Reciprocal carbon subsidies between autotrophs and bacteria in stream food webs under stoichiometric constraints

Soils are currently leaching out their organic matter at an increasing pace and darkening aquatic ecosystems due to climate and land use change, or recovery from acidification. The implications for stream biogeochemistry and food webs remain largely unknown, notably the metabolic balance (biotic CO2 emissions), reciprocal subsidies between autotrophs and bacteria, and trophic transfer efficiencies. We use a flow food web approach to test how a small addition of labile dissolved organic matter affects the strength and dynamics of the autotrophs-bacteria interaction in streams. Our paired streams whole-ecosystem experimental approach combined with continuous whole-stream metabolism and stable isotope probing allowed to unravel carbon fluxes in the control and treatment streams. We increased the natural supply of dissolved organic matter for three weeks by only 12% by continuously adding 0.5 mg L−1 of sucrose with a δ13C signature different from the natural organic matter. Both photosynthesis and heterotrophic respiration increased rapidly following C addition, but this was short lived due to N and P stoichiometric constraints. The resulting peak in heterotrophic respiration was of similar magnitude to natural peaks in the control observed when soils were hydrologically connected to the streams and received soil derived carbon. Carbon reciprocal subsidies between autotrophs and bacteria in the control stream accounted for about 50% of net primary production and 75% of bacterial production, under low flow conditions when stream water was hydrologically disconnected from soil water. The reciprocal subsidies were weaker by 33% (autotrophs to bacteria) and 55% (bacteria to autotrophs) in the treatment relative to the control. Net primary production relied partly (11% in the control) on natural allochthonous dissolved organic carbon via the CO2 produced by bacterial respiration. Many large changes in ecosystem processes were observed in response to the sucrose addition. The light use efficiency of the autotrophs increased by 37%. Ecosystem respiration intensified by 70%, and the metabolic balance became relatively more negative, i.e. biotic CO2 emissions increased by 125%. Heterotrophic respiration and production increased by 89%, and this was reflected by a shorter (−40%) uptake length (SwOC) and faster (+92%) mineralisation velocity of organic carbon. The proportion of DOC flux respired and organic carbon use efficiency by bacteria increased by 112%. Macroinvertebrate consumer density increased by 72% due to sucrose addition and consumer production was 1.8 times higher in the treatment than in the control at the end of the experiment. The trophic transfer efficiencies from resources to consumers were similar between the control and the treatment (2-5%). Synthesis. Part of the carbon derived from natural allochthonous organic matter can feed the autotrophs via the CO2 produced by stream bacterial respiration, intermingling the green and brown webs. The interaction between autotrophs and bacteria shifted from mutualism to competition with carbon addition under nutrient limitation (N, P) increasing biotic CO2 emissions. Without nutrient limitation, mutualism could be reinforced by a positive feedback loop, maintaining the same biotic CO2 emissions. A small increase in dissolved organic carbon supply from climate and land use change could have large effects on stream food web and biogeochemistry with implications for the global C cycle under stoichiometric constraints.


Summary 22
1. Soils are currently leaching out their organic matter at an increasing pace and darkening aquatic 23 ecosystems due to climate and land use change, or recovery from acidification. The implications for 24 stream biogeochemistry and food webs remain largely unknown, notably the metabolic balance (biotic 25 CO2 emissions), reciprocal subsidies between autotrophs and bacteria, and trophic transfer 26 efficiencies. 27 28 2. We use a flow food web approach to test how a small addition of labile dissolved organic matter 29 affects the strength and dynamics of the autotrophs-bacteria interaction in streams. Our paired streams 30 whole-ecosystem experimental approach combined with continuous whole-stream metabolism and 31 stable isotope probing allowed to unravel carbon fluxes in the control and treatment streams. 32 33 3. We increased the natural supply of dissolved organic matter for three weeks by only 12% by 34 continuously adding 0.5 mg L -1 of sucrose with a δ 13 C signature different from the natural organic 35 matter. Both photosynthesis and heterotrophic respiration increased rapidly following C addition, but 36 this was short lived due to N and P stoichiometric constraints. The resulting peak in heterotrophic 37 respiration was of similar magnitude to natural peaks in the control observed when soils were 38 hydrologically connected to the streams and received soil derived carbon. 39 40 4. Carbon reciprocal subsidies between autotrophs and bacteria in the control stream accounted for 41 about 50% of net primary production and 75% of bacterial production, under low flow conditions 42 when stream water was hydrologically disconnected from soil water. The reciprocal subsidies were 43 weaker by 33% (autotrophs to bacteria) and 55% (bacteria to autotrophs) in the treatment relative to 44 the control. Net primary production relied partly (11% in the control) on natural allochthonous 45 dissolved organic carbon via the CO2 produced by bacterial respiration. 46 47 5. Many large changes in ecosystem processes were observed in response to the sucrose addition. The 48 light use efficiency of the autotrophs increased by 37%. Ecosystem respiration intensified by 70%, 49 and the metabolic balance became relatively more negative, i.e. biotic CO2 emissions increased by 50 125%. Heterotrophic respiration and production increased by 89%, and this was reflected by a shorter 51 (-40%) uptake length (SwOC) and faster (+92%) mineralisation velocity of organic carbon. The 52 proportion of DOC flux respired and organic carbon use efficiency by bacteria increased by 112%. 53 54 6. Macroinvertebrate consumer density increased by 72% due to sucrose addition and consumer 55 production was 1.8 times higher in the treatment than in the control at the end of the experiment. The 56 trophic transfer efficiencies from resources to consumers were similar between the control and the 57 treatment (2-5%). 58 59 7. Synthesis. Part of the carbon derived from natural allochthonous organic matter can feed the 60 autotrophs via the CO2 produced by stream bacterial respiration, intermingling the green and brown 61 webs. The interaction between autotrophs and bacteria shifted from mutualism to competition with 62 carbon addition under nutrient limitation (N, P) increasing biotic CO2 emissions. Without nutrient 63 limitation, mutualism could be reinforced by a positive feedback loop, maintaining the same biotic 64 CO2 emissions. A small increase in dissolved organic carbon supply from climate and land use change The global annual riverine flux of organic C (0.26-0.53 Pg C year −1 ) to the oceans is comparable to 75 the annual C sequestration in soil (0.4 Pg C year −1 ), suggesting that terrestrially derived aquatic losses 76 of organic C may contribute to regulating changes in soil organic carbon storage (Dawson 2013). 77 Soils are currently leaching out their organic matter at faster rates to aquatic ecosystems due to 78 climate and land use change, or recovery from acidification (e.g. Freeman  aggregates (Grossart 2010). In theory, bacteria and autotrophs could compete for limiting nutrients 99 (Currie and Kalff 1984), notably when bacteria have lower C:nutrient biomass ratios than autotrophs 100 (Daufresne and Loreau 2001 The potential benefit of bacterial CO2 for primary producers has been hypothesised to explain a small 112 increase (16-20%) in gross primary production following dissolved organic carbon addition and 113 associated increase in bacterial activities (Robbins et al. 2017). The reciprocal carbon subsidies 114 between autotrophs and bacteria in stream food webs has however, to our knowledge, not been 115 studied either empirically or theoretically. It raises the possibility that the carbon of the primary 116 producers may be partly derived from allochthonous organic matter processed by the bacteria, and 117 intermingle the green and brown webs (Zou et  Consumers and the trophic transfer efficiency should benefit from an increase in bacterial production.

136
The design also allows to quantify the potential of priming of allochthonous OM by sucrose via 137 bacterial activity (Kunc et al. 1976, Hotchkiss et al. 2014. 138 We combined whole-ecosystem stable isotope probing (using sucrose from sugar cane) and Bayesian 139 mixing model to characterise the carbon links (sources to mixtures). We converted the relative carbon 140 fluxes from the different sources, into carbon fluxes by estimating the production of the mixtures, 141 after taking into account assimilation efficiencies (e.g. carbon use efficiency, bacterial growth 142 efficiency). This approach relies on the integration of stream metabolism, nutrient cycling, ecological 143 stoichiometry, stable isotope probing of the food web and production estimates (see Welti et al. 2017).

144
In this study, we focus on the basal part of the food web, and notably on the elusive reciprocal C 145 subsidies between autotrophs and bacteria.  Sunnyvale, California). The limits of detections were 0.001 for NO3 and PO4 and 0.003 mg L -1 for Cl. 218 In order to provide a more comparable indicator of nutrient cycling for different hydrological 219 conditions, the uptake velocity vf was also calculated as follows: vf = uz/Sw, with u average water 220 velocity and z average depth. Short uptake lengths and fast uptake velocities indicate fast cycling rates 221 (high exchange rates between water and benthos). distances between oxygen stations corresponded to 80-90% of the oxygen sensor footprints (3u/k2), 240 with u/k2 entirely independent of discharge (R 2 =0.0005), which allowed the manipulated reach to be 241 independent of the control reach. The DOC injection point was 28 m upstream of the top station of the 242 manipulated reach, and this distance corresponded to 69% of the oxygen sensor footprint of the top 243 station. All sondes were deployed from May to October 2007, logging at 5 min time step interval. 244 The net metabolism was only calculated for stable flow conditions (3-32 L s -1 ), as follows (Demars 245 2018): 246 with NEPt net ecosystem production at time t (g O2 m -2 min -1 ), CAV average dissolved oxygen (g O2 m - with Q discharge, a and b constants, permitting to correct for baseflow (first term of the equation) and 254 soil water (second term of the equation) lateral inflows, see Demars (2018). The proportion (± se) of 255 total lateral inflows relative to discharge (Qg/Q) was 10.7±0.6%, 6.6±0.5%, and 2.3±0.4% for the 256 Birnie Burn control, Cairn Burn control and treated reach, respectively, independently of discharge in 257 the range 3.8-32.5 L s -1 (stable flows). 258 All calculations were run in Excel using a preformatted spreadsheet (Demars 2018). The overall 259 uncertainties in daily stream metabolism, including cross-calibration errors, individual parameter 260 uncertainties, spatial heterogeneity (through the average of diel O2 curves) and correction for lateral 261 inflows, were propagated through all the calculations using Monte Carlo simulations (Demars 2018). 262 The corrections for lateral inflows amounted to about 6% of ER for the treated reach (Cairn Burn), 263 19% and 16% in the control reaches, Cairn Burn and Birnie Burn, respectively. green algae and bryophytes were collected by hand along both studied reaches before and after 269 sucrose addition. All samples were freeze dried and milled prior to analyses for C, N, δ 13 C and δ 15 N.

270
The main source of inorganic carbon for primary producers was assumed to be CO2 because of the 271 low alkalinity (remaining below 0.5 meq HCO3 L -1 under low flows). The fractionation factor for CO2 272 assimilation into macrophyte tissue is known to vary with pCO2 and growth rate, and was set at - We suggest another approach: δ 13 C of dissolved CO2 may be estimated indirectly under low flows 290 using the fractionation factor of Rubisco -25.5±3.5‰ and δ 13 C of bryophytes (strict CO2 user and no 291 CO2 transport limitation). In our study, the average δ 13 C of bryophytes was -36.4‰, and assuming the 292 above fractionation coefficient of -25.5‰, Glensaugh δ 13 C of dissolved CO2 would be -10.9‰. This 14.7‰ and bryophytes as -39.2‰, suggesting a fractionation coefficient of -24.5‰ by difference. In 297 our calculations under low flow conditions we therefore assumed δ 13 C of dissolved CO2 as -11±3‰. 298 To quantify the reciprocal subsidies between autotrophs and bacteria, it remained to decompose the 299 overall stable isotope signature of stream dissolved CO2 into the allochthonous (groundwater, soil 300 water and atmospheric exchange) and autochthonous (respiration by heterotrophs and autotrophs) 301 sources. The allochthonous signature, δ 13 CCO2-allochthonous, can be deduced from rearranging: 302 where FCO2 represents CO2 fluxes (g C m -2 day -1 , with all fluxes expressed as positive values) and δ 13 C 306 the isotope signature (‰) of the different sources of CO2. We averaged the δ 13 C of the autotrophs 307 (filamentous green algae and biofilm primary producers). The δ 13 CCO2-allochthonous was only calculated 308 for the control stream under low stable flows and assumed to apply to both streams. Uncertainties 309 were propagated in quadrature using standard deviation δx for sums, and relative uncertainties δx/x for 310 the division. collected by hand along both studied reaches before and after sucrose addition. Since there was hardly 327 any difference in δ 13 C between DOC and CPOM (δ 13 C = -27.4±0.7 ‰, Table S1), we used the δ 13 C of 328 CPOM determined in this study as the signature for allochthonous organic carbon. 329 330 Periphyton. Periphyton (or biofilm) samples represent a mixture of primary producers (algae and 331 cyanobacteria), bacteria and fine particulate organic matter. The samples were collected before and at 332 the end of the sucrose addition from the flumes and stones with a toothbrush, funnel and bottle. All 333 samples were freeze dried and milled. 334 We also placed six pairs (with/without Vaseline) of unglazed ceramic tiles (10 x 10 cm) fixed on 335 bricks and deployed along the studied reaches in both streams three weeks before the start of the 336 manipulation. Vaseline was applied around half the tiles to prevent grazing by invertebrates. After 337 three weeks, there was hardly any growth on the tiles, and so the tiles were left in the stream until the 338 end of the manipulation. One brick in the control stream was lost. At the end of the experiment, the 339 tiles were frozen at -20°C, later freeze dried and the biofilm was scraped with a razor blade. Since 340 there was little biomass per tile (about 1 g C m -2 ), the biofilm was pooled by stream and grazer 341 treatments (leaving two samples per stream). Very little grazing activity was observed on the tiles 342 during the six weeks and unsurprisingly no difference in biofilm dry mass emerged due to grazer 343 exclusion (paired t-tests on ln transformed mass; Birnie, t4, p=0.13; Cairn, t5, p=0.26). Quantification and δ13C values of the PLFAs were both determined by Gas Chromatography-348 Combustion-Isotope Ratio Mass Spectrometry (GC-C-IRMS) as described by Main et al. (2015), and 349 averaged for each stream (δ 13 C,  within the same study (Davidson and Cullen 1957). 446 The flux of CO2 was then related to discharge within the range of low stable flows for which stream 447 metabolism was processed (Cairn n=47, R 2 =0.81; Birnie n=65, R 2 =0.77) to provide daily estimates.

448
For more details, see Demars 2018. 449 450 Biotic CO2 emissions. These were simply calculated as the net ecosystem production (NEP), gross 451 primary production (GPP) plus ecosystem respiration (ER, a negative flux) expressed in g C m -2 day -1 . 452 Bacterial respiration of DOC was calculated as heterotrophic respiration (HR, a negative flux) from: 453 with AR, autotrophic respiration and ER, ecosystem respiration, both negative fluxes (oxygen 455 consumption) and GPP, positive flux (producing oxygen). We partitioned ER into auto and 456 heterotrophic respiration with α=0.5 (see ) and calculated uncertainties using α=0.2 457 and α=0.8. Bacterial respiration of the added sucrose was calculated as the difference in heterotrophic 458 respiration between the treatment and a control reach during sucrose addition, after standardising for 459 site differences using the control period. Light. We derived a conversion factor for photosynthetically active radiation (PAR; 1 mol photon m -2 475 day -1 = 6.13 g C m -2 day -1 ) by relating the ratio of total quanta to total energy within the PAR concentrations (28/08, 5/09, 11/09/2007) multiplied by discharge (Fig S1). With all fluxes expressed in g C m -2 day -1 , we calculated the light use efficiency (net primary 504 production/photosynthetic radiations), organic carbon use efficiency (bacterial production/DOC 505 supply), resource use efficiency (consumers/ [net primary production + bacterial production]), and the 506 proportion of biotic CO2 emissions (net ecosystem production / total CO2 emissions). 507 508 Data analyses 509 We calculated an effect size (i.e. proportional changes) of sucrose addition on our response variables 510 (e.g. nutrient cycling, stoichiometric ratios, metabolic fluxes, ecosystem efficiencies) using the values 511 of the control (C) and treatment (T) reaches, before (B) and after (A) sucrose addition as follows: 512 with all uncertainties propagated in quadrature using standard deviation δx for sums, and relative 514 uncertainties δx/x for the division. The standard error (se) of the effect size was calculated as = 515 /√ with n average number of independent samples or measurements in CB, CA, TB, TA (although 516 strictly speaking the samples were pseudo-replicated because the random collection was within one 517 plot). Since the overall design was unreplicated we simply interpreted the effect size in relation to the 518 standard error (we do not report p values). 519 The control period for the metabolic parameters and trophic transfer efficiencies was limited to the 520 two days prior to sucrose addition (21-22 August), so the whole data series was within a single period 521 of low stable flows, i.e. not interrupted by peak flows, producing more comparable results across 522 sites. Daily metabolic parameters were temporally pseudo-replicated (especially during the control 523 period), so it was only possible to report relative uncertainties (δx/x). 524 All effect sizes and efficiencies were calculated at the reach scale. 525 526

527
DOC addition 528 We were fortunate to have no rainfall and low stable flows during the whole three weeks of carbon 529 addition. The target concentration was achieved at the top of the studied reach, 28 m downstream of 530 the injection point, where the DOC addition averaged 0.52 mg C L -1 (Fig. S1), despite a lower than 531 expected discharge (from 20 L s -1 down to 8 L s -1 by the end of the addition), due to significant uptake 532 and mineralisation within the 28 m mixing zone (an average 55% loss of sucrose flux from the point 533 of injection to the top of the reach).

535 536
Stream metabolism continuous monitoring 537 Bryophytes covered 4, 11 and 17% of the river bed in the Birnie control, Cairn control and Cairn 538 treatment reach, respectively. Filamentous green algae (mostly Microspora sp., Microsporaceae) 539 percentage cover increased during the period of sucrose addition from 1 to 19%, 11 to 33% and 4 to 540 37% of channel width in the Birnie control, Cairn control and Cairn treatment reach, respectively. 541 Gross primary production (GPP) peaked to 7.6 g O2 m -2 day -1 ten / eleven days after the start of 542 sucrose addition in the treatment reach before decreasing sharply down to an average 2.4 g O2 m -2 543 day -1 during the last four days of the carbon addition, and this despite high photosynthetic active 544 radiations. In contrast, GPP remained relatively constant in the control reaches Birnie Burn (about 1.2 545 g O2 m -2 day -1 ) and Cairn Burn (3.2 g O2 m -2 day -1 for the first two weeks declining to 2.1 g O2 m -2 546 day -1 during the last week g O2 m -2 day -1 ) - Fig. 4. 547 Peaks in ecosystem respiration (ER) down to -20 g O2 m -2 day -1 (Birnie Burn) and -35 g O2 m -2 day -1 548 (Cairn Burn) were more visible for the controls than the treatment reach, with respiration activity 549 inversely related to soil hydrological connectivity, as recorded by soil moisture continuous monitoring 550 (Fig. 4). There was an increase in ER in the treatment reach at the start of the sucrose addition, despite 551 the continuing loss of hydrological connectivity. More specifically, heterotrophic respiration activity 552 associated to sucrose addition peaked sharply 15 days after the start of the addition, processing up to 553 59% of the daily sucrose flux (Fig. 5). On average 35±20% of the added sucrose was respired during 554 the addition over just 84 m (or 15 minutes mean travel time). Heterotrophic production ranged 555 between 2% and 10% of the sucrose flux, based on bacterial growth efficiencies of 0.05 and 0.2 556 respectively. 557 558 559 560

Nutrient cycling studies and stoichiometry 561
The background concentrations of nitrate and phosphate were 180 and 90 µg N L -1 and 2 and 4 µg P 562 L -1 in the Birnie control and Cairn treatment reach, respectively. The added geometric mean of N and 563 P were on average 471 µg N L -1 and 24 µg P L -1 . The addition of sucrose had no effect on nitrate and 564 phosphate nutrient uptake length and uptake velocity (Fig. 6, Table S3). The phosphate uptake length 565 was highly related to discharge and became very short, down to 31 m in the treatment reach towards 566 the end of the sucrose addition. Phosphate uptake velocity was about 0.2 mm s -1 and an order of 567 magnitude faster than nitrate with uptake lengths in the kilometre range. 568 The molar C:N:P stoichiometric ratios of coarse particulate organic matter and bryophytes remained 569 stable throughout the experiment. Sucrose addition exerted strong effects on filamentous green algae 570 and periphyton stoichiometry (Fig. 7, Table S4). While the molar C:N:P stoichiometric ratios 571 decreased in the control stream, they increased sharply in the treatment reach following sucrose 572 addition: from 330:29:1 to 632:49:1 in filamentous green algae and from 262:29:1 to 428:38:1 in 573 periphyton.

Quantification of carbon fluxes and efficiencies 594
The flow food web of the control and treatment reach over the three weeks of sucrose addition were 595 quantified according to our conceptual model (see Fig. 1). Figure 10 illustrates the C fluxes of net 596 primary production, bacterial production and secondary production as well as bacterial CO2 flux and 597 overall net ecosystem production (emission of CO2 to the atmosphere). Photosynthetic active radiation 598 (light) decreased slightly from 106 to 80 and 118 to 90 g C m -2 day -1 in the control and treatment 599 respectively. Allochthonous organic matter (DOC) was more than halved from 177 to 85 and 110 to 600 50 g C m -2 day -1 in the control and treatment. This was reflected by a general reduction in the organic 601 carbon uptake length 3214 to 2531 m and 4257 to 1886 m in the control and treatment, respectively, 602 independently of the carbon addition. The organic carbon uptake velocity decreased in the control 603 from 0.82 to 0.55 m day -1 but increased in the treatment from 0.53 to 0.76 m day -1 .

605
Our estimates suggest that a large part of net primary production (0.23, range 0.09-0.38 g C m -2 day -1 ) 606 was used for bacterial production (0.15-0.69 g C m -2 day -1 ) in the control, and in turn nearly half of 607 the CO2 fixed by autotrophs was derived from bacterial CO2, although it represented a small fraction 608 of bacterial respiration driving net ecosystem production (biotic CO2 emissions). These reciprocal C 609 subsidies between autotrophs and bacteria were not as strong relative to net primary production (0.91, 610 range 0.36-1.46 g C m -2 day -1 ) and bacterial production (0.14-0.66 g C m -2 day -1 ) in the treatment 611 during sucrose addition. The estimated flux of allochthonous organic matter assimilated by bacteria 612 was similar in the control (0.04-0.17 g C m -2 day -1 ) and treatment (0.02-0.10 g C m -2 day -1 ). Part of the 613 allochthonous organic matter respired by bacteria was recycled by primary producers and accounted 614 for 11±6% of net primary production in the control.

616
We also derived, from the BACI design, the effect size of sucrose addition on selected whole-617 ecosystem metabolic properties and efficiencies (see Fig. 11, Table S6). All estimated ecosystem 618 properties and efficiencies were summarised in Table S7 for Birnie control and Cairn treatment before 619 and after sucrose addition. We present some key highlights below. 620 621 While GPP increased marginally (12%), there was a small relative increase in light use efficiency 622 (37%). ER intensified by 70%, and the net ecosystem production (NEP) became relatively more 623 negative by 125%, i.e. 125% relative increase in biotic CO2 emissions. Heterotrophic respiration and 624 production increased by 89%, and this was reflected by a shorter (-40%) uptake length (SwOC) and 625 faster mineralisation velocity (92%) of organic carbon. The proportion of DOC flux respired (range 626 2.2-5.3%) and organic carbon use efficiency by bacteria (range 0.1-0.3%) increased by 112%. While  627 there was a relative decrease (-20%) in total CO2 emissions, the proportion of biotic CO2 emission 628 increased by 88%. The reciprocal subsidies between autotrophs and bacteria were weaker by 33% 629 (autotrophs to bacteria) and 55% (bacteria to autotrophs) in the treatment relative to the control.

631
The average consumer biomass per individual was similar between the control (0.20 mg C ind -1 ) and 632 the treatment reach (0.21 mg C ind -1 ) and average production per individual was slightly higher in the 633 treatment reach (6 µg C ind -1 day -1 ) than in the control (5 µg C ind -1 day -1 ), at the end of the sucrose 634 addition. Consumer density (range 1300-6000 ind m -2 ) increased by 72% due to sucrose addition and 635 consumer production was higher in the treatment (36 mg C m -2 day -1 ) than in the control (20 mg C m -2 636 day -1 ) at the end of the experiment. The resource use efficiency (trophic transfer efficiency) by 637 consumers was similar between the control and the treatment reach (2-5%), with a size effect of 638 sucrose addition ranging from -33% to +8% depending on the heterotrophic growth efficiency (0.05 639 and 0.2, respectively) used to calculate heterotrophic production. 640 641

644
Our experiment showed that a small continuous addition of labile DOC (0.52 mg C L -1 as sucrose, 645 12% of total DOC, Fig. S1) can profoundly alter whole-ecosystem behaviour. The use of a before and 646 after control experiment together with the addition of a deliberate tracer with a distinctive δ 13 C 647 signature allowed not only to trace the fate of the added carbon into the treated reach but also to build 648 the flow food web of the control reach, unravel C reciprocal subsidies between autotrophs and 649 bacteria, and demonstrate the potential for some natural allochthonous organic matter to feed the 650 primary producers via bacterial respiration. 651 652 653

Reciprocal subsidies between autotrophs and bacteria 654
The use of autotroph carbon by bacteria has been shown before using a photosystem II inhibitor in 655 biofilm (Neely and Wetzel 1995)  The bacterial CO2 flux to autotrophs assumed its δ 13 C signature was the same as that of the bacteria, 681 but the proportion of C sources used in bacterial production may be different to the proportion of C 682 sources respired by bacteria. The controlled experiment allowed to calculate the bacterial respiration 683 of the added sucrose as 2.16 g C m -2 day -1 over the three weeks of sucrose addition. This represented 684 82% (53-184%) of the estimated treatment bacterial respiration (2.64 ±1.45 g C m -2 day -1 , Table S7) 685 and the proportion of sucrose in bacteria was estimated at 51 ±7 % (Fig. S2). So the estimates were 686 still within measurement errors. 687

688
The autotrophs did not include bryophytes in our flow food web calculations because their 689 contribution to primary production was thought to be negligible over the few weeks of the 690 experiment, and particularly towards the end of the experiment (on which data the flow food webs 691 were based) when filamentous green algae were covering bryophytes. The lack of bryophyte growth 692 under very low phosphorus concentrations (here 2-4 µg P L -1 of soluble reactive P) combined with 693 shading by epiphytes has been well documented (e.g. Finlay and Bowden 1994). This may also 694 explain the lack of changes in bryophyte C:N:P stoichiometry (Fig. 7). 695 696 697 Boom and bust: role of nutrients 698 Gross primary productivity (GPP) appeared to be stimulated by the addition of sucrose but this was 699 short lived, despite sustained light availability during the addition period (Fig. 4). Heterotrophic 700 respiration of sucrose peaked after two weeks (three days after GPP) but crashed within days while 701 the supply of sucrose was continuously flowing through the reach (with sucrose concentration 702 increasing from 0.22 to 0.88 mg C L -1 with falling discharge). This was in sharp contrast to the peaks 703 in respiration followed by a more sustained response of ecosystem (mostly heterotrophic) respiration 704 to hydrological connectivity with soil water in the control reaches (Fig. 4, see Demars 2018). This 705 boom and bust in the treated reach was likely due to nutrient limitation, mostly P according to the 706 changes in filamentous green algae and periphyton C:N:P stoichiometry. Surprisingly, the shortfall of 707 nutrients (N, P) was not compensated by faster nutrient cycling rates, as observed in previous studies 708 with much higher labile DOC additions (e.g. Bernhardt and Likens 2002). This may be due to 709 limitation of phosphate uptake by NH4 availability (long term median in both streams 7 µg N L -1 ) in 710 streams relatively rich in nitrate (see Oviedo-Vargas et al. 2013). The higher primary productivity in 711 the Cairn burn may partly result from higher P availability (4 µg P L -1 ) than in the control Birnie Burn 712 (2 µg P L -1 ), despite lower nitrate availability (90 µg N L -1 versus 180 µg N L -1 , respectively) - Table  713 S3. These different nutrient supply rates may reflect the legacy of past experimental amendments (Ca, 714 N, P, K) in 33 ha of the Cairn burn catchment aimed to increase grassland productivity in the late 715 1970s and early 1980s (Hill Farming Research Organisation 1983).

717 718
Metabolic balance 719 The metabolic balance (or net ecosystem production) was responsible for a quarter of CO2 emissions 720 in the control and increased from 6 to 12 % in the treatment reach during sucrose addition. CO2 721 emissions from these streams were therefore largely dominated by soil CO2 derived from the 722 mineralisation of soil organic matter, rather than rock weathering of Dalradian acid schist drifts 723 (Demars 2018 Fate of carbon 733 Sucrose is very labile and is well known to promote the growth of filamentous bacteria Sphaerotilus 734 natans ("sewage fungus"), even at relatively low concentrations (0.25-1.00 mg L -1 ) in a stable flow 735 forested stream (Warren et al. 1964). This was not observed in this study, likely because the moorland 736 streams studied here were more open and colonised by Microspora, a common genus of filamentous 737 green algae in Scottish streams (Kinross et al. 1993). Microspora was able to uptake sucrose by 738 osmotrophy (Wright and Hobbie 1966) but this accounted to only 10% of C uptake by Microspora. 739 Overall daily uptake of sucrose by autotrophs represented 1.2% of the average daily flux of sucrose. 740

741
The proportion of added labile carbon (sucrose) in the consumers varied widely between species 742 (Table S5) The treatment reach did not show the large peaks in respiration (outside the period of sucrose 754 addition) that the control reaches showed when the catchment was hydrologically connected, as 755 indicated by soil moisture (Fig. 4, Demars 2018). This is likely because the treatment reach is a more 756 constrained reach largely disconnected from the land. It was initially chosen to avoid lateral inflows 757 which were very small (2.3%) and mostly from a spring fed flush (i.e. groundwater), rather than 758 seepage from organic and riparian soils known to stimulate bacterial activity (e.g. Brunke and Gonser 759 1997, Pusch et al. 1998). Interestingly the average bacterial respiration was similar in the control and 760 treatment reaches over the three weeks of sucrose addition (Fig. 10), suggesting part of the organic 761 matter respired in the control was relatively labile and comparable to the 0.5 mg C L -1 of added 762 sucrose. The dynamic of this respiration within the three weeks was very different however, 763 decreasing with the progressive loss of hydrological connectivity with soils in the control, and 764 peaking after two weeks in the treatment (Fig. 4). 765 766 767 By and far, the largest quantity of carbon processed by bacteria was lost as CO2 emission.

768
Heterotrophic respiration over the treated reach respired (on average) 35% of the added sucrose. 769 Bacterial production averaged 2-10% of the sucrose flux. Heterotrophic respiration of natural 770 allochthonous DOC was about 3% in the control stream and 2.2 % prior to sucrose addition in the 771 treatment. The fact that sucrose was processed 10 times faster than natural DOC was well reflected by 772 the shortening of the organic uptake length (SwOC) and increased mineralisation velocity (vf-OC). This uptake lengths in the control and the differences between control and treatment prior to sucrose 777 addition can be explained by the loss of hydrological connectivity with soil water, as indicated by the 778 changes in soil moisture, and the difference in lateral inflows between the control (10.7%) and the 779 treatment (2.3%). The shortening of the organic uptake length (SwOC) reflected more hydrological 780 changes as indicated by the different direction of change in organic carbon uptake (vfOC) between the 781 control and the treatment. In the control vfOC declined by 32% against a 52% decline in DOC supply. 782 In contrast, vfOC increased by 43% in the treatment with the addition of labile carbon, despite a 55% 783 fall in DOC supply (similar to the control). Overall the uptake lengths of natural DOC (i.e. excluding 784 those influenced by sucrose addition) were longer (2.5-4.3 km) than the length of the first order 785 streams studied here (1 km), and so a large part of the carbon is released to downstream ecosystems as 786 previously observed (e.g. Wiegner et al. 2005), especially during time of loss of hydrological 787 connectivity with the soil of the catchment (Demars 2018). The simple comparison between flow food 788 webs (Fig. 10) suggest that the addition of labile carbon did not prime the bacterial use of natural 789 allochthonous DOC.

791
Choice of carbon for DOC addition 792 We initially considered to add natural DOC to the stream after isolating DOC using reverse osmosis 793 (Sun et al. 1995; RealSoft Pro2S, US) but the product was too salty with high pH and high nutrient 794 concentrations (Stutter and Cains 2016). While reverse osmosis may be combined with electrodialysis 795 to avoid co-concentration of salt (Koprivnjak et al. 2006), the quantities needed for our whole-796 ecosystem experiment were simply too large. We also considered several commercial humate sources 797 but rejected their use because of pH, solubility and nutrient issues. Sucrose derived from sugarcane 798 (C4 carbon fixation) has a very distinctive carbon stable isotope signature compared to the autotrophs 799 in temperate ecosystems (C3 carbon fixation). It offered the possibility to trace its fate through the 800 food web, and even to identify the bacterial carbon pathways in the control streams. It turned out that 801 sucrose was a more judicious choice than first thought because labile DOC (polysaccharide, amino 802 acids) is likely driving the respiration of the studied streams at the land-water interface (Demars 803 2018), as found in bioreactors (e.g. Drake et al. 2015). Acknowledgments 817 We thank Carol Taylor and Helen Watson for managing the long-term monitoring, Yvonne Cook and 818 Susan McIntyre for running water chemical analyses, Claire Abel for the phospholipid fatty acid 819 extraction, Maureen Procee for running the compound specific isotope ratio analysis, Gillian Martin 820 for preparing and running the samples for stable isotope ratio analysis, Glensaugh farm manager 821 Donald Barrie for hosting BOLD and JLK during the experiment and facilitating our work, and 822 Baptiste Marteau for help with macroinvertebrate identification and comments on the manuscript.  based on the before and after control impact experimental design, except for resource use efficiency 1283 by consumers and algae-bacteria reciprocal subsidies relying on a simple comparison between the 1284 control and the treatment reach. See Table S6 for uncertainties (most very large) and Table S7 for the 1285 individual values. The resource use efficiency by consumers was estimated for two heterotrophic 1286 growth efficiency (HGE). 1287