Introduction

Organic matter, a complex entity with diverse functional groups and molecular sizes, has important effects on both pristine and engineered aquatic systems (Dryer et al, 2008; Simsek et al., 2012; Yu et al., 2013). It can affect species evolution, nutrient cycles, and pollutant transport in aquatic systems (Zbytniewski and Buszewski, 2005; Goldman et al., 2012). Treated/untreated wastewater into urbanized rivers deeply influences on organic matter components. Furthermore, urban changes in land use and climate continuously impact on organic matter cycling in waters (Evans et al., 2005; Osburn et al., 2012; Meng et al. 2014). The organic matter pools can be divided into particulate (POM) and dissolved (DOM) organic matter by filtration (Raymond and Bauer, 2001; Tremblay and Gagne, 2009; Osburn et al., 2012). POM pool is dominated by fresher materials, whereas DOM is composed of more recalcitrant compounds (Duan and Bianchi, 2007; Tremblay and Gagne, 2009). Their properties usually present temporal and spatial variations in a river-dominated system. A comprehensive study of POM and DOM components in rivers can conduce to reveal biogeochemical cycles of terrestrial and aquatic ecosystems within and between the organic carbon pools (Grace and Bianchi, 2010; Goldman et al., 2012; Osburn et al., 2012).

Excitation-emission matrix (EEM) fluorescence spectroscopy can provide advantageous information on organic matter, which can correctly expound its components and dynamics in different aqueous environments (Baker et al., 2007; Goldman et al., 2012). An EEM is a three-dimensional spectrum where fluorescence intensity is defined as a function of excitation and emission wavelengths (Valencia et al., 2014). Early using the peak-picking method, several peaks can be screened, such as peaks A (UV fulvic-like substance) and C (visible fulvic-like substance), T (tryptophan-like substance) and B (tyrosine-like substance), and M (microbial humic-like substance) (Hudson et al., 2008; Osburn et al., 2012; Korak et al., 2014). In a growing number of the organic matter studies, EEMs have been widely combined with parallel factor analysis (PARAFAC) to reveal component, source, and fate of organic matter. Spectral components are identified by the method, which can provide further insight into the environmental dynamics of the organic matter in diverse aquatic ecosystems.

The aims of this study were (i) to discriminate constituent components of DOM and POM in Baitapuhe River and Xihe River, two typical urbanized rivers located in Shenyang City of northeast China using the EEMs with PARAFAC; (ii) to compare the characteristics of components and spatial dynamics of the DOM and POM; and (iii) to reveal correlations of fluorescent components with water quality.

Material and methods

Study area and sampling strategy

The Hunhe River, a main branch of Liaohe River, is located in the center of Liaoning Province, China (Fig. 1a). The river extends 415 km with the basin of 11,400 km2, where a heavy industry complex has been developed. There are 31 tributary rivers whose basins are more than 100 km2 (Fig. 1b). Among these tributaries Baitapuhe River and Xihe River, two urbanized rivers, flow through Shenyang City (Fig. 1c). Baitapuhe River extends 51.5 km in the left bank of the Hunhe River, and its basin covers 182 km2. Its average runoff volume is approximately 69,200 m3 days−1 dominated by untreated/treated livestock wastewater and domestic sewage. Xihe River feeds into the Hunhe River from the right bank with 78.2 km, and its basin encompasses 365 km2. Its average runoff volume is about 924,500 m3 days−1 dominated by industrial wastewater, which have been treated mostly. The basins experience a temperate monsoon climate with an average annual rainfall of 680.4 mm. Annual temperature is approximately 8.1 °C, and minimum monthly average temperature (−11.6 °C) exists in January and maximum monthly average temperatures (24.6 °C) in July. The dominant wind directions are south-east in summer and north-west in winter. The annual average concentration of total phosphorus (TP) or ammonia(NH4-N) is 0.37 mg L−1 or 4.94 mg L−1 in Baitapuhe River and 3.27 times or 2.35 times less than that in Xihe River. This indicated that water quality is becoming deteriorated, particularly in Xihe River.

Fig. 1
figure 1

Study area and sampling sites. Liaohe River system (a), Hunhe River system (b), Baitapuhe River and Xihe River with the composition of river water (c). B Baitapuhe River, X Xihe River

Full size image

Water sampling was performed during August 14 to 15, 2013. Sampling sites in Baitapuhe River and Xihe River were set along a human impact gradient (Fig. 1c). The sites B1 to B4 were located in Baitapuhe River across the rural region and the sites B5 to B9 across the urban region (Yu et al., 2015a). Along Xihe River, the sites X1 to X4 were set in the centralized wastewater-discharge region, while the sites X5 to X8 in the dispersed wastewater-discharge region (Yu et al., 2015b). At a given sampling site, three samples were obtained from the left, middle, and right positions of the section, which were completely mixed with an equal volume. The three positions of the section were about 30 cm below the surface water, and the left and right positions were 2 m from the relevant shore. Water sample collection was carried out with a 5 L glass water sampler, and replicates were carried out at each sampling site. The samples were immediately stored and shipped to the lab in a cooled container. Fluorescence spectra are measured usually within a day after the sampling (Borisover et al., 2009).

Physico-chemical analyses

Temperature, pH, dissolved oxygen (DO), and electrical conductivity (EC) of each water sample were measured on-site with an YSI 600 multi-probe. In the laboratory, water samples filtered through 0.45 μm membrane were used for measurement of NH4-N and nitrate (NO3-N), while non-filtered samples were used for measurement of total nitrogen (TN), TP, chemical oxygen demand (COD), 5-day biochemical oxygen demand (BOD5), and Chlorophyll a (Chl a) (Specchiulli et al., 2008). These parameters were measured in accordance with national standards (SEPA, 2002).

An empirical relationship between the concentrations of DO and NH4-N is used as an indicator of the aerobic/anaerobic conditions among water samples, which is calculated as (Bleich et al., 2009; Stanhope et al., 2009; Vazquez et al., 2011):

$$ \mathrm{C}\mathrm{I}= \ln \left(\mathrm{DO}/{\mathrm{NH}}_4\hbox{-} \mathrm{N}\right) $$
(1)

The chemical index (CI) may vary with water environmental conditions. In the case of low DO concentration, it is expected to increase solutes in reduced form (NH4-N). Hence, CI will show low values, while high values in aerobic conditions (Vazquez et al., 2011). Moreover, another empirical ratio between BOD5 and COD (B/C) is utilized to indicate biodegradability of wastewater, and contaminants with B/C values higher than 0.3 may be considered easily biodegradable (Chamarro et al., 2001).

DOM-POM extraction and EEM spectroscopy measurement

POM was extracted from the water sample using a precombusted glass fiber filter (0.7 μm porosity, 25-mm diameter, Whatman GF/F), and the filtrate was used for DOM. Organic matter on the 0.7 μm filter defined as POM was extracted into 10 mL of 0.1 N NaOH for 24 h in the dark. The resultant solution was filtered through 0.2 μm polyethersulfone filters (Millipore Sterivex). After adjustment for pH 8.0 ± 0.05 near that of the original sample, the base extracts of POM were corrected with the corresponding solution volume of DOM. Particle (POC) and dissolved (DOC) organic carbon were measured by a TOC analyzer (analyticjena multi N/C 3100 TOC, Germany).

EEM spectra were recorded on a Hitachi Fluorescence Spectrophotometer (F-7000) equipped with the fluorescence solutions 1.00.000 for data processing. Excitation and emission slit widths were set to 5 nm bandpass, and PMT Voltage was fixed at 700 V. Scan speed was set at 2400 nm min−1. Excitation wavelengths were determined every 5 nm from 240 to 450 nm, and emission wavelengths were determined from 280 to 550 nm at intervals of 10 nm (Osburn et al., 2012). Appropriate instrument-specific corrections of the excitation and emission were performed, and all fluorescence spectra were corrected for inner-filtering effects before Quinine-sulfate calibration against Raman signal (Osburn et al., 2012). All spectra were subtracted from their respective procedural blanks.

PARAFAC model and multivariable analyses

Fluorescence EEMs were modeled by PARAFAC, which uses an alternating least squares algorithm to minimize the sum of squared residuals across the dataset and to estimate the underlying structure of the EEM (Harshman and Lundy, 1994; Bro, 1997). The calibrated and corrected EEM datasets were performed in Matlab software. The DOMFluor toolbox 1.7 of the Matlab software package was used for the implementation and validation of the PARAFAC model (Stedmon and Bro 2009; toolbox is freely available). The number of independent fluorescent components was determined by validation methods including split-half and residual analysis.

A non-metric multidimensional scaling (nMDS) analysis was applied to explore the physico-chemical (dis)similarities between Baitapuhe River and Xihe River. Principal component analysis (PCA) was conducted on fluorescent components and physico-chemical parameters across the rivers for finding latent factors and revealing relationships between fluorescent components and water quality. Statistical analyses were performed using SSPS 16.0 and OriginPro 8.0 for windows.

Results and discussion

Physico-chemical characteristics of two urbanized rivers

The variations of physico-chemical parameters were very noticeable in the rivers (Fig. 2). The mean temperature was higher (23.4 ± 0.5 °C) in Xihe River than that (20.8 ± 1.6 °C) in Baitapuhe River. The proportion of effluent from wastewater treatment plants (WWTPs) into Xihe River, whose temperature was about 25 °C in summer, was higher than that into Baitapuhe Rive. The mean pH was similar in the two rivers, but a little higher in Xihe River. The mean EC in Baitapuhe River was lower (472.1 ± 104.4 μS cm−1) than that (855.8 ± 209.8 μS cm−1) in Xihe River. The mean DO in Baitapuhe River was much higher (8.82 ± 1.72 mg L−1) than that (1.41 ± 0.36 mg L−1) in Xihe River, while the mean NH4-N (2.78 ± 1.44 mg L−1) in the former was much lower than that (17.25 ± 2.67 mg L−1) in the latter. Hence, the mean CI in Baitapuhe River was apparently higher (1.25 ± 0.68) than that (−2.53 ± 0.38) in Xihe River, proving that Baitapuhe River was under the aerobic condition, whereas Xihe River was under the anaerobic condition. The mean COD (16.86 ± 4.72 mg L−1) in Baitapuhe River was 2.57 times less than that (60.16 ± 15.73 mg L−1) in Xihe River, while the BOD5 (7.86 ± 3.29 mg L−1) in the former was similar to that (7.79 ± 4.32 mg L−1) in the latter. Consequently, the mean B/C ratio (0.46 ± 0.13) in Baitapuhe River was far higher than that (0.15 ± 0.11) in Xihe River, indicating that biodegradability of containments of the former was considerably higher than that of the latter. In Xihe River, the mean B/C was much lower in the centralized wastewater-discharge region (0.06 ± 0.01) than that in the dispersed wastewater-discharge region (0.26 ± 0.06). The mean TN was lower in Baitapuhe River (4.65 ± 2.04 mg L−1) than that in Xihe River (22.23 ± 2.75 mg L−1), so were the NO3-N, TP, and Chl a. This revealed that the water quality was better in Baitapuhe River than that in Xihe River.

Fig. 2
figure 2

Box plots of temperature (a), pH (b) EC (c), DO (d), COD (e), BOD5 (f), NH4-N (g), NO3-N (h), TN (i), TP (j), Chl a (k), CI (l), and B/C (m) of water samples. Black circles indicate the mean values, boxes show standard error, and whiskers delimit maximum and minimum. BR Baitapuhe River, XR Xihe River

Full size image

The nMDS analysis allowed comparing the variability of physico-chemical parameters among sampling sites in Baitapuhe River and Xihe River (Fig. 3). The graphical representation showed that the sampling sites were clustered into two groups with 85 % confidence level for each dashed ellipse. The confidence level is lower, whose confidence interval is narrower, but the evaluation accuracy is higher. This indirectly verified that the water quality of Xihe River was worse than that of Baitapuhe River. Notice that the major and minor axes of the dashed ellipse for Baitapuhe River have larger values than those for Xihe River. This explained that the variations of physico-chemical parameters in Baitapuhe River were larger than those in Xihe River (Fig. 2).

Fig. 3
figure 3

nMDS map of the distribution of sampling sites in both rivers (black: Baitapuhe River; white: Xihe River) according to its water physico-chemical characteristics. B Baitapuhe Rive, X Xihe River

Full size image

Two expected groups were presented in the sampling sites of Baitapuhe River with 55 % confidence level for each solid-line ellipse, which corresponded to the rural (sites B1 to B4) and urban (sites B5 to B9) regions. This indirectly illustrated that water quality in the rural region was better than that in the urban region. The sampling sites of Xihe River also exhibited two expected groups with 40 % confidence level for the solid-line ellipses, indirectly explaining that a large amount of treated/untreated wastewater entered into the river along the sites X1 to X4, whereas a small amount of wastewater into the river along the sites X5 to X8.

Content and source differences of DOM-POM between the two rivers

The variations of DOM concentrations were very obvious in the rivers (Fig. 4). The mean DOC (10.51 ± 2.69 mg L−1) in Baitapuhe River was 1.14 times less than that (22.51 ± 6.99 mg L−1) in Xihe River. The DOC in Baitapuhe River showed good positive correlations with the NH4-N (r = 0.78, p < 0.05) and the TP (r = 0.76, p < 0.05), and the B/C values were more than 0.3. This indicated that the DOM increased with a rise in the NH4-N or the TP, and it had the strong biodegradability. Since Baitapuhe River water was composed of treated/untreated livestock wastewater and domestic sewage (Fig. 1c), whose typical compounds are the NH4-N and TP (Goldman et al., 2012), the most DOM probably derived from the livestock wastewater and domestic sewage. The DOC in Xihe River showed insignificantly correlations with the NH4-N (r = 0.57, p > 0.05) and TP (r = 0.31, p > 0.05), and the B/C values were far less than 0.3. For Xihe River, water was dominated by treated industrial wastewater with the weak biodegradability (Fig. 1c); the majority of DOM could originate from treated industrial wastewater.

Fig. 4
figure 4

Distributions of the concentration of the DOC (a) and POC (b) in Baitapuhe River and Xihe River

Full size image

In Baitapuhe River, the DOC concentrations at the sites B1 to B4 were lower than those at the sites B5 to B9. This might attribute to much treated/untreated sewage discharge into the river across the urban region, besides accumulation from the rural region. Noticeably, the effluent from the Baita WWTP with 20,000 m3 days−1 entered into the river at the site B6, which could result in the highest total the DOC concentration among the nine sampling sites. In Xihe River, the DOM gradually increased along river water flow in the centralized wastewater-discharge region, while it decreased in the dispersed wastewater-discharge region. The effluent from the Beibu WWTP with 360,000 m3 days−1 runoff at the site X1 was considered as the river head. The effluent from the No. 14 Gateway with 60,000 m3 days−1 runoff flowed into Xihe River at the site X2, which primarily consisted of the domestic sewage. The effluent from the Xiannühe WWTP with 360,000 m3 days−1 runoff discharged into the river at the site X3, and effluent from the Xibu WWTP with 120,000 m3 days−1 runoff entered into the river at the site X4. The influent of the WWTPs was dominated by the industrial wastewater, and containments had lower level biodegradation, for their B/C values were less than 0.07.

The POM concentrations in Baitapuhe River showed slightly lower (except the site B1) than those in Xihe River, but the variations of the former were higher than the latter. In Baitapuhe River, the mean POC was 4.98 ± 0.62 mg L−1, about 1.11 times less than the mean DOC. In Xihe River, the mean POC was 6.51 ± 0.45 mg L−1, about 2.46 times less than the mean DOC.

EEM characteristics of DOM-POM from two urbanized rivers

EEM spectra can exactly describe organic matter components, characterized by excitation/emission (Ex/Em) wavelength pairs and the specific fluorescence intensity. Fiver fluorescence peaks distinctly exhibited in the EEM spectra of the DOM in the rivers (Figs. 5 and 6). Based on previous studies (Coble, 1996), peak B (λ Ex/Em = 260–280/290–320 nm) and peak T (λ Ex/Em = 275–285/330–370 nm) are assigned to tyrosine-like and tryptophan-like substances, which are regarded as anthropogenic contaminants in wastewater-dominated rivers (Meng et al. 2014). Peak A (λ Ex/Em = 300–350/400–450 nm) or peak C (λ Ex/Em = 300–350/400–450 nm) is assigned to UV fulvic-like or visible fulvic-like. Peak M situated between the peaks A and C is assigned to microbial humic substance, which potentially indicates microbial processing.

Fig. 5
figure 5

The EEM spectra of the DOM and POM in Baitapuhe River

Full size image
Fig. 6
figure 6

The EEM spectra of the DOM and POM in Xihe River

Full size image

The distinct peaks B and T existed in the whole EEM spectra of the POM from the two rivers, whose variations of the intensities were relatively subtle. Moreover, the peaks A and M occurred in the EEM spectra of the POM at the sites X5 to X8, i.e., in the dispersed wastewater-discharge region.

PARAFAC modeling EEM datasets of the DOM and POM produced four different spectral components using the DOMFlour toolbox of the Matlab software package (Fig. 7). The broad and long wavelength peak of component 1 (C1) has been presented in a terrestrial component from a variety of aquatic environments, and it is similar to fulvic-like component. Component 2 (C2) occurs universally in agriculturally and industrially/urban impacted rivers, eutrophic lakes, and wastewaters, which resembles the previously identified peak M (Coble, 2007; Seredyńska-Sobecka et al., 2011; Osburn et al., 2012). Component 3 (C3) and component 4 (C4) are the protein-like peaks, which are relative to tryptophan-like and tyrosine-like components.

Fig. 7
figure 7

EEM of PARAFAC components extracted from the DOM and POM in Baitapuhe River and Xihe River

Full size image

Comparison of PARAFAC components of DOM-POM

Since the maximum fluorescence intensity (Fmax) is derived for each component and has the same units (Raman units), it can be referred as the proportional relative content of the homologous component (Stedmon and Bro, 2009; Sharpless and Blough, 2014). The average Fmax C1 of the DOM (C1d) in Baitapuhe River (356.1 ± 65.3) was approximately equal to that in Xihe River (363.1 ± 173.3), while the variations in the latter were the larger than those in the former (Fig. 8a). The trend of the C3d was approximate to the C1d. The mean C2d or C4d in Xihe River was 1097.1 ± 801.2 or 542.8 ± 426.9 and 4.29 times or 2.80 times more than that in Baitapuhe River. Interestingly, in Baitapuhe River, the continuum C1d values in the rural region were lower than those in the urban region, so were the C2d and C3d. The trend of the C4d was opposite to the C1d, C2d, and C3d. In Xihe River, the C2d values in the centralized wastewater-discharge region were lower than those in the dispersed wastewater-discharge region. However, the trends of the C1d, C3d, and C4d were opposite to the C2d.

Fig. 8
figure 8

Distributions of the abundance of the DOM (a) and POM (b) components in Baitapuhe River and Xihe River

Full size image

The average C1 of the POM (C1p) in Baitapuhe River (9.6 ± 4.6) was lower than that in Xihe River (27.3 ± 15.7), so were the C2p and C3p (Fig. 8b). The average C4p in Baitapuhe River was close to that in Xihe River. Strangely, in Baitapuhe River, the C2p was less present; this could be attributed to microbial activity limited by fresher materials (de Beer et al., 2006). In Xihe River, each component of POM in the centralized wastewater-discharge region, especially the C2p, was lower than that in the dispersed wastewater-discharge region.

The amounts of each component, measured as % of the total fluorescence maxima for the four components (%Fmax), varied substantially among the DOM and POM fractions (Fig. 9). Prominently, in Baitapuhe River, the %Fmax for each DOM component stayed relatively consistent, except for the C4d, which ranged between 2.1 and 45.3 %. However, in Xihe River, each component was higher in the centralized wastewater-discharge region than that in the dispersed wastewater-discharge region, except for the C3d, whose trend was reverse. The %Fmax C3d was highest in Baitapuhe River, indicating that tryptophan-like component was the representative component. The C3d ranged from 28.5 to 58.0 % at the sites X1 to X4, and the C4d altered from 13.9 to 39.7 %, proving that protein-like fluorescence comprised the dominant component in Xihe River across the centralized wastewater-discharge region. The C2d stayed relatively consistent (77.6–85.4 %) at the sites X5 to X8, explaining that microbial humic fluorescence was the indicative component in the dispersed wastewater-discharge region.

Fig. 9
figure 9

Distributions of percentages of the DOM (a) and POM (b) components in Baitapuhe River and Xihe River

Full size image

The %Fmax for the C1p was relatively small fraction in Baitapuhe River (0.5–3.1 %) and Xihe River (1.4–4.6 %). The C2p was close to zero in Baitapuhe River but widely ranged from 0.02 to 17.8 % in Xihe River. The C3p remained relatively stable in Baitapuhe River (6.5–15.1 %) and Xihe River (9.4–20.9 %). The C4p changed between 84.4 and 92.3 % in Baitapuhe River and between 66.4 and 88.3 % in Xihe River, indicating that the tyrosine-like component was the major fraction in the POM.

Correlations between PARAFAC components and water quality

Deterioration in water quality has been a lingering problem in the urbanized river, in which excessive anthropogenic-derived contaminants are loaded with rapidly increasing urbanization (Ouyang et al., 2006; Osburn et al., 2012). PCA was conducted on the fluorescent components and water quality data to investigate differences among Baitapuhe River and Xihe River, to seek latent factors, and to reveal correlations between fluorescence components and water quality. PCA of the two data sets produced two PCs for the two rivers, elucidating 91 and 97 % of total variance in the respective data sets. Figure 10 summarized the PCA including the loadings and scores plots. The loading whose absolute value was more than 0.6 of the total variance was highlighted.

Fig. 10
figure 10

Loadings plot and scores plot for PC1 and PC2 of the DOM-POM and water quality in Baitapuhe River (a) and Xihe River (b)

Full size image

In Baitapuhe River, PC1, accounting 79.5 % of the variance, had strong positive loadings on the C3d, C2p, C2d, EC, Chl a, NO3-N, and C1d, while strong negative loading on the C4d (Fig. 10a). The PC1 demonstrated that DOM is a latent factor in water quality, especially the C3d (a representative component in DOM). The C4d showed negative correlations with the C2d (r = −0.68, p < 0.05) and C3d (r = −0.65, p < 0.05), possibly indicating that tyrosine-like substance was biodegraded and partially transformed into tryptophan-like substance. The site B6 had a higher positive score on the PC1, indicating where relatively higher abundance of the C2d or C3d appeared, while the lower C4d compared to the other sites. The site B1 had a lower negative score on the PC1, elaborating where relatively lower C2d and C3d occurred. PC2 (11.5 % of the variance) showed strong positive loading on the C4p, C3p, and NH4-N, whereas strong negative loading on the DO. The C3p (with the loading 0.93) was less appearance, but its variance was high between samples. The NH4-N concentrations showed a close negative correlation with the DO (r = −0.77, p < 0.05), indicating that the NH4-N decreased with a rise in the DO. The site B9 had a higher positive score on the PC2, revealing where relatively higher C4p and NO3-N emerged, while lower DO.

An expected clustering existed between the sites B1-B4 and the sites B5-B9 with 65 % confidence level for each ellipse in the score plot. This indirectly revealed that the PARAFAC components of the DOM and POM and water quality in the urbanized river across the rural region were different from the urban region. Additionally, the major and minor axes of the confidence ellipse for the urban region samples have larger values than those for the rural region samples. This proved that the variations of the organic matter and water quality in the urban were larger than those in the rural region.

In Xihe River, PC1, describing 66.7 % of the variance, had strong positive associations with the DO, C1p, C2d, C2p, and BOD5, while strong negative associations with the C3d and COD (Fig. 10b). The C3d exhibited a significant positive correlation with COD (r = 0.73, p < 0.05), indicating that C3d increased with a rise in COD. This suggested that C3d should be used as an indicator of COD. COD showed a good negative correlation with BOD5 (r = -0.73, p < 0.05), proving that COD increased with the decreasing of BOD5. This indirectly proposed that organic matter had a low level of biodegradability, which could derive from the industrial wastewater. PC2 (30.3 % of the variance) exhibited negative associations with the C1d and C4d, while a strong positive association with the C3p. The site X4 showed a lower negative score on the PC1, while a higher positive score on the PC2, evidencing where relatively higher C3d and C3p, while lower DO and BOD5.

The sites were clustered into two groups with 40 % confidence level for each confidence ellipse in the scores plot, as expected. This indirectly proved that the B/C and C2d values in the centralized wastewater-discharge region were higher than those in the dispersed wastewater-discharge region, while the C1d, C3d, and C4d values were generally lower in the former than those in the latter.

Conclusions

EEM spectra combined with PARAFAC and multivariable analysis can provide an excellent challenge to compare components of the DOM and POM and to reveal their correlations with water quality in the two urbanized rivers. The DOM content in Baitapuhe River was lower than that in Xihe River, so was the POM. The DOM mainly derived from the livestock wastewater and domestic sewage in Baitapuhe River, and it had strong biodegradability. The majority of the DOM originated from the industrial wastewater in Xihe River, and it had weak biodegradability. The DOM and POM consisted of tyrosine-like, tryptophan-like, microbial humic, and fulvic-like components. Tryptophan-like component was the dominant contributor of the DOM in Baitapuhe River. Protein-like component was the representative fraction of the DOM in Xihe River across the centralized wastewater-discharge region, while microbial humic component was the major component across the dispersed wastewater-discharge region. Tyrosine-like component was the main component of the POM in the two rivers. The DOM, DO, Chl a, NO3-N, and NH4-N were the latent factors in Baitapuhe River, while DOM, DO, BOD5, and COD in Xihe River. Baitapuhe River was under the aerobic condition, whereas Xihe River under the anaerobic condition. These will encourage the study on restoration of urbanized rivers, such as treatment of breeding waster and domestic fecal sewage in Baitapuhe River and stricter effluent standards for the urban WWTPs in Xihe River.