High winds, storms and tidal change in estuaries increase shear stress on the benthic sediment, resulting in increased SPM concentrations and transport of particulate material to coastal waters   . SPM has a large exposed surface area compared with benthic sediment, with easier access to dissolved oxygen, organic matter, macronutrients and micronutrients, facilitating the attachment and growth of bacteria   . Therefore resuspension of particulate matter can give rise to chemical and biological changes in the water column.
The release/uptake P mechanisms between SPM and water has been studied in recent years. The key factors are temperature    pH   , dissolved oxygen  , redox conditions (Eh)   , microbes   , SPM size distribution   and SPM texture  , the organic matter content of the sediment and salinity  .
Normally, under natural conditions, the phosphorus concentrations (including organic and inorganic phosphorus) should have a balance between adsorption and desorption processes of SPM. The complexity of SPM composition has the capability to allow P concentration to be constant to a certain level in the aqueous solution to a certain time. The property of SPM to resist fast uptake or release P to the aqueous solution is called buffer capacity. If the SPM has buffer capacity to P compounds, it means that the SPM could control the P transfer and possess at a certain time, in addition to reduce/slow down negative effects of P contaminants on the water quality.
This paper is a cross-disciplinary study which is providing the effects of exogenous nutrient input such as organic and inorganic phosphorus fluxes on water quality. To achieve this, we focused on SPM on the effects of “alive” or “dead”, “short term” or “long term” resuspension, response to the “labile” or “refractory” organic phosphorus and “ideal” or river or sea water matrices. The specific objectives were to: determine P speciation in the autoclaved and non-autoclaved SPM; investigate phosphorus equilibration in UHP water, biotic river water and autoclaved river water over short (5 h) and long (3 day) time scales; quantify DIP (orthophosphate) and DOP (G-6-P and phytic acid) exchange between SPM and the water column at a fixed shear stress in spiked UHP water, biotic river water and autoclaved river water over short (20 h) and long (5 day) time scales; study the effects of the water matrix and bacteria on phosphorus exchange.
The suspended particulate matter and water were collected from the Tamar Estuary which is a typical, turbid macrotidal estuary for which extensive background water quality data are available  -  . The Tamar Estuary drains an area of 590 km2. The length of the estuary, from Gunnislake (the tidal limit) to the mouth of the estuary in Plymouth Sound is 31.5 km. The geology of the region is mainly composed of underlying slate, limestone and grit. Estuarine flushing time is typically one day in winter  , when instantaneous flow can exceed 38 m3∙s−1, increasing to one week in summer when flows can be as low as 5 m3∙s−1. A sedimentation basin particle sampler (SBPS), supplied by Dr Matthias Ricking from the Free University of Berlin, was used to collect resuspended particles (kg amounts) for characterization and experimentation. Before deployment the SBPS sampler was soaked in 0.5% m/v detergent (Nutricon), rinsed thoroughly with ultra high purity (UHP) water and air dried. The SBPS was deployed with the inlets facing upstream at a depth of ~1 m at a riverine sampling station near Gunnislake (SX 426 725) for 2 - 4 week periods during the winter. Samples were then transferred from the SBPS into high density polyethylene (HDPE) bottles using an acid washed plastic scoop. The sample slurries were then frozen at −20˚C until use. The sedimentation basin system consisted of a sedimentation unit and a filtration unit. River water was routed into the sedimentation basin via inlet holes and slowed down by the blades and SPM was deposited in the sedimentation basin by gravitational settling.
2.2. Water Matrices and Sterilisation of SPM and River Water
1) UHP water (≥18.2 MΩ∙cm−1). This was used to simulate phosphorus release from suspended sediment to a theoretically pristine river water (salinity = 0, no phosphorus) at pH ~ 6.
2) Gunnislake river water. This was used to determine phosphorus release from suspended sediment to a real river water (S = 0) at pH ~ 7.
3) Autoclaved Gunnislake river water. This was used to investigate phosphorus release from suspended sediment to a real river water under autoclaved conditions (Salinity = 0) at pH ~ 7. Ten L of filtered Gunnislake water was autoclaved for 30 min at 121˚C and 15 lb in−2 pressure.
2.3. Experimental Design
Phosphorus equilibration and exchange experiments were set up in a laboratory environment to simulate Gunnislake SPM under water conditions. Previous paper had the details of the P model compounds selection  . Resuspension was achieved by stirring the solution with a magnetic stirrer at a fixed stirring rate (1200 rpm, to keep the particulate material in suspension). Gunnislake SPM and water separately autoclaved and then was mixed in a ratio of 4 g slurry to 1 L of water at ambient temperature (10˚C - 22˚C) in a well oxygenated environment (O2 saturation > 90%). Autoclaved experiments were carried out in a class 100 laminar flow cabinet (BassAIRE). To investigate natural DIP and DOP equilibration. The kinetics of P equilibration were investigated over short (5 h) and long (3 day) time scales. Autoclaved conditions were maintained in control experiments; to investigate DIP and DOP exchange in these continuously stirred, pre-equilibrated batch reactors, G-6-P (labile), phytic acid (refractory) and orthophosphate were separately spiked into the UHP and river water matrices to make sure the final concentrations are 80 µg P L−1. The concentration of 80 µg P L−1 was chosen because it was at the upper end of the range of TDP concentrations typically found in the Gunnislake river water and because it provided analytically robust measurements. The kinetics of P exchange were investigated over short (15 h) and long (6 day) time scales.
2.4. Sample Sequence
Stage 1. The kinetics of P exchange was investigated over the short term (5 h) and long term (3 days).
1) The water matrix (1 L) was stirred, the T0 sample was taken with a syringe, and then SPM slurry (0.5 g∙L−1) was added.
2) From T1 - T10 samples (30 mL) were taken every 30 min.
3) Sample T11 was taken after 24 h.
4) Sample T12 was taken after 48 h.
5) Sample T13 was taken after 72 h.
Stage 2. The reactors were separately spiked with ortho-phosphate (potassium dihydrogen phosphate KH2PO4), G-6-P and phytic acid.
6) Sample S1 was taken after each reactor had reached a steady state (defined by a change of <5% in the P concentration in the water column between samples). Then each reactor was immediately spiked with 80 µg P L−1 ortho-phosphate, G-6-P or phytic acid.
7) Samples S2, S3, S4 and S5 were taken every 4 h at 4, 8, 12 and 16 h for the short-term investigation.
8) Sample S6 was taken after 30 - 34 h.
9) Samples S7, S8, S9, S10 were taken every 20 - 24 h at 70, 90, 114 and 138 h for the long-term investigation.
The fixed stirring rate is to make sure the SPM well mixed and we assume that the total volume of sample (420 mL) would not affect the results.
3. Results and Discussion
3.1. Phosphorus Content in the Autoclaved and Non-Autoclaved SPM
Autoclaving is commonly used to sterilise water and sediment/SPM by denaturing the bacteria with high temperatures and pressures  . The autoclaving process can however break down the physical structure of particulate matter, increasing the surface to volume ratio  so for this work the temperature was kept relatively low (121˚C). Autoclaving will lyse any bacterial cells in the SPM but could also increase the rate of hydrolysis of any organic phosphorus compounds naturally present. However, experiment (Table 1) shows that there was no significant difference in the concentration of inorganic or organic P between the untreated and autoclaved Gunnislake SPM.
The relatively large error bars are due to the heterogeneous nature of the SPM slurry and possibly time differences in the storage period and analysis of different sample.
Table 1. Phosphorus concentrations in Gunnislake SPM. The concentration values given are the means of three separate samples analysed in triplicate (n = 9). Error bars represent ± 1 s.d.
3.2. Phosphorus Equilibration in UHP Water, Biotic River Water and Autoclaved River Water
The aim of these first experiments was therefore to investigate the release and equilibration of phosphorus species naturally present in the SPM in the presence of different water masses. The results are shown in Figure 1 and the mean concentration changes after 5 h and 72 h are given in Table 2. Six batch reactors were set up (duplicate experiments in UHP, biotic and autoclaved Gunnislake river water).
Figure 1(a) and Table 2 show that the releases of DIP and DOP at 5 h were 33 and 20 µg∙L−1, respectively, from the SPM to UHP water. Desorption is likely to be the dominant mechanism. The concentration of DIP increased linearly with time over the 5 h period (R2 = 0.986, slope = 7.02 µg P L−1 h−1), whereas DOP concentration increased up to 3.5 h and then gradually decreased. The release of DIP to UHP water continued to increase linearly over the 72 h period (R2 = 0.913, slope = 1.361 µg P L−1 h−1) with an overall increase of 103 µg P L−1 from the start of the experiment (Table 2). The theoretical maximum release of DIP from SPM was 820 µg P L−1 (Table 1); therefore 12.5% of the total IP pool was released from SPM over 72 h. In contrast the DOP profile showed more temporal variability, with an overall net gain of 56 P L−1 in the water column after 72 h, which is 6 % of the total OP pool in the SPM.
Figure 1(b) shows that in biotic river water DIP and DOP had an inverse relationship over the initial 5 h period. DIP decreased for the first 2 h and then increased slightly from 2 - 5 h; in contrast DOP increased and then decreased over the same time period. However, considering the 5 - 72 h period (Figure 1(e)), the DIP concentration remained broadly constant, with no significant net change (−7 µg P L−1). Therefore, although the system is fairly dynamic, the river water matrix appears to inhibit the release of DIP with respect to UHP water. Since the initial DIP concentration in the river water decreases to only 3 µg P L−1, this appears to be controlled by the ionic strength of the water matrix. The DOP concentration increased to a steady state water column concentration of 80 µg P L−1 at 72 h, which is equivalent to 12 % of the total OP pool available from the SPM. The DOP profiles for UHP and river water were similar over the 72 h timescale, suggesting that the water matrix did not have a significant impact, and showed a net release (UHP: 56 µg P L−1; river water: 58 µg P L−1). Desorption from the SPM is the most likely release mechanism after the initial 5 hour
Table 2. Mean water column DIP/DOP concentrations after 5 h and 72 h equilibration. Error bars represent +/− 1 s.d. of 6 experiments with each sample analysed in triplicate (n = 18) in UHP and Biotic river water; Error bars in abiotic river water experiments represent +/− 1 s.d. of 4 experiments with each sample analysed in triplicate (n = 12).
Figure 1. Comparison of P concentrations (♦ = DIP; ■ = DOP) in UHP and Gunnislake river water during short term and long term equilibration experiments. Error bars represent +/− 1 s.d. of 6 experiments with each sample analysed in triplicate (n = 18) in UHP and Biotic river water; Error bars in abiotic river water experiments represent +/− 1 s.d. of 4 experiments with each sample analysed in triplicate (n = 12).
period. The results suggest that the Gunnislake SPM had an initial capacity to take up DIP from the biotic river water, either by bacterial uptake on the SPM surface and/or physical adsorption.
Figure 1(c) shows the changes in the DIP and DOP concentrations during resuspension and equilibration in autoclaved river water. In the first 1.5 h the DIP concentration did not change significantly but the release of DOP to the water column was more extensive than in the biotic equivalent (from 10 to 43.8 µg P L−1). This may be due to the autoclaving process facilitating OP release, e.g. from lysed bacteria  . Therefore over the first 5 h period there was a large increase in DOP (55 µg P L−1) but a relatively small decrease in DIP (9 µg P L−1, Table 2). Figure 1(f) shows that the DIP concentration gradually decreased further to 12 µg P L−1 at 72 h, with a net loss of 33 µg P L−1.
These experiments were intended to be compared to the biotic system to elucidate biological processes; however, autoclaving can sterilize the particle/SPM    , and thus the autoclaved processes taking place. The physicochemical speciation of iron is particularly important for inorganic phosphorus exchange and this can change during autoclaving  . Gunnislake sediment contains 39,500 - 52,100 mg∙kg−1 of iron and sequential extraction of phosphorus from this sediment showed that 465 - 744 µg P g−1, representing 55% - 62% of the TP pool, was bound to iron, making this the dominant P fraction in Gunnislake river sediment  . The effects of autoclaving are therefore likely to be an important factor in the contrasting exchange kinetics of DIP in the biotic and autoclaved systems. The DOP concentration remained relatively stable over the 5 - 72 h period, which is consistent with the idea of short term release of OP from lysed cells and indicates equilibration between the SPM and the water column, with no enzymatic hydrolysis.
3.3. Phosphorus Exchange in Spiked UHP Water, Biotic River Water and Autoclaved River Water
The aim of these experiments was to investigate the effect of adding orthophosphate, G-6-P and phytic acid to an equilibrated SPM-water system, in UHP water and Gunnislake river water. The ability of SPM to adsorb P from the water column is called the SPM phosphorus buffering   capacity. The higher the SPM P-buffering capacity, the more P can potentially be absorbed by the SPM  .
In the UHP water systems, the DIP concentration in the water did not change significantly after spiking with orthophosphate (Figure 2(a)), suggesting rapid adsorption of the orthophosphate. The concentration then stayed constant up to 12 h, and between 12 and 16 h DIP desorbed to reach its equilibrium concentration of 168 ± 34 µg P L−1 (Table 3). This means that the SPM buffering capacity  lasted for 12 h for DIP, thus an exogenous nutrient input of DIP to a turbulent river (the “ideal” river) would be expected to be buffered for around 12h. The uncertainties associated with the DIP results include the differences between duplicate experiments. The DOP concentration did not change in the UHP water over either time scale, suggesting that there was no DOP/DIP inter-conversion during the experiment.
The SPM buffering capacity was also apparent in biotic river water, but the subsequent release of DIP was much slower than into UHP water, and continued over the entire experiment (192 h; Figure 2(b)). Generally, the DIP concentration in the biotic river water had a linear relationship with time (R2 = 0.89, slope = 0.61 µg P L−1 h−1). The final concentration was comparable to that found in the
Figure 2. Comparison of DIP (♦) and DOP (■) concentrations during step change experiments spiked with 80 µg P L−1 orthophosphate. Error bars represent 2 experiments and each sample was analysed in triplicate (n = 6). (a) UHP water; (b) river water. Time 0 represents the DIP and DOP background concentrations and the time at which the orthophosphate was added. The dashed lines separate the short term and long term time scales.
Table 3. Mean water column P concentration 16 h and 190 h after spiking with 80 µg P L−1 as orthophosphate.
UHP water (198 µg P L−1). Therefore, the SPM buffering capacity in real river systems may be longer-lived than suggested by the UHP water experiments. This indicates that water matrices can positively affect the “P-buffering” capacity of SPM as seen previously by  . This is consistent with the higher concentrations of exchangeable Al, Fe and organic matter content in river water   . The addition of orthophosphate resulted in a rapid release of DOP, which increased from 23 to 51 µg P L−1, presumably due to exchange, but there was no further significant change (p > 0.05) over the remaining of the experiment.
G-6-P is quite labile and hence is a potential source of P for bacteria   . Figure 3(a), b and c show P exchange between SPM and the water phases after spiking with G-6-P. The DOP concentrations increased after (4 h) the G-6-P spikes, by 49 µg P L−1 (UHP), 52 µg P L−1 (biotic) and 70 µg P L−1 (autoclaved). Therefore the SPM in UHP and biotic river water had similar buffering capacity for G-6-P of about 30 µg P L−1, whereas in the autoclaved system the buffering capacity was only 10 µg P L−1.
The DOP concentrations generally decreased to 80 - 100 µg P L−1 (Table 4) during the 4 to 16 h in all the matrices, as DIP increased, indicating natural hydrolysis of G-6-P after the initial release. Over the longer term, DOP concentrations in UHP and biotic river water remained constant, whereas DOP in autoclaved system continued to decrease to 45 µg P L−1. The DIP concentrations did
Table 4. Mean water column P concentration 16 h and 140 h after spiking with 80 µg P L−1 as G-6-P.
Figure 3. DIP (♦) and DOP (■) concentration changes during step change experiments with UHP water (a and d), biotic river water (c and e) and autoclaved river water (c and f) spiked with 80 µg P L−1 G-6-P (a)-(c) and phytic acid (d)-(f). Error bars represent 2 experiments and each sample was analysed in triplicate (n = 6). Time 0 represents the DIP and DOP background concentrations and the time at which the G-6-P or phytic acid was added; the dashed lines separate the short term and long term time scales.
not change after 16 h in the UHP water and biotic river water systems, but increased in the autoclaved system, with overall DOP/DIP R2 of 0.936. This suggests that hydrolysis is most extensive in the autoclaved system, with all of the added DOP hydrolysed over the course of the experiment.
The SPM also showed a high buffering capacity for phytic acid in the biotic and autoclaved river water systems (Figure 3), as the DIP and DOP concentrations did not increase significantly on addition of the spike. In the UHP water systems, a slight initial increase in both DIP and DOP was observed (DIP from 73 - 84 µg P L−1; DOP from 30 - 31 µg P L−1; Table 5), but there was no significant (P = 0.77) change after this (Figure 3(d)). These changes in the DOP and
Table 5. Mean water column P concentration 16 h and 190 h after spiking with 80 µg P L−1 as phytic acid.
DIP concentrations indicate that approximately 79 µg P L−1 of phytic acid attached to the SPM and replaced 11 µg P L−1 of DIP.
In the biotic river water experiment, the DIP concentration varied over time, showing a general increase from 77 - 100 µg P L−1 and the DOP concentration did not change measurably (Figure 3(e)). In the autoclaved river system, the DOP concentration increased by 29 µg P L−1 over the first 12 h, and there was no significant DIP release (Figure 3(f)) over the same time scale. The longer term profile (16 - 150 h) showed a “mirror-image” relationship with DOP decreasing and DIP increasing (DOP/DIP, R2 = 0.693). Considering the experimental conditions and the refractory nature of phytic acid, the physical exchange of phytic acid with orthophosphate at the SPM surface should be the only possible process in this experiment. Therefore, since the initial sorption of phytic acid did not displace IP, it appears that the slower second phase of sorption involves a different process from the initial buffering.
In these constantly stirred batch reactor experiments, the water matrix had a significant effect on the release of IP and OP during equilibration with SPM and on the buffering capacity of the SPM when the systems were spiked with orthophosphate. SPM exhibited a greater DIP buffering capacity in the river water matrix compared with UHP water during equilibration. Following addition of P species to the equilibrated system, the SPM was able to buffer all of the orthophosphate and phytic acid added to both UHP and river water. However, the initial SPM buffering capacity was much lower for G-6-P. However hydrolysis of the G-6-P occurred within the first 16 h in the UHP (56%) and biotic river water (55%) experiments, and effectively all of the G-6-P added was hydrolysed in the autoclaved experiments over 140 h. The buffering capacity of the SPM for phytic acid was high (95% - 98% removed) and permanent in UHP and biotic river water, whereas phytic acid was cycled in the autoclaved experiments.
The results in this paper suggest that in a turbulent river, DIP and DOP will reach equilibrium between the particulate matter and the water column within 24 h. If there is an exogenous nutrient input of DIP to the river, the initial buffering capacity of the SPM will be effective for 12 - 16 h. If there is an input of DOP, the SPM will have a high and permanent buffering capacity for refractory species such as phytic acid, but a much lower buffering capacity for labile species such as G-6-P. The water matrix will however have limited influence on the exchange of added DOP between SPM and water. Any input of labile DOP to a river will undergo hydrolysis and release DIP over short timescales. Any input of refractory DOP however will not be hydrolysed but will be involved in P cycling through displacement of DIP from the surface of particulate matter. Finally, the Gunnislake River clearly cannot represent the “ideal” river, therefore P dynamics with different environmental conditions requires further studies for the next step.
This project is financially supported by The National Key Research and Development Program of China (2017YFA0604903) and the National Natural Science Foundation of China (41571510).