Dissolved organic matter and metabolic dynamics in dryland lowland rivers

SAA-117871; No of Pages 10 Spectrochimica Acta Part A: Molecular and Biomolecular Spectroscopy xxx (xxxx) xxx

Contents lists available at ScienceDirect

Spectrochimica Acta Part A: Molecular and Biomolecular Spectroscopy journal homepage: www.elsevier.com/locate/saa

Dissolved organic matter and metabolic dynamics in dryland lowland rivers Aleicia Holland a,⁎, Paul J. McInerney c, Michael E. Shackleton b, Gavin N. Rees c, Nick R. Bond b, Ewen Silvester a a b c

La Trobe University, School of Life Science, Department of Ecology, Environment and Evolution, Centre for Freshwater Ecosystems, Albury/Wodonga Campus, Vic 3690, Australia La Trobe University, Centre for Freshwater Ecosystems, Albury/Wodonga Campus, Vic 3690, Australia CSIRO Land and Water, Thurgoona, NSW 2640, Australia

a r t i c l e

i n f o

Article history: Received 30 July 2019 Received in revised form 13 November 2019 Accepted 26 November 2019 Available online xxxx Keywords: Fluorescence excitation emission scan FEEM Dissolved organic carbon Humic Protein Productivity

a b s t r a c t Dissolved organic matter (DOM) within freshwaters is essential for broad ecosystem function. The concentration and type of DOM within rivers depends on the relative contributions of allochthonous sources and the production and consumption of DOM by microbes. In this work we have examined the temporal patterns in DOM quality and productivity in three lowland rivers in dryland Australia using fluorescence excitation emission scans. We assessed the production and consumption of DOM within light and dark bottle assays to quantify the relative contribution of bacteria and algae to the DOM pool and simultaneously assessed whether the systems were autotrophic or heterotrophic. DOM varied temporally within the three river systems over the course of the study period. Characterisation of DOM within light and dark bottles following a 6-hour incubation revealed microbial consumption of a humic-like component and production of protein-like components similar in nature to the amino acids tryptophan and tyrosine. The lack of a significant difference in DOM quality between the light and dark bottles indicated that the protein-like DOM is likely derived from bacterial activity. Respiration was shown to be higher than gross primary production in both whole river and bottle assays, yielding negative net production values and demonstrating that these rivers were predominately heterotrophic. Our work suggests that bacterial metabolism of DOM may be a significant contributor to the production of protein-like components within heterotrophic freshwater systems. © 2019 Elsevier B.V. All rights reserved.

1. Introduction Stream metabolism, measured as rates of aquatic primary production and whole ecosystem respiration, are functional indicators of aquatic ecosystem condition [1,2]. Gross primary production (GPP) and community respiration (CR) rates are controlled by exchanges between hydrology, vegetation (terrestrial riparian and aquatic), geomorphology, climate, water chemistry and biota of the stream environment, along with overall catchment condition [2,3]. Net production (GPP-CR) of aquatic systems depends on the amount of oxygen produced via primary production and the amount of carbon dioxide respired via respiration [4]. Negative net production is indicative of a heterotrophic ecosystem dominated by bacteria and fungi, whereas autotrophic systems (positive net production) are generally dominated by algae [5,6]. Many aquatic ecosystems world-wide have been found to be heterotrophic [5–7], although in some Australian lowland rivers annual net ecosystem production has been shown to be low, suggesting that

⁎ Corresponding author at: La Trobe University, School of Life Science, Department of Ecology, Environment and Evolution, Albury/Wodonga Campus, Vic, Australia. E-mail address: [email protected] (A. Holland).

photosynthesis and respiration are closely balanced and that allochthonous organic carbon plays a small role in fuelling metabolic activity [8]. Dissolved organic matter (DOM) is essential for a range of aquatic ecosystem functions [9]. DOM is an important food source to heterotrophs within river systems [10,11] and its presence may influence metabolic dynamics within rivers. DOM is a complex heterogenous mixture of humic substances, proteins, amino sugars, nucleic acids, and carbohydrates [12]. The concentration and type of DOM present within an aquatic environment depends on the contributions of allochthonous and autochthonous inputs within the system. Within fresh waters, increased connectivity with adjacent terrestrial environments can generate more complex DOM profiles, with higher relative quantities of humic substances, greater aromaticity and molecules of a higher molecular weight [9,13,14]. In contrast, aquatic ecosystems that are reliant on DOM produced from autochthonous activity (i.e. in-stream algal and microbial processes) generally contain DOM of a lower molecular weight, that is more aliphatic and proteinaceous in nature [9,13,14]. DOM composition differs spatially and temporally within and between freshwater systems [9,12]. These differences may be attributed to spatial and temporal variations in microbial and algal communities, rainfall, land use and urbanisation [15–18]. A common method used to determine changes in DOM quality is fluorescence and absorbance

https://doi.org/10.1016/j.saa.2019.117871 1386-1425/© 2019 Elsevier B.V. All rights reserved.

Please cite this article as: A. Holland, P.J. McInerney, M.E. Shackleton, et al., Dissolved organic matter and metabolic dynamics in dryland lowland rivers, Spectrochimica Acta Part A: Molecular and Biomolecular Spectroscopy, https://doi.org/10.1016/j.saa.2019.117871

2

A. Holland et al. / Spectrochimica Acta Part A: Molecular and Biomolecular Spectroscopy xxx (xxxx) xxx

spectroscopy [14,19,20]. These techniques provide information regarding relative abundance of key components and can give an indication of relative molecular weight, aromaticity and potential sources of DOM in freshwater systems [14,19,21,22]. Changes in abundance of certain DOM components e.g. protein-like components therefore may give an indication of spatial and temporal changes in autochthonous production within rivers, streams and lakes [12,13,20]. This study examined temporal changes in DOM quality (abundance of key components, relative aromaticity and molecular weight and source) and stream metabolism as indicated by gross primary production, respiration and net production rates within three Australian lowland rivers within the Edward-Wakool river system. We assess rates of production and consumption of DOM (and oxygen) within light and dark bottles to quantify the relative contribution of bacteria and algae to the DOM pool and provide information regarding the bioavailability of DOM components.

low (summer) rainfall and likely changes in flow and water temperatures during this period. 2.2. Water quality and nutrient analysis Temperature, conductivity, pH, dissolved oxygen (DO), and turbidity was measured at each location during each sampling trip with a YSI multiprobe (YSI incorporated, Yellow Springs, USA). Three replicate measurements were taken each sampling time and each site sampled 10 times (n = 30). Water-quality samples were collected between 12 pm and 3 pm each sampling trip to minimize temporal confounding of spatial patterns due to different collection times. Water samples (n = 3) were also collected from each location each sampling trip to measure Ammonium (NH4), oxides of nitrogen (NOx), total nitrogen (TN), total phosphorus (TP), filterable reactive phosphorus (FRP), dissolved organic carbon (DOC), and chlorophyll-a (Chl a) (see [23] for full details on methods for analysis of water samples).

2. Materials and methods

2.3. Production

2.1. Sites

Two approaches were used to estimate production; (1) single station open water methods were used to estimate whole-stream production and (2) light and dark bottles were used to estimate pelagic production. For single station whole-stream estimates, D-Opto dissolved oxygen loggers (Zebra-Tech LTD, Nelson, New Zealand) were attached to biofilm platforms at each site, and set to log dissolved oxygen and temperature every 10 min for the project duration. Odyssey® Waterproof Photosynthetic Active Radiation Loggers (Odyssey, Christchurch, New Zealand) measuring photosynthetically active radiation (PAR), were deployed on the bank adjacent to floating platforms, and logged light for the project duration. Gross Primary Production (GPP), Community Respiration (CR) and Net Production (NPP) were calculated using the BAyesian Single-station Estimation method [1] using the R package BASEMETAB [24]. To estimate pelagic production at each site, we constructed floating rigs to support three replicate light and dark bottles, which sat 18 cm

Sites were selected at locations on three rivers within the EdwardWakool River system in the southern Murray-Darling Basin (MDB), Australia; the Wakool River (−35.522576, 144.519656), Yallakool Creek (−35.506229, 144.484940) and Neimur River (−35.267449, 144.237214) (Fig. 1). All three rivers are typical of lowland systems in the MDB, consisting of channels that are relatively incised with clay banks, riparian vegetation was dominated by E. camaldulensis (River Redgum) and A. stenaphylla (River Cooba) at the margins, with E. largiflorens (Black Box) growing on higher elevations. Understories were dominated by introduced grasses and native chenopods, with a range of submerged and emergent macrophytes growing in slack water areas. The study was conducted during the Austral SpringSummer period, between October 2018 and January 2019. The study period was chosen to correspond with periods of likely high (spring) and

Fig. 1. Map of sampling sites: Neimur and Wakool Rivers and Yallakool Creek.

Please cite this article as: A. Holland, P.J. McInerney, M.E. Shackleton, et al., Dissolved organic matter and metabolic dynamics in dryland lowland rivers, Spectrochimica Acta Part A: Molecular and Biomolecular Spectroscopy, https://doi.org/10.1016/j.saa.2019.117871

A. Holland et al. / Spectrochimica Acta Part A: Molecular and Biomolecular Spectroscopy xxx (xxxx) xxx

3

3. Results

below the surface. Dissolved oxygen and temperature within bottles were measured prior to incubation, which occurred over 6 daylight hours, and were then measured again on retrieval. Light and dark bottles were deployed on nine occasions throughout the study. Water samples were collected prior to incubation, which occurred for 6 h during daylight hours, and then again from light and dark bottles on retrieval for characterisation of the dissolved organic matter. Preliminary work showed that the 6 h incubation was enough to alter the DOM characteristics in the water from the original river water. Whole stream metabolism was estimated using the R package BASEmetab [1] and retrieving instantaneous rates. Two hundred thousand MCMC iterations were performed on the data with the first 50,000 discarded as a burnin. For each sample day, the instantaneous rates occurring between the times light/ dark bottles were set to when they were retrieved and were summed to calculate the total production of oxygen within the river over the sample period. Production per hour was calculated by dividing the total production by the number of hours.

3.1. Water quality Ranges of water quality parameters are provided in Table 1. Water quality was similar among the three rivers. Temperature followed a seasonal trend among all sites, with mean values remaining between 18 and 22 °C from October to December before increasing in summer to peak at 27.9 °C in the Neimur River in January (Table 1 and Supplementary Fig. 1). Conductivity was similar in the Niemur River and Yallakool Creek, ranging from 45 to 55 μS/cm (Table 1 and Supplementary Fig. 1). Mean conductivity in the Wakool was slightly higher, ranging from 54 μS/cm to 83 μS/cm. pH ranged from 6.47 in the Wakool River to 7.32. Turbidity was generally highest within the Wakool River, and ranged between 23 and 80 NTU (Table 1 and Supplementary Fig. 1). Dissolved oxygen (% saturation), varied during the study period and ranged between 39 and 92% saturation (Table 1 and Supplementary Fig. 1). Dissolved organic carbon did not vary greatly during the study period and ranged from 1.9 to 3.2 mg/L (Table 1). Nutrients were similar between rivers (Table 1 and Supplementary Fig. 2). Chl-a was also relatively constant, although concentration at all sites dipped below 10 μg/L between 15 November and 6 December (Table 2 and Supplementary Fig. 3).

2.4. Dissolved organic matter characterisation Fluorescence excitation emission (FEEM) scans along with absorbance scans were performed using a quartz cuvette in an HORIBA Aqualog® fluorimeter (HORIBA Scientific) to determine quality of DOM and its key components within river water and light dark bottles. FEEM scans along with simultaneous absorbance measurements were conducted on all samples, with excitation wavelengths in 2-nm steps between 250 and 450 nm, and emission wavelengths of 250–620 nm. The absorbance of a blank of ultrapure water was automatically subtracted from each sample. Inner filter effect along with the 1st and 2nd order Rayleigh and Raman scatter were also removed using the HORIBA software. The excitation emission matrices were then analysed in MATLAB R2014b (MathWorks, Inc. © 1994–2016) to produce threedimensional FEEMs and contour plots. A total of 231 FEEMs were modelled using parallel factor analysis (PARAFAC) (PLS-toolbox in MATLAB: Eigenvectors Research Inc., WA, USA). The PARAFAC model was validated following the recommendations in Murphy et al. [25]. No clear consistent patterns and peaks were visible in the residual plots, core consistency was 100%, and split-half analysis results were consistent with the model. Relative DOM aromaticity, molecular weight, and source was determined using a range of absorbance and fluorescence indices such as the Specific UV absorbance at 254 nm (SUVA254: [22]), Biological Index (BIX: [53]) and the Fluorescence Index (FI: [13]). Further details provided in Supplementary Table 1.

3.2. Dissolved organic matter Fluorescence excitation emission scans of dissolved organic matter in Neimur, Wakool Rivers and Yallakool Creek water followed by PARAFAC analysis revealed the presence of three main DOM components: protein-like1 characterised by excitation of 240 and 290, emission 340; humic-like characterised by an excitation of 255 and 340, emission 444; and protein-like2 characterised by excitation 280 and emission 320 (Fig. 2). These components explained 98.7% of the variation and showed N95% similarity to 25 models within the Openfluor database (https://openfluor.lablicate.com) (Supplementary Table 1). Temporal and spatial variation in relative abundance of DOM components occurred between rivers and across sampling times (P b .001) (Fig. 6), with a significant interaction also shown between river and time (P b .001) (Fig. 3). All sites were dominated by the humic-like component except during the sampling periods of 23 November and the 4th of December in Wakool and Neimur rivers, respectively, where the protein-like1 component became the most dominant (Fig. 3). Similarly, the protein-like 2 component increased during the 15 November sampling period in the Wakool river and during the 4th December sampling period in Yallakool Creek (Fig. 3). Significant differences in relative abundance of all three components was found between fresh river water and river water incubated within light and dark bottles for 6 h (P b .001), with no significant differences between the light and dark treatments (P = .146) (Fig. 4). Over the incubation period, humic-like DOM was found to decrease while both protein-like components increased, in both the light and dark treatments (Fig. 4), however, temporal differences were shown in the abundance of the two protein-like DOM produced within the bottles. The protein-like 2 component dominated production during the first three sampling periods in November 2018 and the last sampling period in January 2019, whereas the protein-like 1 component was produced in higher amounts during December 2018 (Fig. 4).

2.5. Statistical analysis All graphs were created in R using ggplot2 [26]. Two-way ANOVAs comparing the relative abundance of each DOM component between rivers and between river water samples and light/dark bottles were conducted. Post-hoc Tukey tests were performed on pair-wise comparisons. Pearson correlations were performed to determine any significant relationships between DOM components and indices and water quality parameters and chlorophyll-a in water column and biofilms. Productivity (GPP, respiration and NPP) relationships to water quality parameters were also assessed using Pearson correlations in R using the Hmisc package. Table 1 Minimum and maximum water quality parameters for each river over the sampling period.

Neimur Wakool Yallakool

Temp °C

Cond μS/cm

DO % sat

pH

Turb NTU

DOC mg/L

Chl a μg/L

TN μg/L

NH4 μg/L

NOx μg/L

TP μg/L

FRP μg/L

19.6–27.9 18.6–26.4 18.7–26.8

46–54 54–83 45–55

66.4–79.8 38.5–77.1 72.6–91.5

6.59–7.31 6.47–6.86 6.57–7.13

23.1–55.8 57.8–80.1 24.9–72.6

2.3–3.0 1.9–3.1 2.1–3.2

6.7–14 8.5–19 5.9–12

340–433 452–507 362–396

5.3–8.0 4.2–6.7 4.5–6.3

1.7–2.3 2.0–2.7 1.7–2.7

50–63 57–66 49–62

5.3–7.6 2.5–14 2.5–6.3

Three replicate measurements were taken each sampling time and each site was sampled 10 times (n = 30). *DO = Dissolved oxygen; Cond = Electrical conductivity; Turb = Turbidity; DOC = Dissolved organic carbon; Chl a = Chlorophyll-a; TN = Total Nitrogen; NH4 = ammonium; NOx = Nitrate + nitrite; TP = Total Phosphorous; FRP = Filterable reactive phosphorus.

Please cite this article as: A. Holland, P.J. McInerney, M.E. Shackleton, et al., Dissolved organic matter and metabolic dynamics in dryland lowland rivers, Spectrochimica Acta Part A: Molecular and Biomolecular Spectroscopy, https://doi.org/10.1016/j.saa.2019.117871

4

A. Holland et al. / Spectrochimica Acta Part A: Molecular and Biomolecular Spectroscopy xxx (xxxx) xxx

Fig. 2. Dissolved organic matter (DOM) components derived using fluorescence excitation emission scans followed by PARAFAC analysis.

Dissolved organic matter aromaticity as indicated by specific UV absorbance @ 254 nm (SUVA254) values decreased in all three rivers during the experiment (Fig. 5). Fluorescence Index (FI) ranged from 1.43–1.69 (Fig. 5). Relative molecular weight, as indicated by abs254/ 365 and spectral slope275–295 varied between sampling periods with DOM in the Neimur and Yallkool rivers having the lowest molecular weight DOM (i.e. highest absorbance ratio of 254 nm to 365 nm (abs254/365)) on the 4th of December 2018 (Fig. 5). Temporal variation in autochthonous production as indicated by Biological Index (BIX) was shown in all three rivers. Increased autochthonous production (higher BIX) occurred in late November in the Wakool River and early December in the Neimur (Fig. 5). Significant positive relationships were shown between Protein-like1, BIX and FI, while a negative relationship was found between humic-like, BIX, FI and abs254/365 (Table 2). Chlorophyll–a was negatively related to the protein-like1 component and positively related to protein-like 2 DOM (Table 3).

3.3. Production Gross primary production (GPP), respiration, and net primary production (NPP) varied across the study period (Fig. 6). Whole river GPP estimates was generally low across all three sites, ranging from 0.01–0.47 mg O2 L −1 h−1 (Fig. 6). Community respiration ranged from 0.13–5.67 mg O2 L−1 h−1 (Fig. 6). Whole stream NPP was therefore largely negative, ranging from −0.09 to −5.66 mg O2 L−1 h−1 (Fig. 6). GPP estimates calculated from the pelagic photic zone were within the range recorded for the whole river (0.13–0.42 mg O2 L−1 h−1), whereas, mean respiration in the pelagic zone was lower than whole stream estimates, ranging between 0 and 0.06 mg O2 L−1 h−1 (Fig. 6). Pelagic NPP estimates were higher than whole stream estimates, ranging from 0.09–0.42 mg O 2 L −1 h−1 (Fig. 6). Whole river GPP and whole river respiration was positively related to DOC and temperature, while NPP on the other hand was negatively related (Fig. 7). No relationships were shown between whole stream and pelagic GPP, NPP, respiration and DOM components were found.

4. Discussion 4.1. Dissolved organic matter characterisation The three components described in this study showed N95% similarity to 25 models within the Openfluor database (https://openfluor. lablicate.com) (Supplementary Table 1) and have been found in all aquatic environments including rivers, wetlands, groundwater and oceans. Temporal and spatial variation in relative abundance of DOM components occurred across sampling times and between rivers, with the DOM pool generally dominated by the humic-like component (Fig. 3). The high abundance of the humic-like component found within these systems is indicative of the important role allochthonous DOM plays in lowland Australian rivers as this component is described in the literature as being terrestrially derived [27–31]. The high abundance of protein-like1 components shown over the period between 23rd November to the 4th December, is likely to represent an increase in microbial production during this period as protein-like1 and 2 both have been described elsewhere as indicative of freshly produced microbially derived carbon [32–37]. The increase in protein-like1 DOM also coincided with a decline in chlorophyll a (Supplementary Fig. 3) during the same period suggests that the increase in this protein-like DOM component during this period is driven by an increase in bacterial abundance rather than an increase in algae. On the other hand the protein-like 2 component was related to chlorophyll a concentrations suggesting an algae contribute to the production of this component. Dissolved organic matter aromaticity as indicated by SUVA254 values in freshwater generally range from 1 to 6 L mg C−1 m−1, with higher values reported from systems with a high terrestrial inputs [9,12]. All three river systems during the first two sampling periods reported SUVA254 values N6, suggesting a strong presence of terrestrial DOM during this period. This aligns with high abundance of the humic-like DOM component. The decline in SUVA254 over the sampling period within all three systems suggests that DOM decreased in aromaticity and may indicate increased autochthonous production, given autochthonous DOM produced by bacteria or algae is generally characterised by low SUVA254 [38]. Fluorescence Index (FI) is used as a proxy to determine the likelihood that DOM is derived from allochthonous (lower FI) or

Please cite this article as: A. Holland, P.J. McInerney, M.E. Shackleton, et al., Dissolved organic matter and metabolic dynamics in dryland lowland rivers, Spectrochimica Acta Part A: Molecular and Biomolecular Spectroscopy, https://doi.org/10.1016/j.saa.2019.117871

A. Holland et al. / Spectrochimica Acta Part A: Molecular and Biomolecular Spectroscopy xxx (xxxx) xxx

5

Fig. 3. Percent relative contribution of dissolved organic matter (DOM) components: Protein-like 1 (C1); Humic (C2); Protein-like 2 (C3), within the three rivers over the period October 2018 to January 2019; a) Fluorescence excitation emission matrices of DOM in river water on the 24th October 2018; bi and biii) 4th of December 2018; bii) 23rd of November 2018 and c) 3rd January 2019; i) Neimur River; ii) Wakool River; iii) Yallakool Creek.

autochthonous (higher FI) sources. Values from our three lowland rivers ranged from 1.43–1.69, indicating that DOM pool consisted of a mixture of allochthonous and autochthonous DOM [13]. Higher FI was also related to increased abundance of the protein-like1 DOM which has also been suggested to be a by-product of increased autochthonous production [32–37]. Relative molecular weights varied between sampling periods (Fig. 4) and this likely reflects changes in the contribution of allochthonous and autochthonous carbon to the DOM pool within these systems. Abs254/365 values generally range between 2.5 and 15 in freshwaters globally, with lower values representing higher molecular weight DOM [9]. The negative relationship between humic-like DOM and a bs254/365 supports the idea that humic-like DOM consists of a higher molecular weight than protein-like components. Similar relationships between molecular weight (abs254/365) and humic-like DOM

were also found by Holland et al. [9] for other Australian freshwaters. Temporal variations in autochthonous production as indicated by Biological Index (BIX) was also shown in all three rivers. Increased autochthonous production (higher BIX) was shown to occur in late November in the Wakool River and early December in the Neimur. BIX values N1.0 indicates the presence of freshly produced autochthonous DOM, whereas values b0.6 indicates lower levels of autochthonous DOM [38,39]. BIX values in this study ranged from 0.5–8.03 indicating that the autochthonous production within these systems varied greatly during the study period, but was extremely dominant on some occasions. This was supported by variations in the abundance of various DOM components during the study period with an increase in abundance of protein-like1 DOM and a decrease in humic-like DOM significantly related to an increase in autochthonous production as indicated by BIX index.

Please cite this article as: A. Holland, P.J. McInerney, M.E. Shackleton, et al., Dissolved organic matter and metabolic dynamics in dryland lowland rivers, Spectrochimica Acta Part A: Molecular and Biomolecular Spectroscopy, https://doi.org/10.1016/j.saa.2019.117871

6

A. Holland et al. / Spectrochimica Acta Part A: Molecular and Biomolecular Spectroscopy xxx (xxxx) xxx

Fig. 4. Relative abundance of protein-like1 (C1), humic-like (C2) and protein-like2 (C3) dissolved organic matter (DOM) in light (green) and dark (pink) bottles compared to the original river water (blue). Errors bars are ±1 SE. (For interpretation of the references to colour in this figure legend, the reader is referred to the web version of this article.)

4.2. Dissolved organic matter production and consumption Significant differences in the relative abundance of all three DOM components was shown between fresh river water and river water incubated within light and dark bottles for 6 h, with no differences between the light or dark treatments (Fig. 4). Humic-like DOM was shown to significantly decrease while both protein-like components increased over the 6 h incubation within both the light and dark treatments (Fig. 4). Clear temporal differences in the dominant protein-like component produced within the bottles were shown, with proteinlike 2 component dominant in November and January and the protein-like1 component dominant during December. The increase in protein-like DOM and decrease in the humic-like component within the bottles suggests microbial degradation of the humic-like component and production of protein-like material. It has been suggested that humic-like DOM may be more bioavailable to bacteria than lower molecular weight DOM [40] and that some bacteria prefer terrestrially derived DOM over algal DOM [41]. However, this contradicts numerous other reports that suggest humic-like DOM is less bioavailable than autochthonous protein-like DOM derived from algae and/or bacteria [54, 55]. Extracellular polymeric substance (EPS) produced by algae and bacteria are generally dominated by fluorescence DOM peaks similar to those displayed by both protein-like components in this study [42–44]. Li et al. [42] reported that there was no clear difference in the type of protein-like DOM produced by two types of bacteria compared to DOM produced by two algal species. The lack of a significant difference in the abundance of protein-like components between light and dark bottles suggests that bacterial activity is responsible for some of the patterns we observed, since if protein-like components were produced predominately by algae, we would have expected higher abundances of protein like components in light bottles. This is further supported by the significant negative relationship between chlorophyll-a and protein-like1 DOM in the water column, suggesting that this component was not related to increased algal abundance

rather likely to be bacterial derived (Table 2). However, the production of the protein-like 2 component is likely to be associated with both algal and bacterial production given its concentration in the water column was significantly related to higher chlorophyll a concentrations. Increases in protein-like DOM has been directly linked to bacterial cell counts [45–47] and the production of protein-like DOM within the bottles supports our contention that protein-like components within the riverine samples represent freshly produced autochthonous (microbial) DOM. The release of labile DOM compounds such as the protein-like DOM described in this study has been suggested to enhance the bacterial degradation of the more recalcitrant terrestrial DOM pool (Lambert and Perga, 2019). Therefore, bacteria and/or algae within the light and dark bottles might be releasing the protein-like substances to help aid in the degradation of the humic-like DOM in this study.

4.3. Production Microbial decomposition of organic matter and oxygen consumption by heterotrophs within the rivers in this study is greater than the production of oxygen from autotrophic activity, ultimately resulting in negative NPP values for the whole stream. Thus, overall the systems were heterotrophic over this study period. Heterotrophy tends to dominate many rivers, lakes and streams [5,48,49]. Other Australian rivers have also been found to be net heterotrophic [6,50]. In all three rivers metabolism appeared to be dominated by respiration occurring on benthic surfaces, with whole river estimates of respiration being much higher than pelagic respiration. Pelagic and whole river estimates showed that GPP in all three rivers was low (b0.5 mg/L h−1) and the similarity between whole river and pelagic values suggest that in our study systems, most GPP is derived from pelagic production by phytoplankton. This is supported by the significant positive relationship between pelagic GPP and chlorophyll-a; if surface biofilms were contributing significantly to overall GPP, we would expect higher whole river values comparable to our pelagic assay.

Please cite this article as: A. Holland, P.J. McInerney, M.E. Shackleton, et al., Dissolved organic matter and metabolic dynamics in dryland lowland rivers, Spectrochimica Acta Part A: Molecular and Biomolecular Spectroscopy, https://doi.org/10.1016/j.saa.2019.117871

A. Holland et al. / Spectrochimica Acta Part A: Molecular and Biomolecular Spectroscopy xxx (xxxx) xxx

7

Fig. 5. Dissolved organic matter (DOM) indices for the Neimur, Wakool Rivers and Yallakool Creek from the 24/10/2018 to the 3/1/2019 (n = 3); a) specific UV absorbance coefficient at 254 nm (SUVA254) as indicator of aromaticity; b) Fluorescence Index indicator of source of DOM; c) absorbance ratio at 254 nm to 365 nm as indicator of molecular weight; d) Biological Index (BIX) as indicator of autochthonous production and e) Spectral Slope between absorbance at 275–295 as another indicator of relative molecular weight. Boxplots represents mean and the upper and lower quartiles, whiskers are the 95th percentiles.

Gross primary production and respiration both increased with increased DOC concentration and increased temperatures, while NPP decreased. Thus, increases in DOC and temperature increased heterotrophy of these rivers. Respiration and primary production has previously been reported to be positively correlated with DOC and temperature [7,49,51], while NPP has also been reported to be negatively correlated with DOC, in a range of different freshwater environments [7,49]. However, it has been suggested that freshwaters, especially lakes with DOC concentrations b6 mg/L are generally autotrophic, with systems containing DOC concentrations above this said to be net heterotrophic [49,52]. The rivers in this study contained DOC concentrations ranging from 1.9–3.2 mg/L lower than the concentrations typically associated with heterotrophy. This may be due to decreased light penetration caused by high turbidity within these systems. No relationship was found Table 2 Pearson Correlation coefficients relating DOM components with DOM quality indices and chlorophyll-a. P values ≤.05 are highlighted in bold and indicate a significant relationship. Protein-like1

abs254/365 BIX FI SUVA254 Chl a Biofilm Chl a

Humic-like

Protein-like2

Correlation Coefficient

P Value

Correlation Coefficient

P Value

Correlation Coefficient

P Value

0.091 0.854 0.288 −0.095 −0.311 0.008

0.455 b0.001 0.016 0.432 0.008 0.944

−0.268 −0.683 −0.328 0.077 0.173 −0.036

0.026 b0.001 0.006 0.531 0.155 0.770

0.206 −0.206 0.082 −0.003 0.623 0.045

0.089 0.088 0.500 0.978 0.003 0.714

between the three DOM components described in this study and whole stream or pelagic indicators of productivity (GPP, NPP and respiration). Other authors, however, have reported that increases in GPP is related to increases in autochthonous protein-like DOM [56]. 5. Conclusion Our three rivers were net heterotrophic and increased heterotrophy was significantly related to (slight) increases in DOC and temperature. The relative abundance of key DOM components were also shown to vary temporally within the three river systems over the study period. Characterisation of DOM within the light and dark bottles after a 6hour incubation revealed the microbial consumption of humic-like component and production of the protein-like components. The lack of a significant difference between the light and dark bottles, the heterotrophic nature of the three systems and the negative correlation between protein-like1 DOM with chlorophyll suggests that the proteinlike1 component is likely bacterial in nature. However caution must be taken in regards to the negative relationship between chlorophyll a and protein-like1 DOM due to the low correlation coefficient. The positive relationship between the protein-like2 component and chlorophyll also suggests that algal abundance contributes to the production of this component in freshwaters, however, the lack of difference between light and dark bottles also suggests bacteria may also contribute to the production of this component. Therefore bacteria may be contributing to the production of protein-like DOM within heterotrophic freshwater systems and further research is needed to directly link this DOM component with bacterial cell counts in freshwaters.

Please cite this article as: A. Holland, P.J. McInerney, M.E. Shackleton, et al., Dissolved organic matter and metabolic dynamics in dryland lowland rivers, Spectrochimica Acta Part A: Molecular and Biomolecular Spectroscopy, https://doi.org/10.1016/j.saa.2019.117871

8

A. Holland et al. / Spectrochimica Acta Part A: Molecular and Biomolecular Spectroscopy xxx (xxxx) xxx

Fig. 6. Stream metabolism dynamics a) Respiration; b) Net Primary Production (NPP), c) Gross Primary Production (GPP), estimated as oxygen mg/L per hour calculated from open water estimates (Whole) and light and dark bottles (Pelagic) from the Niemur River, Yallakool Creek and Wakool River from 13 November–3 January 2019 (symbols represent the average of three readings).

Author contribution Aleicia Holland: Conceptualization, Methodology, Investigation, formal analysis, Data curation, Writing- Original draft preparation; Visualisation, Supervision.

Paul J. McInerney: Conceptualization, Methodology, Data curation, Writing- Original draft preparation; Visualisation, Supervision, Project administration; Funding acquisition; Michael E. Shackleton: Data curation, formal analysis, software, Writing – Review & Editing; Gavin N. Rees: Writing – Review & Editing,

Fig. 7. Whole river gross primary production, respiration, and net primary production (estimated as oxygen mg/L per hour) relationships with dissolved organic carbon concentrations (mg/L) and temperature (°C).

Please cite this article as: A. Holland, P.J. McInerney, M.E. Shackleton, et al., Dissolved organic matter and metabolic dynamics in dryland lowland rivers, Spectrochimica Acta Part A: Molecular and Biomolecular Spectroscopy, https://doi.org/10.1016/j.saa.2019.117871

A. Holland et al. / Spectrochimica Acta Part A: Molecular and Biomolecular Spectroscopy xxx (xxxx) xxx

Funding acquisition; Nick R. Bond: Software, Writing – Review & Editing and Ewen Silvester: Writing – Review & Editing, Supervision. Declaration of competing interest The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper. Acknowledgements This study was funded by the Murray–Darling Basin Authority (MDBA) and Joint Governments (New South Wales, Victoria, South Australia and Queensland). Dr. Aleicia Holland was funded via an Australian Research Council Discovery Early Career Research Award DE160100628. Robyn Watts and Nicole McCasker generously assisted in selecting and arranging access to study sites. Matthew O'Brien (MDBA), Daniel Coleman (NSW Department of Industry) provided valuable feedback over the course of the research. Field work assistance was provided by Rochelle Petrie and Chris Davey. FEEMs were conducted by Stuart Smith as part of a student placement. Appendix A. Supplementary data Supplementary data to this article can be found online at https://doi. org/10.1016/j.saa.2019.117871. References [1] M.R. Grace, D.P. Giling, S. Hladyz, V. Caron, R.M. Thompson, R. Mac Nally, Fast clestation estimation, 13 (2015), e10011. [2] P.J. Mulholland, C.S. Fellows, J.L. Tank, N.B. Grimm, J.R. Webster, S.K. Hamilton, E. Martí, L. Ashkenas, W.B. Bowden, W.K. Dodds, W.H. McDowell, M.J. Paul, B.J. Peterson, Inter-biome comparison of factors controlling stream metabolism, Freshw. Biol. 46 (2001) 1503–1517. [3] M.R. Grace, S.J. Imberger, Stream Metabolism: Performing and Interpreting Measurements, Water Studies Centre Monash University, Murray Darling Basin Commission and New South Wales Department of Environment and Climate Change, 2006 204. [4] P.J. McInerney, G.N. Rees, B. Gawne, P. Suter, Invasive Salix fragilis: altered metabolic patterns in Australian streams, Hydrobiologia 767 (1) (2016) 267–277. [5] C.M. Duarte, Y.T. Prairie, Prevalence of heterotrophy and atmospheric CO2 emissions from aquatic ecosystems, Ecosystems 8 (2005) 862–870. [6] B. Gawne, C. Merrick, D.G. Williams, G. Rees, R. Oliver, P.M. Bowen, S. Treadwell, G. Beattie, I. Ellis, J. Frankenberg, Z. Lorenz, Patterns of primary and heterotrophic productivity in an arid lowland river, River Res. Appl. 23 (2007) 1070–1087. [7] J. Ask, J. Karlsson, M. Jansson, Net ecosystem production in clear-water and brownwater lakes, Glob. Biogeochem. Cycles 26 (2012). [8] R.L. Oliver, C.J. Merrick, Partitioning of river metabolism identifies phytoplankton as a major contributor in the regulated Murray River (Australia), Freshw. Biol. 51 (6) (2006) 1131–1148. [9] A. Holland, J. Stauber, C.M. Wood, M. Trenfield, D.F. Jolley, Dissolved organic matter signatures vary between naturally acidic, circumneutral and groundwater-fed freshwaters in Australia, Water Res. 137 (2018) 184–192. [10] A. Holland, L.J. Duivenvoorden, S.H.W. Kinnear, Naturally acidic waterways: conceptual food webs for better management and understanding of ecological functioning, Aquat. Conserv. Mar. Freshwat. Ecosyst. 22 (2012) 836–847. [11] J.L. Meyer, The microbial loop in flowing waters, Microb. Ecol. 28 (1994) 195–199. [12] R. Jaffé, D. McKnight, N. Maie, R. Cory, W.H. McDowell, J.L. Campbell, Spatial and temporal variations in DOM composition in ecosystems: the importance of longterm monitoring of optical properties, Journal of Geophysical Research: Biogeosciences 113 (2008). [13] R.M. Cory, D.M. McKnight, Fluorescence spectroscopy reveals ubiquitous presence of oxidized and reduced quinones in dissolved organic matter, Environmental Science & Technology 39 (2005) 8142–8149. [14] J.B. Fellman, E. Hood, R.G.M. Spencer, Fluorescence spectroscopy opens new windows into dissolved organic matter dynamics in freshwater ecosystems: a review, Limnol. Oceanogr. 55 (2010) 2452–2462. [15] T. Lambert, F. Darchambeau, S. Bouillon, B. Alhou, J.D. Mbega, C.R. Teodoru, F.C. Nyoni, P. Massicotte, A.V. Borges, Landscape control on the spatial and temporal variability of chromophoric dissolved organic matter and dissolved organic carbon in large African Rivers, Ecosystems 18 (2015) 1224–1239. [16] J.H. Larson, P.C. Frost, M.A. Xenopoulos, C.J. Williams, A.M. Morales-Williams, J.M. Vallazza, J.C. Nelson, W.B. Richardson, Relationships between land cover and dissolved organic matter change along the river to lake transition, Ecosystems 17 (8) (2014) 1413–1425, https://doi.org/10.1007/s10021-014-9804-2.

9

[17] G.A. Weyhenmeyer, R.A. Müller, M. Norman, L.J. Tranvik, Sensitivity of freshwaters to browning in response to future climate change, Clim. Chang. 134 (2016) 225–239. [18] M. Winterdahl, M. Erlandsson, M.N. Futter, G.A. Weyhenmeyer, K. Bishop, Intraannual variability of organic carbon concentrations in running waters: drivers along a climatic gradient, Glob. Biogeochem. Cycles 28 (2014) 451–464. [19] P.J. Curtis, D.W. Schindler, Hydrologic control of dissolved organic matter in loworder Precambrian shield lakes, Biogeochemistry 36 (1997) 125–138. [20] D.M. McKnight, E.W. Boyer, P.K. Westerhoff, P.T. Doran, T. Kulbe, D.T. Andersen, Spectrofluorometric characterization of dissolved organic matter for indication of precursor organic material and aromaticity, Limnol. Oceanogr. 46 (2001) 38–48. [21] J. Dahlén, S. Bertilsson, C. Pettersson, Effects of UV-A irradiation on dissolved organic matter in humic surface waters, Environ. Int. 22 (1996) 501–506. [22] J.L. Weishaar, G.R. Aiken, B.A. Bergamaschi, M.S. Fram, R. Fujii, K. Mopper, Evaluation of specific ultraviolet absorbance as an Indicator of the chemical composition and reactivity of dissolved organic carbon, Environmental Science & Technology 37 (2003) 4702–4708. [23] P.J. McInerney, R.J. Stoffels, M.E. Shackleton, C.D. Davey, Flooding drives a macroinvertebrate biomass boom in ephemeral floodplain wetlands, Freshwater Science 36 (2017) 726–738. [24] D. Giling, N. Bond, BASEmetab: Estimate Single-station Whole-stream Metabolic Rates from Diel Dissolved Oxygen (DO) Curves, 2019 (R package version 2.3.2). [25] K.R. Murphy, C.A. Stedmon, D. Graeber, R. Bro, Fluorescence spectroscopy and multiway techniques. PARAFAC, Anal. Methods 5 (2013) 6557–6566. [26] H. Wickham, ggplot2: Elegant Graphics for Data Analysis, Springer-Verlag, New York, 2016. [27] M. Chen, S.-H. Kim, H.-J. Jung, J.-H. Hyun, J.H. Choi, H.-J. Lee, I.-A. Huh, J. Hur, Dynamics of dissolved organic matter in riverine sediments affected by weir impoundments: production, benthic flux, and environmental implications, Water Res. 121 (2017) 150–161. [28] J. D’Andrilli, J.R. Junker, H.J. Smith, E.A. Scholl, C.M. Foreman, DOM composition alters ecosystem function during microbial processing of isolated sources, Biogeochemistry 142 (2019) 281–298. [29] M. Derrien, M.-S. Kim, G. Ock, S. Hong, J. Cho, K.-H. Shin, J. Hur, Estimation of different source contributions to sediment organic matter in an agricultural-forested watershed using end member mixing analyses based on stable isotope ratios and fluorescence spectroscopy, Sci. Total Environ. 618 (2018) 569–578. [30] M. Derrien, K.-H. Shin, J. Hur, Assessment on applicability of common source tracking tools for particulate organic matter in controlled end member mixing experiments, Sci. Total Environ. 666 (2019) 187–196. [31] L. Jørgensen, C.A. Stedmon, T. Kragh, S. Markager, M. Middelboe, M. Søndergaard, Global trends in the fluorescence characteristics and distribution of marine dissolved organic matter, Mar. Chem. 126 (2011) 139–148. [32] A. Brym, H.W. Paerl, M.T. Montgomery, L.T. Handsel, K. Ziervogel, C.L. Osburn, Optical and chemical characterization of base-extracted particulate organic matter in coastal marine environments, Mar. Chem. 162 (2014) 96–113. [33] K.R. Murphy, A. Hambly, S. Singh, R.K. Henderson, A. Baker, R. Stuetz, S.J. Khan, Organic matter fluorescence in municipal water recycling schemes: toward a unified PARAFAC model, Environmental Science & Technology 45 (2011) 2909–2916. [34] A.-R. Schittich, U.J. Wünsch, H.V. Kulkarni, M. Battistel, H. Bregnhøj, C.A. Stedmon, U.S. McKnight, Investigating fluorescent organic-matter composition as a key predictor for arsenic mobility in groundwater aquifers, Environmental Science & Technology 52 (2018) 13027–13036. [35] C.A. Stedmon, D.N. Thomas, M. Granskog, H. Kaartokallio, S. Papadimitriou, H. Kuosa, Characteristics of dissolved organic matter in Baltic Coastal Sea ice:Allochthonous or autochthonous origins? Environmental Science & Technology 41 (2007) 7273–7279. [36] Y. Yamashita, J.N. Boyer, R. Jaffé, Evaluating the distribution of terrestrial dissolved organic matter in a complex coastal ecosystem using fluorescence spectroscopy, Cont. Shelf Res. 66 (2013) 136–144. [37] Y. Yamashita, L.J. Scinto, N. Maie, R. Jaffé, Dissolved organic matter characteristics across a subtropical Wetland's landscape: application of optical properties in the assessment of environmental dynamics, Ecosystems 13 (2010) 1006–1019. [38] A.M. Hansen, T.E.C. Kraus, B.A. Pellerin, J.A. Fleck, B.D. Downing, B.A. Bergamaschi, Optical properties of dissolved organic matter (DOM): effects of biological and photolytic degradation, Limnol. Oceanogr. 61 (2016) 1015–1032. [39] H.J. Dalmagro, M.J. Lathuillière, F.d.S. Sallo, M.F. Guerreiro, O.B. Pinto, P.H.Z. de Arruda, E.G. Couto, M.S. Johnson, Streams with riparian Forest buffers versus impoundments differ in discharge and DOM characteristics for pasture catchments in southern Amazonia, Water 11 (2019) 390. [40] Y. Yamashita, C.G. Fichot, Y. Shen, R. Jaffé, R. Benner, Linkages among fluorescent dissolved organic matter, dissolved amino acids and lignin-derived phenols in a riverinfluenced ocean margin, Front. Mar. Sci. 2 (2015). [41] T. Roiha, S. Peura, M. Cusson, M. Rautio, Allochthonous carbon is a major regulator to bacterial growth and community composition in subarctic freshwaters, Sci. Rep. 6 (2016), 34456. [42] S. Li, H. Winters, L.O. Villacorte, Y. Ekowati, A.-H. Emwas, M.D. Kennedy, G.L. Amy, Compositional similarities and differences between transparent exopolymer particles (TEPs) from two marine bacteria and two marine algae: significance to surface biofouling, Mar. Chem. 174 (2015) 131–140. [43] L.O. Villacorte, Y. Ekowati, T.R. Neu, J.M. Kleijn, H. Winters, G. Amy, J.C. Schippers, M.D. Kennedy, Characterisation of algal organic matter produced by bloomforming marine and freshwater algae, Water Res. 73 (2015) 216–230. [44] H. Xu, Z. Yan, H. Cai, G. Yu, L. Yang, H. Jiang, Heterogeneity in metal binding by individual fluorescent components in a eutrophic algae-rich lake, Ecotoxicol. Environ. Saf. 98 (2013) 266–272.

Please cite this article as: A. Holland, P.J. McInerney, M.E. Shackleton, et al., Dissolved organic matter and metabolic dynamics in dryland lowland rivers, Spectrochimica Acta Part A: Molecular and Biomolecular Spectroscopy, https://doi.org/10.1016/j.saa.2019.117871

10

A. Holland et al. / Spectrochimica Acta Part A: Molecular and Biomolecular Spectroscopy xxx (xxxx) xxx

[45] S. Cumberland, J. Bridgeman, A. Baker, M. Sterling, D. Ward, Fluorescence spectroscopy as a tool for determining microbial quality in potable water applications, Environ. Technol. 33 (2012) 687–693. [46] S. Elliott, J.R. Lead, A. Baker, Characterisation of the fluorescence from freshwater, planktonic bacteria, Water Res. 40 (2006) 2075–2083. [47] B.G. Fox, R.M.S. Thorn, A.M. Anesio, D.M. Reynolds, The in situ bacterial production of fluorescent organic matter; an investigation at a species level, Water Res. 125 (2017) 350–359. [48] W.K. Dodds, A.M. Veach, C.M. Ruffing, D.M. Larson, J.L. Fischer, K.H. Costigan, Abiotic controls and temporal variability of river metabolism: multiyear analyses of Mississippi and Chattahoochee River data, Freshwater Science 32 (2013) 1073–1087. [49] P.C. Hanson, D.L. Bade, S.R. Carpenter, T.K. Kratz, Lake metabolism: relationships with dissolved organic carbon and phosphorus, Limnol. Oceanogr. 48 (2003) 1112–1119. [50] S. Vink, M. Bormans, P.W. Ford, N.J. Grigg, Quantifying ecosystem metabolism in the middle reaches of Murrumbidgee River during irrigation flow releases, Mar. Freshw. Res. 56 (2005) 227–241.

[51] B.O.L. Demars, J. Russell Manson, J.S. Ólafsson, G.M. Gíslason, R. Gudmundsdóttir, G. Woodward, J. Reiss, D.E. Pichler, J.J. Rasmussen, N. Friberg, Temperature and the metabolic balance of streams, Freshw. Biol. 56 (2011) 1106–1121. [52] Y.T. Prairie, D.F. Bird, J.J. Cole, The summer metabolic balance in the epilimnion of southeastern Quebec lakes, Limnol. Oceanogr. 47 (2002) 316–321. [53] A. Huguet, L. Vacher, S. Relexans, S. Saubusse, J. Froidefond, E. Parlanti, Properties of fluorescent dissolved organic matter in the Gironde Estuary, Organic Geochemistry 40 (6) (2009) 706–719, https://doi.org/10.1016/j.orggeochem.2009.03.002. [54] F. Guillemette, S.L. McCallister, P.A. del Giorgio, Differentiating the degradation dynamics of algal and terrestrial carbon within complex natural dissolved organic carbon in temperate lakes, J. Geophys. Res. Biogeosci. 118 (2013) https://doi.org/10. 1002/jgrg.20077. [55] T. Lambert, M. Perga, Non-conservative patterns of dissolved organic matter degradation when and where lake water mixes, Aquatic Sciences 81 (4) (2019) 64, https://doi.org/10.1007/s00027-019-0662-z. [56] S. Halbedel, O. Büttner, M. Weitere, Linkage between the temporal and spatial variability of dissolved organic matter and whole-stream metabolism. Biogeosciences 10 (8) (2013) 5555–5569, https://doi.org/10.1007/s00027-019-0662-z.

Please cite this article as: A. Holland, P.J. McInerney, M.E. Shackleton, et al., Dissolved organic matter and metabolic dynamics in dryland lowland rivers, Spectrochimica Acta Part A: Molecular and Biomolecular Spectroscopy, https://doi.org/10.1016/j.saa.2019.117871