with continued anthropogenic nutrient input. Dredging needs integrated planning related to decreases in nutrient inputs from the watershed, especially in tropical shallow lakes where aquatic macrophytes may be important players in nutrient cycling.
The authors would like to acknowledge PETROBRAS for financial support of the ECOLagoas Project and logistic support in NUPEM/UFRJ. We wish to thank Dr. David Bastviken for useful advice, and the two anonymous reviewers for valuable comments on an earlier draft of this manuscript. We are also grateful to the Brazilian funding agencies FAPERJ, CAPES and CNPq for scholarships and financial support.