Introduction

Erhai Lake (100°05′–100°07′E, 25°35′–25°58′N) is in the Dali Bai Autonomous Prefecture, and it is the second largest plateau freshwater lake in Yunnan Province, China. It is a rift lake. It is known for its attractive scenery, and is often referred to as the “Pearl on the Plateau”. Erhai Lake, which has a drainage area of 2565 km2, is part of the Lancangjiang and Mekonghe River system. It has many functions such as regulating the climate and providing water for industrial and agricultural production. Erhai Lake is also an important location within the whole river basin, and nature conservation and recreational area (Shen et al. 2005; Yu et al. 2014; Kitagawa et al. 2007; Kaiserli et al. 2002; Zhang et al. 2013; Sun et al. 1999).

But the lake is transforming from a mesotrophic state to a state of eutrophication, and is currently at an early stage of eutrophication (Tang et al. 2012). Its eutrophication greatly affects social and economic sustainable development of the whole basin. To improve the aquatic environment of Erhai Lake, the state has invested funds into the integrated management of the basin. At the same time, there is an increasing interest in using Erhai Lake as a model for other lakes.

Point-source control has been identified not efficacious measure to protect water bodies (Guo et al. 2004). The study of non-point-source (NPS) pollution has attracted great attention. The rivers, mainly accepting NPS pollution, around the lake bring main pollution load into Erhai Lake. However, NPS pollution from agriculture areas is a great challenge with the excessive use of chemical fertilizer and pesticide for scientists and the government (Gao et al. 2005; Chen et al. 2010; Shen et al. 2012).

The buffer zone of the lake refers to an area of protected land under lots of plant cover that is used to intercept pollutants or hazardous substances. It is effectively a biological treatment to control NPS (Natural Resources Conservation Service, USDA 1998). It plays a role in maintaining water quality and protecting water bodies (Thiagarajah et al. 2010). Current domestic and international research on buffer zones has concentrated mainly on the purification effect and ecological restoration in terms of water quality (Wang et al. 2012; Bernal et al. 2007), but research on variation of water quality in the buffer zone of lake is limited.

The buffer zone is close to the lake (including waterbody and lakeside), the pollutants into the lake major by the following four ways. The first way is the pollutants from dry and wet deposition directly down to the lake. The second way is the pollutants from underground water. The third way is the pollutants from rivers around the lake, the river mainly run from the buffer zone outside, then run through the buffer zone, and then enter the lake. The forth way is the pollutants from the overflow into the lake. Some information about the pollutants carried into Erhai Lake, the water quality situation in the whole basin and the variation time within 1 year was obtained by monitoring the water into Erhai Lake, the water quality situation in quality of the buffer zone of the rivers. So the buffer zone and the upper river is very important for the control of pollution in the whole Erhai Lake Basin.

So before harnessing the lake water, basic information such as the water quality in the rivers is essential for effective management. Determining the spatial distribution of water quality parameters in the rivers is also important.

The purpose of this paper is to evaluate the water quality and analysis of the distribution of water quality parameters for inflow rivers of Erhai Lake to provide baseline data for the characterization of potentially polluted water, and simulate the NPS water to select appropriate vegetation planed in the buffer zone.

Study area description

The distribution of selected water quality parameters in sections of 18 main rivers (the Mijuhe, Yonganjiang, Luoshijiang, Yangxi, Mangyongxi, Jinxi, Lingquanxi, Baishixi, Shuangyuanxi, Yinxianxi, Meixi, Taoxi, Zhonghexi, Baihexi, Longxi, Qingbixi, Mocanxi, and Boluojiang rivers) in the buffer zone around Erhai Lake was investigated. Mijuhe, Yonganjiang, and Luoshijiang rivers are three main rivers flowing into Erhai Lake from the north. The basin of Mijuhe River (1236.88 km2) is the largest sub-basin, accounting for 33.3 % of the average annual runoff, and the water of Haixihai Reservoir and Cibihu Reservoir also flows into Mijuhe River. Yonganjiang River is a tributary of Mijuhe River, collecting the runoff from mountains on the eastern side of the lake. With a length of 18.35 km, Yonganjiang River accounts for 7.6 % of the average annual runoff. Luoshijiang Basin, as one of the three major agricultural areas in Dali Bai Nationality Autonomous Prefecture, is the nearest basin to Erhai Lake, and is seriously populated. Luoshijiang River is a short river, with a basin area of 112.75 km2, and it collects water flowing from Xihu Lake and Lvyu Reservoir. The Water Environment Monitoring Center of Dali monitored 23 main streams in the dry and rainy seasons, and results had revealed that the rivers flowing into the north of the lake account for 47.0, 28.7, and 61.9 % of the total nitrogen, total phosphorus, and CODCr load flowing into Erhai Lake, respectively (Yan et al. 2005).

The Cangshan Eighteen-xi River flows into the west of the lake. The total area of the Cangshan Eighteen-xi River is 336.9 km2, accounting for 13.1 % of the drainage area of Erhai Lake, and the average annual runoff is 2.33 × 108 m3. It has an important impact on the ecological environment of Erhai Lake. Boluojiang River runs for 11 km from Fengyi Town in Dali City, and has a length of 17.5 km from its origin in Dingxiling village.

Materials and methods

Surface water from 18 rivers in the buffer zone area is sampled during the dry (from October to next May) season and rainy season (from June to October) in 2013 by a water sampler (Fig. 1). Surface water samples were stored in precleaned polyethylene bottles and transported to the laboratory in a small refrigerator (~4 °C) to minimize deterioration prior to chemical analysis. The following water quality parameters were measured: total nitrogen (TN), nitrate nitrogen (NO3 −-N), ammonia nitrogen (NH3-N), total phosphorus (TP) and the permanganate index (CODMn).

Fig. 1
figure 1

Location of sampling sites in rivers flowing into the Erhai Lake

Full size image

The following standard methods were used for analysis: ρ(TN) was determined in unfiltered samples by ultraviolet spectrophotometry method (GB11894-89), ρ(NH3-N) was determined in filtrates by Nash-reagent photometry method (GB7479-87), ρ(TP) was determined in unfiltered samples by antimony molybdenum method (GB11893-89), and ρ(NO3 −-N) and ρ(CODMn) were determined with reference to “Water and wastewater monitoring analysis method” (Yang et al. 2013). SPSS17.0 was used for the statistical data analysis.

Results and discussion

The spatial distribution of nitrogen

The change of ρ(TN) in water samples from the rivers in buffer zone area during the dry and rainy seasons is shown in Fig. 2. In the dry season, the range of ρ(TN) values was 0.21–5.85 mg/L, and the average value was 1.52 mg/L. While, in the rainy season, the range was 0.84–8.63 mg/L, and the average value was 2.12 mg/L. Compared with 29 rivers around Taihu Lake, and 16 channel openings around Danchi Lake, the ρ(TN) in the 18 inflow rivers of Erhai Lake was lower (Lu et al. 2011; Wang et al. 2009). This is because Erhai Lake is in a transition period toward eutrophication. In recent years, a series of engineering measures have been implemented for ecological protection, water conservation, sewage collection, and water treatment projects, while ecological rehabilitation projects have also been undertaken in the basin.

Fig. 2
figure 2

Variation of TN in the water of different rivers flowing into the lake. Class II Class II of “Surface Water Environmental Quality Standards of China”, Class IV Class IV of “Surface Water Environmental Quality Standards of China”

Full size image

The average ρ(TN) value of inflow rivers in the north of the lake was 2.06 and 2.16 mg/L during the dry and rainy seasons, respectively. The average value of ρ(TN) of inflow rivers in the west of the lake was 1.41 and 1.94 mg/L during the dry and rainy seasons, respectively. The average value of ρ(TN) of inflow rivers in the north of the lake was higher than that in the west (P = 0.341). The main reason is large number of agricultural non-point sources and high levels of aquaculture pollution in the northern part of the basin (Li et al. 2011). The northern area contains 58 % of the total farmland and 70 % of the total livestock in the whole basin according to “Yunnan Erhai Green Basin Construction and Water Pollution Prevention and Control Planning”.

In the rainy season, the order of ρ(TN) measured concentrations from high to low was as follows: Zhonghexi River > Jinxi River (P = 0.30) > Luoshijiang River (P = 0.88) > Mocanxi River (P = 0.73) > Baihexi River (P = 0.25) > Boluojiang River (P = 0.23) > Yangxi River (P = 1.00) > Meixi River (P = 0.88) > Taoxi River (P = 1.00) > Mangyongxi River (P = 0.98) > Baishixi River (P = 0.98) > Qingbixi River (P = 0.92) > Mijuhe River (P = 0.98) > Yonganjiang River (P = 0.96) > Lingquanxi River (P = 0.93)Longxi River (P = 0.98) > Yinxianxi River (P = 0.99) > Shuangyuanxi River(P = 0.89).

In the dry season, the order of ρ(TN) measured concentrations from high to low was as follows: Yinxianxi River > Yonganjiang River (P = 0.78) > Longxi River (P = 0.22) > Boluojiang River (P = 0.56) > Mocanxi River (P = 0.34) > Zhonghexi River (P = 0.79) > Jinxi River (P = 0.89) > Taoxi River (P = 0.96) > Mangyongxi River (P = 0.95) > Baihexi River (P = 0.63) > Shuangyuanxi River (P = 0.81) > Qingbixi River (P = 0.99) > Yangxi River (P = 0.97) > Luoshijiang River (P = 1.00) > Lingquanxi River (P = 0.87) > Meixi River (P = 0.99) > Baishixi River (P = 0.80) > Mijuhe River (P = 0.78).

The ρ(TN) in most of the rivers changes in large magnitude in rainy season and dry season and the ρ(TN) in the rainy season was clearly higher than that in the dry season (P = 0.419); the reason is that ρ(TN) was affected seriously by NPS pollution. Large amounts of nitrogen in the soil were transferred into the river by rain, which was in addition to the large number of NPS emissions from economic activities (Cheng et al. 2008).

The change of ρ(NO3 −-N) and ρ(NH3-N) in water samples from the rivers in the buffer zone during the dry and rainy seasons is shown in Figs. 3 and 4. In the dry season, the average value of ρ(NO3 −-N) and ρ(NH3-N) was 0.31 and 0.04 mg/L. In the rainy season, the average value of ρ(NO3 −-N) and ρ(NH3-N) was 0.61 and 0.70 mg/L, respectively.

Fig. 3
figure 3

Variation of NO3 −-N in the water of different rivers flowing into the lake

Full size image
Fig. 4
figure 4

Variation of NH3-N in the water of different rivers flowing into the lake. Class II Class II of “Surface Water Environmental Quality Standards of China”, Class IV Class IV of“Surface Water Environmental Quality Standards of China”

Full size image

In the dry season, the range of ρ(NO3 −-N) values was 0.00–3.48 mg/L, and the order of measured concentrations from high to low was as follows: Yonganjiang River > Luoshijiang River (P = 0.00) > Boluojiang River (P = 0.00) > Longxi River (P = 0.00) > Taoxi River (P = 0.24) > Mocanxi River (P = 0.93) Jinxi River (P = 0.98) > Mangyongxi River (P = 0.90) > Zhonghexi River (P = 0.99) > Lingquanxi River (P = 1.00) > Baihexi River (P = 1.00) > Qingbixi River (P = 0.98) > Yinxianxi River (P = 0.97) > Mijuhe River (P = 1.00) > Meixi River (P = 1.00) > Yangxi River (P = 0.98) > Shuangyuanxi River (P = 1.00) > Baishixi River (P = 0.99).

In the rainy season, the range of ρ(NO3 −-N) values was 0.46–1.16 mg/L, and the order of measured concentrations from high to low was as follows: Mocanxi River > Jinxi River (P = 0.05) > Mangyongxi River (P = 0.89) > Baihexi River (P = 0.66) > Qingbixi River (P = 0.75) > Longxi River (P = 0.84) > Zhonghexi River (P = 0.76) > Yonganjiang River (P = 0.62) > Mijuhe River (P = 0.93) > Yangxi River (P = 0.99) > Boluojiang River (P = 0.90) > Taoxi River (P = 0.94) > Luoshijiang River (P = 1.00) > Baishixi River (P = 0.93) > Meixi River (P = 0.98) > Yinxianxi River (P = 0.96) > Shuangyuanxi River (P = 0.92) > Lingquanxi River (P = 0.96).

In the dry season, the range of ρ(NH3-N) values was 0.001–0.28 mg/L, and the order of measured concentrations from high to low was as follows: Mijuhe River > Lingquanxi River (P = 0.99) > Jinxi River (P = 1.00) > Mangyongxi River (P = 1.00) > Qingbixi River (P = 1.00) > Baishixi River (P = 1.00) > Mocanxi River (P = 1.00) > Shuangyuanxi River (P = 1.00) > Meixi River (P = 1.00) > Longxi River (P = 1.00) > Baihexi River (P = 1.00) > Taoxi River (P = 0.97) > Yangxi River (P = 0.98) > Yinxianxi River (P = 0.94) > Zhonghexi River (P = 0.98) > Luoshijiang River (P = 0.00) > Yonganjiang River (P = 0.36) > Boluojiang River (P = 0.36). The water quality of the 18 main rivers in the buffer zone area was within the class II of “Surface Water Environmental Quality Standards of China”.

In the rainy season, the range of ρ(NH3-N) values was 0.08–6.47 mg/L, and the order of measured concentrations from high to low was as follows: Mangyongxi River > Shuangyuanxi River (P = 1.00) > Mocanxi River (P = 0.98) > Lingquanxi River (P = 1.00) > Longxi River (P = 0.98) > Baishixi River (P = 1.00) > Qingbixi River (P = 1.00) > Yinxianxi River (P = 0.99) > Mijuhe River (P = 0.99) > Yonganjiang River (P = 0.99) > Baihexi River (P = 0.10) > Taoxi River (P = 0.14) > Boluojiang River (P = 0.20) > Yangxi River (P = 0.99) > Luoshijiang River (P = 0.98) > Meixi River (P = 0.87) > Jinxi River (P = 0.93) > Zhonghexi River (P = 0.82). The water quality of all 18 rivers was within the Class IV of “Surface Water Environmental Quality Standards of China”, except for Zhonghexi River where the water quality exceeded the Class V of “Surface Water Environmental Quality Standards of China”.

Overall, the average value of ρ(NO3 −-N) and ρ(NH3-N) in water samples from inflow rivers in the buffer zone in the dry season was higher than that in the rainy season. The change of ρ(NO3 −-N) and ρ(NH3-N) in the different rivers between two seasons was considerable. Ammonia nitrogen mainly comes from breeding of livestock and poultry, waste of animals and decomposition of plants (Cai et al. 2013). Ammonia nitrogen was transformed into nitrate nitrogen by nitrobacterium, and then transferred into N2 releasing into the atmosphere finally.

In the dry season, ρ(NO3 −-N) and ρ(NH3-N) accounted for 20.4 and 2.60 % of ρ(TN), respectively. While, in the rainy season, ρ(NO3 −-N) and ρ(NH3-N) accounted for 28.8 and 33.0 % of ρ(TN), respectively. The proportion of ρ(NH3-N) in the total ρ(TN) was higher in the rainy season that was because the NO3 −-N adsorptivity of soil particle and colloid is lower than NH3-N adsorptivity, and NO3 −-N is easier to enter into the rivers leached by runoff (Ding et al. 2014; Tian et al. 2007).The ratio of ρ(NH3-N) to ρ(TN) was lower compared with the rivers flowing into Taihu Lake in the rainy season. Because the terrain of Taihu Lake was flat (Lu et al. 2011), there were many reciprocating rivers, which were not conducive to the conversion of NH3-N to NO3 −-N.

In the dry season, it was found that ρ(TN) and ρ(NO3 −-N) had a significant positive correlation (n = 18, R = 0.541, P < 0.05), but the correlation of ρ(TN) and ρ(NH3-N) was weak. In the rainy season, ρ(TN) and ρ(NH3-N) had a highly significant positive correlation (n = 18, R = 0.921, P < 0.01); the correlation of ρ(TN) and ρ(NO3 −-N) was weak.

The spatial distribution of phosphorus

Phosphorus is very important to phytoplankton and cyanobacteria, but can also be a factor contributing to the eutrophication of water. The concentration and chemical form of phosphorus in river water influence the pollution abatement policies adopted (Piotr and Agnieszka 2005; Lu et al. 2013; San Clements et al. 2009). The change of ρ(TP) in river water samples from the buffer zone area during the dry and rainy seasons is shown in Fig. 5. In the dry season, the range of ρ(TP) was 0.02–0.29 mg/L, and the average value was 0.08 mg/L, which was within the class II surface water standard. In the rainy season, the range was 0.04–1.05 mg/L, and the average value was 0.24 mg/L, which exceeded the class III surface water standard, but it was slightly lower than the value recorded in the 9 rivers flowing into Chaohu Lake (Wang et al. 2011). In the dry season, ρ(TP) and ρ(TN) were significantly positively correlated (n = 18, R = 0.854, P < 0.01), but the correlation was not obvious in the rainy season, which may be due to the complexity of NPS (Kaiserli et al. 2002).

Fig. 5
figure 5

Variation of TP in the water of different rivers flowing into the lake. Class II Class II of “Surface Water Environmental Quality Standards of China”, Class III Class III of“Surface Water Environmental Quality Standards of China”

Full size image

In the rainy season, the order of ρ(TP) measured concentrations from high to low was as follows: Zhonghexi River > Jinxi River (P = 0.58) > Yangxi River (P = 0.94) > Taoxi River (P = 0.49) > Meixi River (P = 0.83) > Shuangyuanxi River (P = 0.90) > Yinxianxi River (P = 0.98) > Luoshijiang River (P = 0.71) > Baihexi River (P = 0.38) > Mijuhe River (P = 0.22) > Baishixi River (P = 0.99) > Yonganjiang River (P = 0.95) > Mocanxi River (P = 0.92) > Qingbixi River (P = 0.49) > Mangyongxi River (P = 0.98) > Longxi River (P = 1.00) > Boluojiang River (P = 0.98) > Lingquanxi River (P = 0.96).

In the dry season, the order of ρ(TP) measured concentrations from high to low was as follows: Meixi River > Yinxianxi River (P = 0.76) > Zhonghexi River (P = 0.00) > Baishixi River (P = 0.99) > Luoshijiang River (P = 0.59) > Yonganjiang River (P = 0.72) > Boluojiang River (P = 0.62) > Longxi River (P = 0.74) > Taoxi River (P = 0.96) > Mocanxi River (P = 0.71) > Yangxi River (P = 0.93) > Lingquanxi River (P = 0.87) > Baihexi River (P = 0.97) > Qingbixi River (P = 0.85) > Shuangyuanxi River (P = 0.95) > Mijuhe River (P = 0.87) > Mangyongxi River (P = 0.98) > Jinxi River (P = 0.97).

The ρ(TN) and ρ(TP) in the Yonganjiang River and Boluojiang River were higher in the dry season than in the rainy season, which was due to point-source pollution (e.g., Yun Nong Feng Wine Industry and Die Quan Emulsion Co. Ltd located in the upstream of Yonganjiang River). Point-source pollution of the Yonganjiang River and Boluojiang River is relatively fixed.

The average values of ρ(TP) in the northern inflow rivers of the lake were 0.07 and 0.16 mg/L during the dry and rainy seasons, respectively. While the average values of ρ(TN) in the western inflow rivers of the lake were 0.09 and 0.25 mg/L, which was higher than the northern inflow rivers at the same time. That was because there were many villages in the west, where untreated sewage water and agricultural runoff entered the rivers.

The spatial distribution of the permanganate index (CODMn)

CODMn measures the amount of organic matter in water (Niemi 2012). The change of ρ(CODMn) in river water samples from the buffer zone and lakeside area in the dry and rainy seasons is shown in Fig. 6.

Fig. 6
figure 6

Variation of CODMn in the water of different rivers flowing into the lake. Class I Class I of “Surface Water Environmental Quality Standards of China”, Class II Class II of“Surface Water Environmental Quality Standards of China”

Full size image

In the dry season, the range of ρ(CODMn) was 1.65–12.96 mg/L, and the average was 5.58 mg/L. In the rainy season, the range was 3.56–13.05 mg/L, and the average was 8.51 mg/L, which was within the Class I of “Surface Water Environmental Quality Standards of China”. The average ρ(CODMn) in the dry season was higher than that in the rainy season, because a large amount of organic matter in the soil entered the rivers through runoff in the rainy season.

In the rainy season, the order of ρ(CODMn) measured concentrations from high to low was as follows: Luoshijiang River > Yangxi River (P = 0.94) > Meixi River (P = 0.93) > Yonganjiang River (P = 0.13) > Baihexi River (P = 0.53) > Boluojiang River (P = 0.48) > Mangyongxi River (P = 0.96) > Zhonghexi River (P = 0.33) > Yinxianxi River (P = 0.40) > Taoxi River (P = 0.82) > Baishixi River (P = 0.82) > Jinxi River (P = 0.91) > Shuangyuanxi River (P = 0.68) > Mijuhe River (P = 0.86) > Qingbixi River (P = 1.00) > Longxi River (P = 0.93) > Mocanxi River (P = 0.44) > Lingquanxi River (P = 0.44).

In the dry season, the order of ρ(CODMn) measured concentrations from high to low was as follows: Taoxi River (P = 0.76) > Yinxianxi River (P = 0.15) > Baishixi River (P = 0.68) > Zhonghexi River (P = 0.46) > Meixi River (P = 0.87) > Longxi River (P = 0.92) > Baihexi River (P = 0.82) > Shuangyuanxi River (P = 0.97) > Qingbixi River (P = 0.81) > Mijuhe River (P = 0.82) > Boluojiang River (P = 0.97) > Yangxi River (P = 0.99 > Mocanxi River (P = 0.83) > Yonganjiang River (P = 0.84) > Lingquanxi River (P = 0.89) > Mangyongxi River (P = 0.76) > Jinxi River (P = 0.94) > Luoshijiang River (P = 0.91).

The average value of ρ(CODMn) in the northern inflow rivers of the lake was 3.14 and 9.71 mg/L during the dry and rainy seasons, respectively. The average value of ρ(CODMn) in the rainy season was higher than that in the dry season. This is likely to be due to the large number of livestock in the “North Sanjiang” river basin, where the manure treatment rate is low, and waste accumulates on streets and the riverside.

Conclusion

  1. 1.

    The ρ(TN) in water samples from 18 main inflow rivers in the buffer zone displayed a seasonal variation, and was higher in the rainy season than in the dry season. The average value of ρ(TN) in the northern inflow rivers was higher than that in the western inflow rivers of the lake. Pollution control should, therefore, be strengthened in the north. The average value of ρ(NO3 −-N) and ρ(NH3-N) in river water from the buffer zone in the dry season was lower than that in the rainy season. In the dry season, ρ(NO3 −-N) accounted for high proportion of the ρ(TN), whereas in the rainy season the proportion of ρ(NH3-N) was higher.

  2. 2.

    The average value of ρ(TP) in water samples from the inflow rivers in the buffer zone during the rainy and dry seasons was 0.08 and 0.24 mg/L, respectively. In the dry season, ρ(TP) and ρ(TN) were significantly positively correlated, but there was no correlation in the rainy season. The average value of ρ(TP) in the northern inflow rivers was higher than that in the western inflow rivers of the lake.

  3. 3.

    The ρ(CODMn) in water samples from the inflow rivers in the buffer zone during the rainy and dry seasons was within the class I surface water standard. The average value of ρ(CODMn) in the northern inflow rivers in the rainy season was higher than that in the dry season.

  4. 4.

    A large variability in nutrient speciation for nitrogen and phosphorus was found among different sites during various seasons. Agricultural runoff, domestic sewage and livestock manure are the major factors influencing the speciation of the water quality.