TYPE Original Research PUBLISHED 26 October 2022 DOI 10.3389/fmars.2022.1017052 Contrasting particle fluxes and OPEN ACCESS composition in a submarine EDITED BY Kostas Kiriakoulakis, Liverpool John Moores University, canyon affected by natural United Kingdom REVIEWED BY sediment transport events Manuel Vargas-Yáñez, Spanish Institute of Oceanography (IEO), Spain and bottom trawling Mauro Celussi, Istituto Nazionale di Oceanografia e di Geofisica Sperimentale, Italy Sarah Paradis 1,2,3*, Marta Arjona-Camas 2, Miguel Goñi 4, *CORRESPONDENCE Albert Palanques 2, Pere Masqué 1,5,6 and Pere Puig 2 Sarah Paradis 1 Institute of Environmental Science and Technology (ICTA), Universitat Autònoma de Barcelona, sparadis@ethz.ch Bellaterra, Spain, 2 Marine Sciences Institute, Consejo Superior de Investigaciones Cientı´ficas (ICM- SPECIALTY SECTION CSIC), Barcelona, Spain, 3 Geological Institute, Department of Earth Sciences, Eidgenössische This article was submitted to Technische Hochschule (ETH) Zürich, Zürich, Switzerland, 4 College of Earth, Ocean, and Deep-Sea Environments and Ecology, Atmospheric Sciences, Oregon State University, Corvallis, OR, United States, 5 School of Natural a section of the journal Sciences, Centre for Marine Ecosystems Research, Edith Cowan University, Joondalup, WA, Frontiers in Marine Science Australia, 6 Marine Environment Laboratories, International Atomic Energy, Principality of Monaco, Monaco RECEIVED 11 August 2022 ACCEPTED 28 September 2022 PUBLISHED 26 October 2022 CITATION Submarine canyons are important conduits of sediment and organic matter to Paradis S, Arjona-Camas M, Goñi M, Palanques A, Masqué P and Puig P deep-sea environments, mainly during high-energy natural events such as (2022) Contrasting particle fluxes and storms, river floods, or dense shelf water cascading, but also due to human composition in a submarine canyon affected by natural sediment transport activities such as bottom trawling. The contributions of natural and trawling- events and bottom trawling. induced sediment and organic matter inputs into Palamó s Canyon (NW Front. Mar. Sci. 9:1017052. Mediterranean) were assessed from three instrumented moorings deployed doi: 10.3389/fmars.2022.1017052 in the axis and northern flank of the canyon covering the trawling closure COPYRIGHT © 2022 Paradis, Arjona-Camas, Goñi, (February) and the trawling season (March-December) of 2017. During the Palanques, Masqué and Puig. This is an trawling closure, large sediment fluxes with high contents of labile marine open-access article distributed under organic matter content were registered in the canyon axis, associated to storm the terms of the Creative Commons Attribution License (CC BY). The use, resuspension on the shelf that coincided with dense shelf water cascading and distribution or reproduction in other high surface water productivity. Although no major natural sediment transport forums is permitted, provided the original author(s) and the copyright events occurred during the following spring and summer months, near-daily owner(s) are credited and that the trawling-induced sediment gravity flows were recorded in the northern flank original publication in this journal is mooring, placed directly below a fishing ground, which sometimes reached the cited, in accordance with accepted academic practice. No use, canyon axis. Compositionally, the organic matter transferred by trawling distribution or reproduction is resuspension was impoverished in the most labile biomarkers (fatty acids, permitted which does not comply with these terms. amino acids, and dicarboxylic acids) and had a high degree of degradation, which was similar to surficial sediment from the adjacent fishing ground. Trawling resuspended particles masked the transfer of organic matter enriched in labile biomarkers that naturally occur during the quiescent Frontiers in Marine Science 01 frontiersin.org Paradis et al. 10.3389/fmars.2022.1017052 summer months. Overall, bottom trawling enhances the magnitude of particle fluxes while modifying its organic carbon composition, increasing the re- exposure and transfer of degraded organic carbon and potentially affecting benthic communities that rely on the arrival of fresh organic matter. KEYWORDS submarine canyon, NW Mediterranean, surface sediment, sediment transport, carbon cycle, biomarkers, storms, bottom trawling 1 Introduction submarine canyons (Fabres et al., 2008; Pasqual et al., 2010; Lopez-Fernandez et al., 2013a). Submarine canyons act as preferential conduits of particulate Anthropogenic activities can also modify the transport of matter connecting the productive coastal shelf with the nutrient- sediment in submarine canyons. Several studies have proven that limited abyss (Allen and Durrieu de Madron, 2009; Puig et al., 2014; deep bottom trawling around submarine canyons can be a main Porter et al., 2016). The enhanced flux of organic matter into driver of particulate matter transfer (Martín et al., 2006; Palanques canyons combined with their heterogenous set of habitats makes et al., 2006; Puig et al., 2012; Martı́n et al., 2014a; Wilson et al., these morphological features important nursery and refuge areas for 2015; Arjona-Camas et al., 2019; 2021). This enhanced transferal diverse marine life (Vetter et al., 2010; De Leo et al., 2014; of particulate matter has caused a general increase in Fernandez-Arcaya et al., 2017). Consequently, several benthic sedimentation rates within the axes of several trawled species of commercial interest tend to congregate near submarine submarine canyons (Martı́n et al., 2008; Puig et al., 2015; canyons, supporting many fishing grounds (Farrugio, 2012). Paradis et al., 2017; 2018a; 2018b; 2021a), which has modified The ecological importance of submarine canyons relies on the abundance and biodiversity of meiobenthos (Pusceddu et al., the transfer of organic matter to the canyon’s interior, which is 2014; Romá n et al., 2018). These shifts in meiobenthic community enhanced by natural processes that increase the flux of structures are attributed to a presumed alteration in both the particulate matter into canyons (see review by Puig et al., amount and the composition of organic matter transferred into 2014). These processes include river floods that deliver canyons, since bottom sediments from fishing grounds tend to be elevated amounts of suspended sediment to the margin and depleted in labile organic matter compounds (Martı́n et al., 2014b; high-energy storm-waves, which resuspend sediment from Pusceddu et al., 2014; Paradis et al., 2021b) and the organic matter shallower regions of the shelf. Combined, both processes can associated with local resuspended sediment tends to be highly transport elevated amounts of suspended particulate matter into degraded (Pusceddu et al., 2015; Wilson et al., 2015; Wu et al., canyons as turbidity currents and/or hyperpycnal flows (Martı́n 2016; Daly et al., 2018). However, the organic matter composition et al., 2011; Liu et al., 2012; Puig et al., 2014). Dense shelf water of downward particulate matter transferred antropogenically from cascading (DSWC) events generated by cooling and/or fishing grounds into canyon axes have not been studied yet. evaporation of coastal surface waters can also transport large The NW Mediterranean margin is incised by several canyons amounts of particulate matter into canyons. Their excess in that present large particle fluxes from both natural sediment density forces these waters to sink to greater depths through transport processes and from bottom trawling activities, and submarine canyons as gravity-driven currents that carry Palamó s Canyon (also known as La Fonera Canyon) is one of the resuspended sediment until reaching neutral buoyancy most prominent of these. Northern storms are frequent and (Durrieu de Madron et al., 2005; Canals et al., 2006; Palanques persistent in this margin, but they generate relatively small waves et al., 2012). Particulate fluxes derived from both storms and (~2 m) on the inner shelf that have limited capacity to resuspend DSWC tend to be highest in the upper and middle canyon sediment. In contrast, the rarer eastern storms generally present reaches, and decrease with depth as the hydrodynamic forces larger waves (> 4 m) and are accompanied by heavy rains and subside, rarely reaching the lower canyon (de Stigter et al., 2007; torrential river discharges carrying large amounts of sediment to the Palanques et al., 2012; Pedrosa-Pàmies et al., 2013). Sediment coast. In the case of Palamó s Canyon, the Ter River is the closest to transported during storms and DSWC generally have relatively its canyon head and contributes significant amounts of terrigenous low particulate organic matter content and is composed of aged material to the system (Palanques et al., 2005). These eastern storms terrestrial organic carbon (OC) eroded from the continental also resuspend sediments that are then intercepted by Palamó s shelf (Sanchez-Vidal et al., 2008; Tesi et al., 2010; Martı́n et al., Canyon, supplying large amounts of OC-poor sediment into the 2011; Hsu et al., 2014; Sparkes et al., 2015). However, when these canyon (Martı́n et al., 2006). DSWC in the NW Mediterranean events occur in coincidence with phytoplankton blooms, margin originates in the Gulf of Lions during winter and generates a particulate matter enriched with marine OC is transferred into southward flow of cold dense water that reaches Palamó s Canyon in Frontiers in Marine Science 02 frontiersin.org Paradis et al. 10.3389/fmars.2022.1017052 specific years, supplying large amounts of sediment into it (Ribó canyon axis (Axis-900), at 975 m water depth in the Montgrı́ et al., 2011; Arjona-Camas et al., 2021). tributary gully located ~0.5 km from Sant Sebastià trawling Palamó s Canyon supports important trawling grounds along ground (Flank-1000), and at 1230 m water depth in the its northern and southern flanks, named Sant Sebastià and confluence of the tributary gully and the canyon axis, at ~2 Rostoll fishing grounds, respectively (Figure 1). These fishing km from the same fishing ground (Axis-1200), and ~4 km grounds are concentrated at the 400-800 m depth range in down-canyon from Axis-900 mooring (Figure 1). The accordance to the life-cycle of their most targeted commercial moorings were deployed in three consecutive periods: from species, the highly coveted red-and-blue deep-sea shrimp February 7 to June 5 and from June 9 to October 3 2017, and Aristeus antennatus (Sardà et al., 1994; Tudela et al., 2003). from October 3 2017 to March 5 2018, but this study was Chronic bottom trawling activities along the canyon’s flanks centered on data of 2017 (Figure S1). cause almost daily sediment resuspension evolving into gravity Mooring Axis-900 was equipped with an autonomous flows towards the canyon’s interior (Palanques et al., 2006; Puig hydrographic profiler (Aqualog), as well as a near-bottom et al., 2012; Martín et al., 2014b), which can be detached as autonomous turbidimeter (AQUA-logger 210TY) with a persistent nepheloid layers at the same depth range of the fishing SeaPoint turbidity sensor set to record turbidity in Formazin grounds (Martín et al., 2014b; Arjona-Camas et al., 2021). Turbidity Unit (FTU) at 1-min intervals in auto-gain mode, Trawling-derived sediment gravity flows are then channeled placed at 5 m above the bottom (mab), and a current meter through canyon flank tributaries, such as the Montgrı́ gully in (Aquadopp, Nortek) coupled with a temperature sensor (SeaBird the northern flank (Figure 1), and can reach the canyon axis, SBE-37) set to record at 5-min intervals and placed at 6 mab. transporting eroded sediment with high silt content, as well as The Aqualog only registered data from February 7 to April 7, lower OC and opal content (Martı́n et al., 2006; Palanques et al., and the results are explained in detail in Arjona-Camas et al. 2006). However, these studies have not assessed the variations in (2021). The current meter stopped some days before the end of the composition of downward particulate organic matter the first and second deployments acquiring data from February 7 transported by natural and trawling activities. to May 20 and from June 9 to September 8, respectively, whereas Hence, the aim of this study is to evaluate how sediment it recorded data during all of the third deployment (Figure S1). transport events linked to both natural processes and trawling Mooring Flank-1000 had a string of 5 single-point activities influence the composition of downward particulate autonomous turbidimeters (AQUA logger 210TY, AQUATEC) organic matter into Palamó s Canyon, using biomarkers yielded placed at 5, 10, 20, 50 and 80 mab, each one equipped with from the oxidation of CuO which have been used to assess the temperature, pressure, and a SeaPoint turbidity sensor set to origin and degree of degradation of organic matter transferred into record FTUs at 1-min sampling interval. The mooring line also canyons (Tesi et al., 2008; Goñi et al., 2013; Pasqual et al., 2010; held a downward-looking 300 kHz Teledyne RDI Acoustic 2013). To address this, three instrumented moorings were deployed Doppler Current Profiler (ADCP) placed above the in Palamó s Canyon to measure the sedimentary dynamics and the turbidimeters, which provided current velocity and direction downward particle fluxes and their composition from February to from 10 to 80 mab in 2-m bins at 5-min intervals. The December 2017. Since 2013, a two-month trawling closure occurs in turbidimeter at 20 mab from the first deployment presented Palamó s Canyon from early January to early March to allow the increasing FTUs over time, indicating the effect of biofouling. recruitment of shrimp juveniles and avoid the risk of Hence, data from this sensor were omitted (Figure S1). overexploitation of the fishing stock (Bjørkan et al., 2020), Mooring Axis-1200 was equipped with a single-point providing a time-window to study natural particle fluxes in autonomous turbidimeter (AQUA logger 210TY, AQUATEC) February and the effect of trawling activities for the remainder of set to record at 1-min intervals in auto-gain mode, placed at 5 the year. To identify the alteration of organic matter compounds mab, and a current meter (Aquadopp, Nortek) paired with a transferred into the canyon by natural sediment transport processes conductivity and temperature sensor (SeaBird SBE-37) and a and bottom trawling activities, we compared the sedimentological SeaPoint turbidimeter (0-25 FTU) set to record at 5-min and biogeochemical composition of sediment trap samples along intervals placed at 6 mab. The autonomous turbidity sensor of with the composition of surface sediment obtained on the adjacent the first deployment did not work, so the turbidity data were trawling grounds. extracted from the turbidimeter paired with the current meter, providing values up to 25 FTUs. Unfortunately, the current meter only acquired data from February 7 to March 21 and from 2 Materials and methods June 6 to October 3 (Figure S1). Additionally, this mooring had a 24-cup sediment trap Technicap PPS3/3 model, with a 0.125 m2 2.1 Fieldwork and instrumentation cylindrical opening and an aspect ratio of 2.5 (1 m long/0.4 m wide), placed at 22 mab and programmed to collect downward To conduct this observational study, three moorings were particulate fluxes at 7-day intervals, from Monday to Sunday, to deployed in Palamó s Canyon: at 929 m water depth in the assess the weekly particle fluxes. The sediment trap overfilled in Frontiers in Marine Science 03 frontiersin.org Paradis et al. 10.3389/fmars.2022.1017052 FIGURE 1 Bathymetric map of Palamó s Canyon, with the location of Axis-900 (pink square) mooring at the canyon axis, Flank-1000 mooring (orange triangle) deployed in the Montgrı ́ tributary gully (MT) and Axis-1200 mooring (blue circle) deployed in the confluence between MT and the canyon axis. The yellow rectangle above the Flank-1000 mooring delimits the section of the fishing ground where daily hauls were extracted for Figure 2. Composition of the sediment trap (Axis-1200) was compared with surficial composition of the sediment core collected in the trawled flank (green cross). Fishing effort is given as cumulative number of hauls per hectare obtained from AIS data from February to December 2017, identifying the main fishing grounds (Sant Sebastià, Rostoll, Llevant). The location of the Cap Begur buoy is given by a yellow star. the 4th sample of the first deployment (March 6-13) and in the components were reported with respect to the main canyon 8th sample of the third deployment (December 4-11), providing axis orientation: along-canyon (i.e. across-gully for Flank-1000) valid samples only from February 13 to March 9 and from June and across-canyon (i.e. along-gully for Flank-1000). FTUs from 12 to December 7 (Figure S1). the different turbidimeters were converted to estimates of Current speeds and directions from the different current suspended sediment concentration (SSC; mg·L-1) following the meters were decomposed in E-W and N-S current velocity calibration described in Arjona-Camas et al. (2021). Cumulative components. In order to obtain the along- and across-canyon sediment transport (kg·m-2) in the across- and along-canyon directions for the Axis-900 and Axis-1200 moorings, data from direction were calculated through the cumulative sum of the these moorings were rotated 70° clockwise to align their product of the decomposed and rotated current speed and the reference frames to the canyon axis orientation, converting the SSC. Given the maximum acquisition of turbidity values of Axis- N-S velocity component into the along-canyon velocity 1200 in the first deployment and the missing data between component (positive up-canyon) and the E-W velocity mooring deployments, cumulative suspended sediment component into the across-canyon velocity component transport must be regarded as lower estimates. (positive towards the NNE). Data of the Flank-1000 mooring In addition, three sediment cores were taken on the trawled were rotated 11° counterclockwise to align the reference frame to northern flank of Palamó s Canyon (Figure 1), at the time of the the canyon gully orientation, converting the N-S velocity first deployment and during the mooring turnaround cruises: component into the along-gully velocity component (positive one during the temporal trawling close season in February 7, and up-gully), and the E-W velocity component into the across-gully two during the trawling season, in June 5 and in October 3. The velocity component (positive towards ESE). To make the three cores were collected using a K/C Denmark multicorer at ∼500-m moorings comparable among them, all current direction depth (see Paradis et al. (2021b) for further details). Frontiers in Marine Science 04 frontiersin.org Paradis et al. 10.3389/fmars.2022.1017052 2.2 Sediment sample treatment and cutin acids, and certain p-hydroxybenzenes), soil (certain benzoic analytical procedure acids), as well as compound classes (fatty-acids, di-carboxylic acids, amino acids) that can be of both marine and terrestrial origin (Table Sediment trap samples were prepared and processed in the S1). Mean concentrations of the parameters were calculated as flux- laboratory using the same method as described in Palanques and weighted mean concentrations over the whole sampling period, as Puig (2018), obtaining sub-sample aliquots for each sample. explained by Heussner et al. (1990). Total mass flux (TMF; g·m-2·d-1) was calculated from the sample Sediment samples from the sediment cores taken on the dry weight of two sub-sample aliquots, which were averaged and trawled northern flank of Palamó s Canyon during the cruises upscaled based on their dilution factor, and divided by the were analyzed for grain size, OC, TN, CaCO3, and CuO collecting area (0.125 m2) and the sampling interval (7 days) biomarkers, following the same procedure as the sediment trap of the sediment trap, with an average standard error of 2.4%. samples. The procedures and the results of these analysis were Average grain size and the fraction of clay (< 4 µm), published in Paradis et al. (2021b). For this study we use the silt (4 – 63 µm), and sand (63 µm – 2 mm) in downward results of the surface sediment samples (0-1 cm) of these cores to particulate fluxes were analyzed using a Horiba Partica LA- compare them with the results of the sediment trap samples. 950 V2m particle-size analyzer by oxidizing approximately 1 g of sample using 20% H 2 O 2 and disaggregating with 2.5% P2 O4− 7 . 2.3 Forcing conditions Elemental analysis of total carbon (TC), organic carbon (OC) and total nitrogen (TN) contents were carried out with an elemental Significant wave height (Hs) was obtained from the Cap de analyzer (Costech ECS Analyzer 4010), according to the procedure Begur buoy (3.65° E, 41.92° N), located ~10 km NE at the described by Nieuwenhuize et al. (1994). Prior to OC analysis, continental slope region next to Palamó s Canyon (Figure 1), samples were first decarbonated through acid-fuming in the which is managed by the Spanish Ports Authority. Storms were presence of 12 N HCl during 24 h and repeatedly adding 100 µL defined as Hs greater than 2 m for more than 6 h (Mendoza and of 2 N HCl to the sample until effervescence ceased. Inorganic Jimé nez, 2009) and eastern storms were identified when the carbon, quantified as the difference between total carbon and wave direction was between 60° and 120°. Daily Ter River organic carbon, was converted to CaCO3 contents using the discharge was obtained from the Catalan Water Agency. Data molecular mass ratio of 100/12, assuming all inorganic carbon of net primary productivity (NPP) were extracted from the present is in the form of CaCO3. Opal (biogenic silica) was analyzed Ocean Productivity database (http://www.science.oregonstate. using a wet-alkaline extraction with sodium carbonate using the edu/ocean.productivity/index.php). NPP data from this method described by Mortlock and Froelich (1989). Finally, the database are calculated using the Vertically Generalized lithogenic fraction was calculated as the difference between the total Production Model, which uses both surface chlorophyll mass and the main mass constituents (CaCO3, organic matter, and concentrations and sea surface temperature data extracted opal), where organic matter content was estimated as twice the from satellite imagery (Behrenfeld and Falkowski, 1997). OC content. Trawling activity was obtained from vessel Automatic Specific biomarkers yielded from the CuO oxidation of Identification System (AIS) extracted from Shiplocus, the official sediment trap samples were analyzed following the method Spanish Port System module that manages and evaluates maritime described by Goñi and Montgomery (2000). Briefly, between 250 traffic data. AIS data were filtered following the method described and 400 mg of homogenized ground sample was oxidized with CuO by Paradis et al. (2021b) to obtain the number of hauls per hectare under basic conditions (2 N NaOH) in an oxygen-free atmosphere in each fishing ground. The number of hauls above Montgrı́ at 150°C for 90 min in a microwave. After oxidation, samples were tributary gully was then extracted from the polygon delimited spiked with ethyl vanillin and trans-cinnamic acid as recovery in Figure 1. standards, after which the samples were acidified to 1 pH using small volumes of concentrated HCl, and subsequently extracted twice using ethyl acetate. Samples were then evaporated under a constant flow of nitrogen gas, reconditioned in pyridine, and 2.4 Statistical analyses derivatized by adding bistrimethylsilyltrifluoroacetamide (BSTFA) + 10% trimethylchlorosilane (TMCS). Specific reaction products Principal Component Analysis (PCA) was performed on CuO (Table S1) were quantified by GC-MS (Hewlett Packard 6890 GC biomarkers of downward particulate fluxes obtained from the System) with DB1 column and the HP5973 mass selective detector sediment trap (Axis-1200) to reduce the dimensions of the through selective ion monitoring as previously described by Goni dataset and maximize the variance between samples. PCA was et al. (2009). CuO yields were grouped into specific compound performed using Python’s Scikit-learn module (Pedregosa et al., classes that have distinct chemical precursors including those from 2011) by first normalizing the data. Based on this PCA, PC loadings specific terrestrial sources such as vascular plants (lignin phenols, of OC composition of surface sediment collected in the adjacent Frontiers in Marine Science 05 frontiersin.org Paradis et al. 10.3389/fmars.2022.1017052 trawling ground (Paradis et al., 2021b) were used to study its NPP also followed this general seasonal pattern, values were low relationship with the composition of downward particulate fluxes. during the first eastern storm in January (550 mg C·m-2·d-1) but increased during the second eastern storm in February due to a late- winter phytoplankton bloom (1200 mg C·m-2·d-1) (Figure 2C). 3 Results During the more quiescent late-spring and summer months, these natural forcing conditions receded to background values 3.1 Forcing conditions (~1 m Hs, ~3 m3·s-1 Ter River discharge, and ~450 mg C·m-2·d-1 NPP) (Figures 2A–C). The general decrease of NPP was During the sampling period, natural forcing conditions (Hs, Ter interrupted by a second phytoplankton bloom in early-June River discharge, and NPP) followed a similar seasonal pattern, with (870 mg C·m-2·d-1) (Figure 2C). generally stronger events occurring during the winter (January- From October to December, Hs increased again due to March) and autumn (October-December) months (Figures 2A– seasonal autumn storms (Figure 2A). These events were C). Although the majority of the storms were caused by northern generally dry storms caused by northern winds that did not lead winds, exceptionally strong eastern storms occurred in January 19- to significant increases in Ter River discharges (Figures 2A, B). 23 and in February 12-15 with Hs reaching 6 m and 5 m respectively, NPP steadily increased during autumn until reaching ~780 mg and strong Ter River floods, reaching 133 m3·s-1 and 30 m3·s-1 C·m-2·d-1 in late-November and decreased again to background respectively during each eastern storm (Figures 2A, B). Although values in the following winter months (Figure 2C). A B C D FIGURE 2 Forcing conditions in 2017. (A) Significant wave height (Hs) measured at the Cap de Begur buoy, where eastern storms are identified in red. (B) Ter River discharge. (C) 8-day average net primary productivity (NPP). (D) Daily hauls in the trawling ground above the Montgrı ́ tributary gully (see yellow polygon in Figure 1). The horizontal arrow on top indicates the monitoring period. Frontiers in Marine Science 06 frontiersin.org Paradis et al. 10.3389/fmars.2022.1017052 Trawling effort in the Sant Sebastià fishing ground above the particle fluxes. Temperature recorded at 6 mab fluctuated Montgrı́ tributary gully varied throughout the year (Figure 2D). No between 13.2 and 13.6°C throughout the year, with a sharp drop hauls were detected at the beginning of the year during the seasonal to 12.6°C between February 13 and 15 (Figure 3A) due to an eastern trawling closure, nor during the weekends or national holidays, storm concurrent with a DSWC event that caused a maximum peak when trawlers are not allowed to operate (Bjørkan et al., 2020). in SSC of ~230 mg·L-1. This event was initially directed down- During working days, an average of 2 hauls per day were recorded, canyon at 0.3 m·s-1, but was followed by a quick up-canyon reversal but trawling effort varied from 0 to 12 hauls per working day. of current direction reaching maximum values of 0.6 m·s-1 Highest fishing effort occurred at the beginning of the trawling (Figure 3D), following the predominant up-canyon bottom season in March, in late August, and again in late December, with 7 current direction (~280°) at this site (Figure 3E). to 12 hauls per day (Figure 2D). During late October and throughout Since the onset of trawling activities in March, several sporadic November, there was negligible trawling activity on the Sant Sebastià peaks of ~20 mg·L-1 were recorded only during weekdays, with fishing ground (Figure 2D) since bottom trawlers had moved to exceptionally higher SSC of 190, 140, and 120 mg·L-1 at the end of shallower fishing grounds on the open slope (Figure S2). July, in mid-August and at the end of September (Figure 3B). SSC was negligible during November, but small peaks of SSC (~60 mg·L- 1 ) resumed again in mid-December (Figure 3B). Current velocities 3.2 Sediment transport in canyon axis presented fluctuating up-canyon and down-canyon flows (~0.15 (929 m water depth) m·s-1), with increasing speeds of up to 0.4 m·s-1 towards the end of the year in November and December (Figure 3D). The The time series of the Axis-900 mooring located in the canyon predominant direction of near-bottom currents was up-canyon axis are presented in detail in Arjona-Camas et al. (2021), but (~280°), although currents came from the northern flank during described here briefly to contextualize the observed downward some high SSC events (Figure 3E). A B C E D FIGURE 3 Axis-900 time-series of (A) temperature at 6 mab, (B) suspended sediment concentration (SSC), (C) across-canyon current velocity, and (D) along-canyon current velocity. (E) Polar plot of non-rotated current direction (angle), current speed (radius) and turbidity (color bar) at 5 mab. The inverted triangle on top indicates the occurrence of a dense shelf water cascading event. Note the different axis of along canyon and across canyon current speeds. Frontiers in Marine Science 07 frontiersin.org Paradis et al. 10.3389/fmars.2022.1017052 3.3 Sediment transport in the Montgrı ́ events generally decreased throughout the year to SSC of ~50 mg·L-1 tributary (975 m water depth) and were limited to between 5-20 mab (Figure 4B). Despite this general decreasing trend, exceptionally high sporadic near-bottom In the Flank-1000 mooring located in the Montgrı́ tributary SSC of 250-770 mg·L-1 were observed in summer months that were gully, temperatures at 5 mab fluctuated between 13.2 and 13.6°C accompanied with strong down-flank current speeds of between during the whole sampling period, with no apparent change 0.25 m·s-1 to 0.6 m·s-1 (Figures 4C, D). Sediment gravity flows (Figure 4A). In contrast, recordings of SSC, current speed, and through the Montgrı́ tributary gully were absent in late October and direction indicate frequent sediment transport events into the throughout November when bottom trawlers moved to the open canyon’s interior that follow a distinct temporal trend slope (Figure S2), but resumed again in December, with near- (Figures 4B-D). During the trawling closure in February, SSC bottom SSC between 100 and 260 mg·L-1 and current speeds of ~0.2 was negligible at all depths, aside for a small near-bottom (5-10 m·s-1 (Figures 4B-D). mab) SSC peak of ~60 mg·L-1 as well as a sharp off-shore and down-flank increase in current speed of 0.4 m·s-1 and 0.2 m·s-1, respectively, that occurred in coincidence with the eastern storm 3.4 Sediment transport in canyon axis and DSWC event of mid-February (Figures 4B-D). (1230 m water depth) At the onset of the trawling period in March, peaks of near- bottom SSC (~400 mg·L-1) and current speeds (~0.4 m·s-1) caused In the Axis-1200 mooring, temperature remained relatively by trawling-induced sediment gravity flows were recorded down- constant, fluctuating between 13.1-13.4°C (Figure 5A) and did flank through the tributary (~190°; Figures 4B-D) on a daily basis, not present the sharp drop in temperature caused by the storm only during weekdays. The magnitude of these sediment transport and DSWC event observed in the mooring further up-canyon A B C E D FIGURE 4 Flank-1000 time-series of (A) temperature at 5 mab, (B) suspended sediment concentration (SSC) at different water depths, (C) across-canyon current velocity at 10-80 mab, and (D) along-canyon current velocity at 10-80 mab. (E) Polar plot of non-rotated current direction (angle), current speed (radius) and turbidity (color bar) at 10 mab. The inverted triangle on top indicates the occurrence of a dense shelf water cascading event. Frontiers in Marine Science 08 frontiersin.org Paradis et al. 10.3389/fmars.2022.1017052 (Figure 3A). Instead, temperature presented a slight increase and came from the Montgrı́ tributary gully (Figures 5C, E), from 13.2 °C to 13.4 °C at the beginning of the event (Figure 5A) occurring in coincidence to the sediment gravity flows recorded concurrent with an increase of near-bottom SSC >28.5 mg·L-1, in the Flank-1000 mooring that had SSC exceeding 250 mg·L-1 above the sensor range, and a peak of near-bottom current speed (Figure 4). The remaining sediment gravity flows were directed of 0.66 m·s-1 directed across-canyon and coming from the towards the SSW and came from an adjacent tributary gully to the northern flank (Figures 5B, C). One day later, on February 15, Montgrı́, or towards the SE, coming from the upper canyon axis a second peak of SSC >28.5 mg·L-1 was recorded, coinciding with (Figure 5E). A strong sediment gravity flow was again registered in another strong bottom current speed that reached 0.66 m·s-1 October 24, reaching high near-bottom SSC of 135 mg·L-1 coupled directed down-canyon (Figures 5B, D). A second sharp increase with strong down-canyon (0.25 m·s-1) and NNW (0.15 m·s-1) in SSC occurred on February 27, reaching the maximum bottom currents (Figures 5B, D). Aside from the down-flank turbidity reading values of >28.5 mg·L-1, with strong across- sediment gravity flows registered during the summer, the canyon bottom current speed of 0.58 m·s-1 coming from the predominant bottom currents during the sampling period were northern flank (Figures 5B, C). At the onset of bottom trawling essentially directed down-canyon (Figure 5E). in March, near-bottom SSC presented sharp increases of ~20 mg·L-1 only during weekdays (Figures 5A, B). Unfortunately, sensors in this first deployment stopped working on March 21. 3.5 Downward particle fluxes and During the second deployment in summer, frequent increases composition in canyon axis (1230 m in SSC from ~1 mg·L-1 to between 60 and 230 mg·L-1 were recorded water depth) from June to early September, accompanied by increases in across- canyon current speed from 0.1 m·s-1 to 0.45 m·s-1 (Figures 5B–D). Downward TMF was highly variable during the whole Some of these sediment gravity flows were directed towards the S sampling period, averaging 45.9 g·m-2·d-1 but ranging from 7.4 A B C E D FIGURE 5 Axis-1200 time-series of (A) temperature, (B) suspended sediment concentration (SSC), (C) across-canyon current velocity, and (D) along- canyon current velocity. Note the maximum data acquisition for SSC in the first deployment. (E) Polar plot of non-rotated current direction (angle), current speed (radius) and turbidity (color bar) at 5 mab. The inverted triangle on top indicates the occurrence of a dense shelf water cascading event. Frontiers in Marine Science 09 frontiersin.org Paradis et al. 10.3389/fmars.2022.1017052 g·m-2·d-1 to >140 g·m-2·d-1 (Figure 6A). Due to the lack of data (SP) displayed low concentrations during the February storm between March and June, a complete analysis of seasonal and DSWC event (44 and 40 µg·100 mg-1 OC, respectively), but variability is not possible. At the beginning of the sampling reached maximum concentrations in the overfilled sample in period, TMF values increased from 29 g·m-2·d-1 in mid-February March at the beginning of the trawling season (65 and 50 µg·100 to >140 g·m-2·d-1 when the sediment trap overfilled in early mg-1 OC, respectively), with values similar to those of the surface March, at the onset of the bottom trawling season. Downward sediment from the adjacent trawled flank (Figure 7A). In TMF presented sporadic increases from mid-June to late- contrast, terrigenous cinnamyl phenols (CP) and cutin acids September that ranged from approximately 50 g·m-2·d-1 to 125 (CA) presented relatively constant concentrations during this g·m-2·d-1, which occurred in coincidence with sediment gravity period (18 and 59 µg·100 mg-1 OC, respectively). During the flows coming from the northern flank (Figures 5A, 6A). Lower summer, concentrations of terrestrial compounds were similar TMF values of approximately 14 g·m-2·d-1 were generally to those quantified in the surficial sediment of the core collected observed from October to early December, interrupted by a in the adjacent trawling flank, although small sporadic increases high TMF in late October (55 g·m-2·d-1) in coincidence with the occurred in mid-July and early September (Figures 7A, B). Small sporadic sediment gravity flow coming from the southern increases in these compounds were also detected during the late- canyon flank. The sediment trap overfilled in early December October sediment gravity flow, as well as the overfilled sediment (>93 g·m-2·d-1), which resulted in an overfilled sample cup and trap in early-December. Ratios of SP/VP and CP/VP in the the clogging of the sediment trap funnel. sediment trap ranged from 0.6 to 1.1 and from 0.2 to 0.4, The flux weighed average of the mean grain size was 12.8 µm respectively, whereas ratios of acid to aldehyde of vanillyl during the whole sampling period, and ranged from 7.7 µm to phenols (Vd/Vl) and syringyl phenols (Sd/Sl) ranged from 0.4 20.0 µm, generally following the patterns of TMF (Figure 6A). to 1.1 and from 0.2 to 0.8, respectively (Figure S3). Sediment grain size consisted of mostly silt (71%) and clay (27%) From the heterogenous compound classes, p- fractions (Figure 6B), with slight increases in sand content from hydroxybenzenes (PB) and benzoic acids (BA) presented a 0.5 to 6.5% during periods with highest TMF (> 40 g·m-2·d-1) similar pattern as VP and SP, with low concentrations during (Figures 6A, B). The highest sand contents of 6.5% and 5.2% were the February storm and DSWC event (31 and 9 µg·100 mg-1 OC, detected in the overfilled sediment trap samples from March and respectively) that peaked with the overfilling of the sediment trap December, respectively (Figure 6C). In comparison to surficial in March (51 and 22 µg·100 mg-1 OC, respectively), reaching sediment from the adjacent trawled flank, sediment trap samples similar values as those from the surface sediment sampled in the had finer grain size (Figures 6A, B). The predominant major adjacent trawled area (Figure 7C). Concentrations of PB and BA constituents of the downward particle fluxes were lithogenic and were initially high in early-summer (64 and 27 µg·100 mg-1 OC, CaCO3 fractions with relatively constant values (average 70% and respectively) and then decreased to similar values as those 26% respectively). Downward particle fluxes presented similar observed in the surficial sediment of the trawled site CaCO3 contents as surficial sediment in the sediment core (Figure 7C). Towards the end of summer, concentrations of PB collected in the adjacent trawled site (Figure 6C). and BA increased again to 64 and 22 µg·100 mg-1 OC, respectively, The flux weighed mean of OC and TN contents was 0.91% but dropped to 35 and 8 µg·100 mg-1 OC, respectively, during the and 0.11%, respectively, and followed a clear seasonal pattern late-October sediment gravity flow (Figure 7C). with high contents at the beginning of the year (0.99% OC, The fatty acid (FA), dicarboxylic acid (DA) and amino acid 0.12% TN) that generally decreased in late-spring and summer (AA) CuO products are highly enriched in marine organic to minimum values (0.86% OC, 0.10% TN), and increased again matter sources (e.g., phytoplankton, zooplankton) relative to in autumn towards the end of the year (1.05% OC, 0.13% TN) terrestrial plants (Goñi and Hedges, 1995). Given their high (Figure 6D). The OC/TN molar ratio had a flux weighted mean degradability, these compounds are rapidly transformed of 8.2, and ranged from 7.2 to 10.2, with highest values occurring during their transport towards the seafloor, so presence of during high TMF in the summer (Figure 6E). Concentrations of these compounds in downward particulate fluxes generally OC and TN in the sediment trap samples during the winter indicate the arrival of fresh marine organic matter, often months were 45 and 57% higher, respectively, than those derived from periods of high net primary productivity (Goni quantified in the surficial sediment of the trawled sediment et al., 2009). These compounds presented higher OC- core, but presented similar values during the summer normalized concentrations during the mid-February eastern months (Figure 6D). storm and DSWC event (89, 65 and 568 µg·100 mg-1 OC) and The composition of organic matter was investigated by decreased during the following weeks (72, 52 and 390 µg·100 calculating the OC-normalized contents of different types of mg-1 OC) (Figures 7D, E). However, concentrations increased biomarkers (in µg of biomarker per 100 mg of OC), which again in the overfilled sediment trap in March. The summer showed considerable variations throughout the sampling period months generally presented lower concentrations (~55, ~40 (Figure 7). Terrestrial vanillyl phenols (VP) and syringyl phenols and ~245 µg·100 mg-1 OC, respectively), similar to the surficial Frontiers in Marine Science 10 frontiersin.org Paradis et al. 10.3389/fmars.2022.1017052 sediment of the core collected in the trawled flank. The variations in OC composition in downward particulate Exceptionally high concentrations (103, 78 and 621 µg·100 fluxes are summarized in a PCA, which in total explained 74% of mg-1 OC) were observed in early June, as observed with PB and the variance (Figure 8). Eigenvalue scores show that higher PC1 BA, as well as in mid-July and mid-September. Concentrations values reflect greater concentrations of all CuO compounds, remained low and similar to the surface sediment of the consistent with the presence of organic matter with higher trawled area during the remaining sampling period, with overall biomarker abundances, whereas different groups of another increase in concentration in the overfilled sediment CuO compounds are differentiated along the PC2 axis in trap in early December (68, 39 and 380 µg·100 mg-1 OC) terms of their general reactivity, as observed by Paradis et al. (Figures 7D, E). (2021b) (Figure 8A). With the exception of the last sediment A B C D E FIGURE 6 Axis-1200 time series of the sediment trap: (A) total mass flux (TMF) in grey bars and mean grain size in blue circles, (B) grain size classes, (C) relative concentration of the main mass composition: lithogenics (brown diamonds), calcium carbonate (CaCO3, white circles), organic matter (OM, pink squares), and biogenic opal (green triangles), (D) organic carbon (OC) in blue and total nitrogen (TN) in green, and (E) OC/TN molar ratio. Question marks above TMF bars indicate that the sediment trap had overfilled in that sample and the inverted triangle indicates the occurrence of a dense shelf water cascading event. Mean values of grain size, CaCO3, OC, and TN of surficial sediment of the sediment core retrieved on the trawled flank in February, June, and October (Paradis et al., 2021b) are given with crosses. Note the break in the axis of the grain size fraction (B). Frontiers in Marine Science 11 frontiersin.org Paradis et al. 10.3389/fmars.2022.1017052 A B C D E FIGURE 7 Temporal evolution of OC-normalized CuO compound classes in the sediment trap samples (Axis-1200). (A) terrestrial vanillyl (blue triangles) and syringyl (green circles) phenols, (B) terrestrial cinnamyl phenols (blue triangles) and cutin acids (green circles), (C) p-hydroxy benzenes (blue triangles) and benzoic acids (green circles), (D) fatty acids (blue triangles) and di-carboxylic acids (green circles), (E) amino acids (blue triangles). Values of surficial sediment of the sediment core retrieved on the trawled flank in February, June, and October (Paradis et al., 2021b) are given with crosses. Note the different scales for each compound classes. The inverted triangle on top indicates the occurrence of a dense shelf water cascading event. trap sample from early-March, the temporal evolution of PC1 4 Discussion generally reflects higher values during the trawling closure than during the trawling season (Figure 8B). In contrast, during the During the 2017 sampling period, several sediment transport trawling season, higher TMF values are normally associated to events of both natural and anthropic origin were identified in lower PC1 and PC2 values, similar to values of surficial samples Palamó s Canyon. Storms occurred during winter and autumn from the adjacent trawling ground (Figures 8B, C). months, characteristic of seasonal patterns of this margin Frontiers in Marine Science 12 frontiersin.org Paradis et al. 10.3389/fmars.2022.1017052 (Palanques et al., 2005; Zú ñiga et al., 2009; Lopez-Fernandez turbidity (~230 mg·L -1 ) and current speed (0.6 m·s -1 ) et al., 2013b) with exceptionally strong eastern storms in mid- (Figures 3A, B; Arjona-Camas et al., 2021). No drop of January and mid-February (Figures 2A, B). These natural temperatures was observed either in Flank-1000 nor in Axis- forcing conditions abated during the spring and summer 1200, the latter mooring located ~4 km down-canyon from Axis- months, when bottom trawling activities were predominant in 900, indicating that these dense waters had reached neutral the surroundings of Palamó s Canyon (Figure 2; Palanques et al., buoyancy at shallower depths. However, this DSWC event 2006; Puig et al., 2012; Martı́n et al., 2014a). Both storms and prompted the transport of suspended sediment through the trawling activities transported sediment into the canyon through entire canyon. The Flank-1000 mooring registered SSC of ~50 different pathways and with different sediment composition. The mg·L-1 and current speeds of 0.4 m s-1, while the Axis-1200 following sections describe and discuss in detail each of these mooring registered SSC >28.5 mg·L-1 and current speeds of events chronologically. 0.6 m s-1 (Figure S4). However, the turbidimeter at Axis-1200 recorded maximum values during that event due to the sensor range limitation (28.5 mg·L-1), and we expect that actual SSC 4.1 Winter storm and dense shelf water were significantly higher given the simultaneous increases in SSC cascading (February) in all moorings (Figure S4). The strong sediment transport event associated to DSWC in An exceptionally strong eastern storm occurred a week prior the Axis-1200 mooring did not lead to high downward sediment to the mooring deployment in mid-January, causing a flash flood fluxes in the sediment trap (Figure 6A), which could indicate of the Ter River (Figures 2A, B). This flood event transported either a stronger horizontal rather than vertical sediment flux terrestrial organic matter to the margin and reached the that would have bypassed the sediment trap, or that the sediment northern flank of Palamó s Canyon (Figure 1), which we transport occurred close to the seafloor and was missed by the sampled two weeks after this storm. The surface sediment sediment trap located at 22 mab, as hypothesized in other studies recovered from this core had high concentrations of lignin (Bonnin et al., 2008; Rumı́n-Caparró s et al., 2016). Moreover, phenols and cutin acids from vascular plants, reflecting fresh high current speeds can affect sediment trap collection efficiency terrestrial inputs from the storm (Figures 7A, B; Paradis (Baker et al., 1988; Heussner et al., 2006), so downward et al., 2021b). particulate fluxes during this event should be considered as This storm was followed by another eastern storm a week semi-quantitative. after the deployment of the moorings in mid-February that Downward sediment fluxes from the mid-February eastern drove the cascading of dense shelf water into the upper Palamó s storm and DSWC event had higher concentrations of OC and Canyon axis, reaching the Axis-900 mooring, which registered a TN (1.1% and 0.14%, respectively) (Figure 6D) enriched in drop in temperatures (< 12.6°C), as well as an increase in compound classes of marine origin (i.e. fatty acids, A B C FIGURE 8 PCA results. (A) Eigenvalues of the PCA performed of OC-normalized CuO products in the sediment trap (Axis-1200). Temporal evolution of (B) PC1 and (C) PC2 values of the sediment trap samples, where the size of the circles refers to TMF and the green crosses show the PC value of surficial sediment of the sediment core retrieved on the trawled flank in February, June, and October (Paradis et al., 2021b). The inverted triangle on top indicates the occurrence of a dense shelf water cascading event. Frontiers in Marine Science 13 frontiersin.org Paradis et al. 10.3389/fmars.2022.1017052 dicarboxylic acids, and amino acids) (Figures 7D, E) in induced sediment gravity flows from a sediment trap deployed in comparison to particulate matter accumulated during the this same site in 2002 (Palanques et al., 2006). following week (Figure S4). This peak in marine biomarkers Downward sediment flux during the onset of the trawling coincided with elevated estimates of NPP that were consistent season was enriched in most biomarkers, including terrestrial with a winter phytoplankton bloom. Although this storm caused compounds (lignin phenols) and labile marine compounds (fatty high Ter River discharge (Figure 2B) that reached off-shore acids, di-carboxylic acids, and amino acids) in comparison to the sediments on the Palamó s northern flank (Paradis et al., 2021b), previous trap sample (Figures 7, S5), with a similar overall the organic matter in the sediment traps had low OC- composition to surficial sediment from the trawled flank normalized concentrations of terrestrial compounds because (Figure 8; Paradis et al., 2021b). This confirms that bottom they were diluted by the enhanced contribution of fresh trawling is translocating sediment from the fishing grounds into marine organic compounds from the phytoplankton bloom the canyon. The higher terrestrial content was associated to the (Figure 2). Other studies have shown similar trends, where the transfer of surficial sediment from the fishing ground which was usual transfer of OC-poor terrigenous organic matter into enriched in these compounds due to the Ter River flash flood canyons during storms (Tesi et al., 2010; Sanchez-Vidal et al., that occurred in January (Paradis et al., 2021b), while the higher 2012) is offset by the transfer of fresh marine organic matter marine compounds indicate the transferal of the remnants of the during phytoplankton blooms (Fabres et al., 2008; Pasqual et al., phytoplankton bloom (see section 4.1). 2010; Lopez-Fernandez et al., 2013a). Hence, these high-energy DSWC events can result in important drivers of fresh and labile organic matter from surface shelf and slope waters to deep-sea 4.3 Trawling season (March-October) environments, while also transferring fluvial derived terrestrial material. Previous studies observed that bottom trawling during the During the following 2 weeks, downward sediment fluxes in summer causes almost-daily sediment gravity flows through the the sediment trap increased from ~30 to ~100 g·m-2·d-1 Montgrı́ tributary only during working days (Palanques et al., (Figure 6A), indicating the settling of sediment that had 2006; Puig et al., 2012; Martín et al., 2014a), which were also remained in suspension after the DSWC event (Arjona-Camas evident during this study (Figures S5, S6). The intensity of these et al., 2021). These samples displayed lower OC and TN contents sediment gravity flows in Flank-1000 mooring decreased as the and decreasing OC-normalized concentrations of the most labile trawling season unfolded, from ~500 mg·L-1 in early March to compounds (i.e. amino acids, fatty acids, dicarboxylic acids; ~15 mg·L-1 in late October, although some large isolated peaks Figures 7, 8), which indicate the resuspension and transfer of were still registered (Figure 4A). This decreasing trend proves a OC-poor sediment resuspended by the strong DSWC currents progressive depletion of sediment available for resuspension by (Puig et al., 2008). Alternatively, these trends could also reflect a bottom trawling gears during the trawling season, as general degradation of these compounds given their long transit hypothesized in previous studies (Martı́n et al., 2014a). time in the water column before reaching the sediment trap. Although these sediment gravity flows were directed towards the canyon axis, only the exceptionally intense sediment gravity flows (SSC >250 mg·L-1) at Flank-1000 reached the Axis-1200 4.2 Onset of trawling season (March) mooring, causing SSC and TMF to increase from ~1 to 60-230 mg·L-1 and from ~16 to 120 g·m-2·d-1 respectively (Figure S6), with Bottom trawling activities began on March 8, after 2 months similar magnitude as those registered during the summer months at of a trawling closure on the fishing grounds surrounding the same site in 2001 (Martı́n et al., 2006). These sediment gravity Palamó s Canyon (BOE, 2017). At the onset of bottom trawling flows mainly originated from the Montgrı́ tributary gully (~190°), activities, high sediment fluxes were recorded at the Axis-900 but also from an adjacent tributary gully towards the NNE (~220°) mooring (Arjona-Camas et al., 2021), and sediment gravity flows (Figure S6), as hypothesized in previous studies (Martı́n et al., 2006; with high SSC (200-400 mg·L-1) and strong bottom currents Palanques et al., 2006). This proves that the various tributary gullies (~0.4 m·s-1) were channeled through the Montgrı́ tributary gully, along the flanks of Palamó s Canyon (Lastras et al., 2011) would recorded by the Flank-1000 mooring, some reaching the Axis- likely serve as pathways of resuspended sediment from the fishing 1200 mooring ~1.5 km further down-flank (Figure S5). The grounds along the canyon flanks into the canyon axis (Figure 1), strong sediment gravity flows occurring on Thursday and Friday presumably leading to increases in sedimentation rates not only at (March 10 and 11) that reached Axis-1200 with SSC between 10- 1750 m in depth, as reported in previous studies (Martı́n et al., 2008; 20 mg·L-1 would have caused the overfilling of that week’s Puig et al., 2015), but likely throughout the whole canyon axis, as sediment trap cup, with a total mass flux of >140 g·m-2·d-1 observed in the adjacent Blanes Canyon (Paradis et al., 2018b). (Figure S5). Sediment transported at the onset of the trawling Trawling-derived sediment gravity flows transported coarser season had coarser grain size with higher sand content than the grain sizes in comparison to periods with naturally low TMFs previous trap samples (Figures 6B, S5), as observed in trawling- (Figure S6), as observed by Palanques et al. (2006). However, Frontiers in Marine Science 14 frontiersin.org Paradis et al. 10.3389/fmars.2022.1017052 downward particulate matter collected by the sediment trap had (Figure 2C) and the lower dilution of organic matter due to finer grain size than surface sediment of trawling grounds lower lithogenic transport during this period in comparison to (Figures 7A, B) as a result of grain size sorting and winnowing of periods with trawling-derived sediment gravity flows. This is finer grained sediment by bottom trawling (Martı́n et al., 2014b; supported by the more labile organic matter during this period Paradis et al., 2021b). The flux weighed mean concentration of OC (Figures 7, S7), indicating the transferal of fresh material into the and TN during this period (March-October) was lower (0.84% OC, canyon in the absence of trawling activities. 0.10% TN) than during the rest of the study period (1.00% OC, An isolated late-October sediment gravity flow coming from 0.13% TN) (Figures 6D, S6), which also coincided with the periods the southern flank interrupted the low TMF during the absence with lower net primary productivity in the area (Figure 2C). This of trawling activities, which led to high TMF of ~60 g·m-2·d-1 could indicate that marine productivity is the main driver increasing with high lithogenic fraction that diluted the OC contents OC and TN contents sinking to the seafloor, where submarine (Figure S7). In contrast to trawling-derived sediment gravity canyons play an important role in its transfer to deeper flows coming from the northern flank (see section 4.3), organic environments (Fabres et al., 2008; Lopez-Fernandez et al., 2013a; matter presented higher lability (Figure 8C), indicating the Grinyó et al., 2017), and that bottom trawling does not alter the OC transferal of fresh material, albeit diluted by the high and TN content of the downward particulate matter fluxes. lithogenic fraction, into the canyon. Therefore, this sporadic However, the composition of OC obtained from CuO oxidation sediment gravity flow seems to be associated to a natural indicates that periods with high downward particulate fluxes due to sediment destabilization from the untrawled southern canyon bottom trawling-derived sediment gravity flows transferred highly flank, as previously observed in deeper parts of this canyon degraded terrestrial compounds and lower content of the most (Martín et al., 2006; 2007). labile marine compound classes, and its biogeochemical After five weeks of trawling intermission in Sant Sebastià composition was similar to surficial sediment from the adjacent fishing ground, bottom trawlers resumed their activity in trawling ground (Figures 8, S6). Despite the similar biogeochemical December (Figure S2), increasing near-bottom SSCs to 130- composition to the trawled site, in the majority of the samples, OC, 260 mg·L-1 in Flank-1000 and to ~50 mg·L-1 in Axis-900 during TN and the most labile compound classes (i.e. fatty acids, di- that month, while the sediment trap in Axis-1200 overfilled once carboxylic acids, and amino acids) had slightly higher content in the more, attaining a TMF of at least 94 g·m-2·d-1 (Figure S8). This sediment trap than in the trawling ground (Figures 6, 7), which increase in sediment transport in all moorings confirm that the reveals that the downward particle flux consisted of a combination fishing ground was replenished with new erodible sediment of eroded sediment transferred from fishing grounds with during the fishing intermittence, generating large sediment hemipelagic settling of particulate organic matter enriched in gravity flows when trawling activities resumed, as was biomarkers of marine origin (Figure 7). In contrast, in the observed at the onset of trawling season in early March (see absence of trawling-derived sediment gravity flows during the section 4.2). In both overfilled trap samples, the sediment summer, particulate organic matter had higher content of reactive transported into the canyon consisted of coarser particles with marine compound classes, such as in early June, associated to the high terrigenous and marine OC (Figures 6-8, S8), indicating the high NPP occurring then (Figures 2C, 7). translocation of freshly-deposited sediment with higher organic matter contents into the canyon. This corroborates that chronic trawling disturbance depletes organic matter from fishing 4.4 Intermission of trawling activities grounds (Pusceddu et al., 2014), while short periods without (November-December) trawling disturbance are insufficient to allow its recovery (Paradis et al., 2021b), which swiftly remove the most OC-rich Throughout October and November, SSCs in all three surficial sediment from fishing grounds when trawling moorings were negligible, while the downward sediment fluxes activities resume. in the sediment trap were only ~10 g·m-2·d-1 (Figures 3-6) with the exception of an isolated sediment gravity flow in late- October (Figure S7). This reduced sediment flux occurred in 4.5 Annual sediment transport and the absence of fishing activities in the Sant Sebastià fishing composition in Palamó s Canyon ground (Figure S2), confirming the strong relationship between sediment transport into Palamó s Canyon and this Sediment transport into Palamó s Canyon during 2017 was anthropogenic activity. During this low sediment flux period, highly dynamic, dominated by natural high-energy events such as sediment collected in the Axis-1200 trap had higher OC and TN eastern storms and DSWC, as well as by bottom trawling, both of contents (1.04% OC, 0.13% TN), than during the summer which transferred large volumes of sediment into the canyon. months (0.84% OC, 0.10% TN) when frequent trawling- Sediment transport through the Montgrı́ tributary gully was derived sediment gravity flows were registered in the canyon dominated by trawling-induced sediment gravity flows, which axis (Figures 6D, S7). This was likely the result of increasing NPP accounted for ~1500 kg·m-2 sediment whereas the eastern storm Frontiers in Marine Science 15 frontiersin.org Paradis et al. 10.3389/fmars.2022.1017052 A B C FIGURE 9 Time series of cumulative sediment transport at all moorings: (A) at the Flank-1000 mooring in the across-canyon direction (10 mab), (B) at Axis-900 and Axis-1200 in the across-canyon direction (5 mab) and (C) in the along-canyon direction (5 mab). Cumulative sediment transport for Axis-900 and Axis-1200 are highly underestimated due to their incomplete timeseries and the maximum data acquisition of the Axis-1200 turbidimeter during the first deployment (Figure 5). Note the different axes for the Flank-1000 cumulative sediment transport in comparison to the across-canyon and along-canyon transport of Axis-900 and Axis-1200. The inverted triangle on top indicates the occurrence of a dense shelf water cascading event. FIGURE 10 Map of main sediment transport pathways of the three mooring locations in Palamó s Canyon: up-canyon (UC) and across-canyon (AC) for Axis- 900 mooring (pink square), Montgrı ́ tributary gully (MT) for Flank-900 mooring (orange triangle), other tributary gully (OT), and down-canyon (DC) for Axis-1200 mooring (blue circle). Annotated in the figure is the cumulative sediment transport for natural high-energy events (N) and trawling-derived events (T). Note that the across-canyon cumulative transport for the Axis-1200 mooring combines sediment transported through both tributaries (MT and OT). Frontiers in Marine Science 16 frontiersin.org Paradis et al. 10.3389/fmars.2022.1017052 and DSWC at the beginning of the year transported ~20 kg·m-2 of Despite the lower OC content of these trawling-derived sediment through this tributary and into the canyon (Figures 9, 10). sediment fluxes (~0.8%), the higher TMF associated to bottom In the case of the moorings located along the canyon axis (Axis-900, trawling activities (>54 kg·m-2; Figure 6A), indicate that a Axis-1200), the magnitudes of sediment fluxes from natural and substantial amount of organic carbon is transferred into the anthropogenic events were comparable (Figures 9, 10). In Axis-900, canyon (> 0.43 kg OC·m-2), where it will eventually be buried in total sediment flux from storms and DSWC was at least ~120 kg·m- the canyon axis. Since efficient OC burial is generally associated 2 during 2017, while bottom trawling accounted for at least ~75 to high sedimentation rates (Blair and Aller, 2012; Bianchi et al., kg·m-2. Further down-canyon, natural high-energy events 2018), the enhanced sediment transfer and resulting increase in transported at least ~600 kg·m-2 of sediment through Axis-1200 sedimentation rates in the canyon axis by bottom trawling whereas trawling activities accounted for at least ~680 kg·m-2 of activities (Martı́n et al., 2008; Puig et al., 2015) could lead to sediment (Figures 9, 10). Despite the comparable magnitude of high OC burial in the canyon axis. However, this enhanced sediment fluxes, natural events tend to be sporadic and seasonal, transport of OC by bottom trawling activities could partially whereas bottom trawling causes continuous and persistent offset the OC lost from fishing grounds (Martı́n et al., 2014b; sediment fluxes into the canyon. Paradis et al., 2021b), and further studies will be necessary to With regards to the composition of sediment transported properly quantify the sediment and the OC budgets in Palamó s into the canyon, it is worth noting that terrestrial compounds Canyon, and assess the potential effects on benthic communities. (i.e. lignin phenols and cutin acids) were detected during the whole sampling period (Figure 7), irrespectively of the sediment transport process, as observed in other submarine canyons 5 Conclusion nearby (Tesi et al., 2010; Pasqual et al., 2013), which confirms the role of submarine canyons as across-margin conduits of High energy events such as river floods and storms are terrestrial organic matter towards deep-sea basins. The high acid important mechanisms that enhance the transport of particulate to aldehyde ratio of vanillyl and syringyl phenols, SP/VP and matter into submarine canyons, but recent studies have emphasized CP/VP ratios (Figure S3) indicates that terrestrial organic matter that bottom trawling activities modify natural sedimentary was highly degraded and mostly originated from non-woody dynamics within these geomorphological features. This study angiosperm leaves and grasses (Hedges and Mann, 1979; Goñi shows how bottom trawling is altering sedimentary dynamics in and Hedges, 1995). The highly degraded compositions are submarine canyons and reveals that the organic matter composition expected given the distance of the Axis-1200 mooring to the of down-canyon particulate matter is modified by this shore and the long transit times that promote its alteration anthropogenic activity. While winter storms transport terrigenous (Pasqual et al., 2013). Presence of non-woody organic matter and fresh marine organic matter into the canyon in the presence of signatures is also expected given the association of these phytoplankton blooms, trawling-induced sediment gravity flows compounds to fine-grained sediment that can be more easily transfer more degraded organic matter. Trawling events interrupt transported offshore than larger-sized woody detritus with high the naturally high concentrations of fresh and reactive organic vanillyl phenol contents, which tend to accumulate in more matter compounds that are usually transferred during the calm nearshore environments (Gordon and Goñi, 2004; Pasqual et al., summer months, which could affect benthic communities that 2013; Zhang et al., 2014). inhabit the canyon and rely on the input of fresh organic matter. Bottom trawling-derived sediment fluxes in the canyon axis had However, the higher downward particulate fluxes into the canyon lower OC content (~0.8%) and consisted of more degraded organic by bottom trawling also increases organic carbon fluxes, urging to matter in comparison to downward sediment fluxes in the absence better constrain the sediment and organic carbon budgets in bottom of bottom trawling during the summer, as well as in comparison to trawling-affected submarine canyons. the sediment fluxes during the storms in February (~1% OC), since these natural processes occurred in coincidence with high net primary productivity in surface waters or an intermittence of Data availability statement fishing activities. The enhanced particulate matter fluxes of eroded sediment by bottom trawling modify the OC composition The raw data supporting the conclusions of this article will of downward particulate fluxes in submarine canyons from fresh be made available by the authors, without undue reservation. marine organic matter to degraded and less reactive terrestrial organic matter. This alteration in the nutritional value of organic matter would affect the benthic community structure inhabiting Author contributions submarine canyons, and could explain the lower diversity of meiofauna mainly consisting of opportunistic nematode species SP, MA-C, AP, and PP conceptualized the study design and as observed within the axes of the studied canyon (Pusceddu et al., performed the fieldwork. MA-C and SP analyzed sensor data from 2014) as well as in the adjacent Blanes Canyon (Romá n et al., 2018). moorings. SP performed the laboratory analyses and interpreted the Frontiers in Marine Science 17 frontiersin.org Paradis et al. 10.3389/fmars.2022.1017052 results alongside MG. AP, PM, and PP provided funding for the the support provided to its Environment Laboratories by the study. SP wrote the manuscript with the contribution of all co- Government of the Principality of Monaco. authors, who approve the submitted version. Conflict of interest Funding The authors declare that the research was conducted in the The results presented in this study were obtained within the absence of any commercial or financial relationships that could “Assessment of Bottom-trawling Impacts in Deep-sea Sediments” be construed as a potential conflict of interest. (ABIDES) Spanish Research Project (CTM2015-65142-R). Additional funds were provided by the Generalitat de Catalunya (2017 SGR-663 and SGR- 1588), by the Australian Research Publisher’s note Council LIEF Project (LE170100219) and by the “Assessment of Bottom-trawling Resuspension Impacts in deep benthic All claims expressed in this article are solely those of the Communities” (ABRIC) Spanish Research Project (RTI2018- authors and do not necessarily represent those of their affiliated 096434-B-I00). This work is contributing to the ICTA’s “Unit of organizations, or those of the publisher, the editors and the Excellence” Maria de Maetzu (CEX2019-000940-M), and ICM- reviewers. Any product that may be evaluated in this article, or CSIC’s “Center of Excellence” Severo Ochoa (CEX2019-000928-S). claim that may be made by its manufacturer, is not guaranteed Open access funding provided by ETH Zurich. or endorsed by the publisher. Acknowledgments Supplementary material We would like to thank the crew of the R/V Garcı́a del Cid as The Supplementary Material for this article can be found well as Elena Martinez, Neus Maestro, and Silvia de Diago, who online at: https://www.frontiersin.org/articles/10.3389/fmars. helped collect and process the samples. The IAEA is grateful for 2022.1017052/full#supplementary-material References Allen, S. E., and Durrieu de Madron, X. (2009). A review of the role of submarine antennatus) con arte de arrastre de fondo en determinadas zonas marı́timas canyons in deep-ocean exchange with the shelf. Ocean Sci. 5, 607–620. pró ximas a palamó s. Boletin Oficial del Estado 30, 8169. doi: 10.5194/os-5-607-2009 Bonnin, J., Heussner, S., Calafat, A., Fabres, J., Palanques, A., Durrieu de Arjona-Camas, M., Puig, P., Palanques, A., Durá n, R., White, M., Paradis, S., Madron, X., et al. (2008). Comparison of horizontal and downward particle et al. (2021). Natural vs. trawling-induced water turbidity and sediment transport fluxes across canyons of the gulf of lions (NW mediterranean): Meteorological variability within the palamó s canyon (NW Mediterranean). Mar. Geophys. Res. 42, and hydrodynamical forcing. Cont. Shelf Res. 28, 1957–1970. doi: 10.1016/ 38. doi: 10.1007/s11001-021-09457-7 j.csr.2008.06.004 Arjona-Camas, M., Puig, P., Palanques, A., Emelianov, M., and Durá n, R. Canals, M., Puig, P., de Madron, X. D., Heussner, S., Palanques, A., and Fabres, J. (2019). Evidence of trawling-induced resuspension events in the generation of (2006). Flushing submarine canyons. Nature 444, 354. doi: 10.1038/nature05271 nepheloid layers in the foix submarine canyon (NW Mediterranean). J. Mar. Syst. 196, 86–96. doi: 10.1016/j.jmarsys.2019.05.003 Daly, E., Johnson, M. P., Wilson, A. M., Gerritsen, H. D., Kiriakoulakis, K., Allcock, A. L., et al. (2018). Bottom trawling at whittard canyon: Evidence for Baker, E. T., Milburn, H. B., and Tennant, D. A. (1988). Field assessment of seabed modification, trawl plumes and food source heterogeneity. Prog. Oceanogr. sediment trap efficiency under varying flow conditions. J. Mar. Res. 46, 573–592. 169, 227–240. doi: 10.1016/j.pocean.2017.12.010 doi: 10.1357/002224088785113522 De Leo, F. C., Vetter, E. W., Smith, C. R., Rowden, A. A., and McGranaghan, M. Behrenfeld, M. J., and Falkowski, P. G. (1997). Photosynthetic rates derived from (2014). Spatial scale-dependent habitat heterogeneity influences submarine canyon satellite-based chlorophyll concentration. Limnol. Oceanogr. 42, 1–20. doi: 10.4319/ macrofaunal abundance and diversity off the main and Northwest Hawaiian lo.1997.42.1.0001 islands. Deep Sea Res. Part II Top. Stud. Oceanogr. 104, 267–290. doi: 10.1016/ Bianchi, T. S., Cui, X., Blair, N. E., Burdige, D. J., Eglinton, T. I., and Galy, V. j.dsr2.2013.06.015 (2018). Centers of organic carbon burial and oxidation at the land-ocean interface. de Stigter, H. C., Boer, W., de Jesus Mendes, P. A., Jesus, C. C., Thomsen, L., van Org. Geochem. 115, 138–155. doi: 10.1016/j.orggeochem.2017.09.008 den Bergh, G. D., et al. (2007). Recent sediment transport and deposition in the Bjørkan, M., Company, J. B., Gorelli, G., Sardà, F., and Massaguer, C. (2020). nazaré canyon, Portuguese continental margin. Mar. Geol. 246, 144–164. When fishemen take charge: The development of a management plan for the red doi: 10.1016/j.margeo.2007.04.011 shrimp fishery in Mediterranean Sea (NE Spain) (Cham: Springer), 159–178. Durrieu de Madron, X., Ferré , B., Le Corre, G., Grenz, C., Conan, P., Pujo-Pay, doi: 10.1007/978-3-030-26784-1_10 M., et al. (2005). Trawling-induced resuspension and dispersal of muddy sediments Blair, N. E., and Aller, R. C. (2012). The fate of terrestrial organic carbon in the and dissolved elements in the gulf of lion (NW Mediterranean). Cont. shelf Res. 25 marine environment. Ann. Rev. Mar. Sci. 4, 401–423. doi: 10.1146/annurev- (19-20), 2387–2409. doi: 10.1016/j.pocean.2004.08.004 marine-120709-142717 Fabres, J., Tesi, T., Velez, J., Batista, F., Lee, C., Calafat, A., et al. (2008). Seasonal BOE (2017). Resolució n de 26 de enero de 2017, de la secretarı́a general de pesca, and event-controlled export of organic matter from the shelf towards the gulf of por la que se fija, para 2017, el periodo de veda establecido en la orden AAA/923/ lions continental slope. Cont. Shelf Res. 28, 1971–1983. doi: 10.1016/ 2013, de 16 de mayo, por la que se regula la pesca de gamba rosada (Aristeus j.csr.2008.04.010 Frontiers in Marine Science 18 frontiersin.org Paradis et al. 10.3389/fmars.2022.1017052 Farrugio, H. (2012). “A refugium for the spawners of exploited Mediterranean Martı́n, J., Puig, P., Palanques, A., and Ribó , M. (2014b). Trawling-induced daily marine species: the canyons of the continental slope of the gulf of lion,” in sediment resuspension in the flank of a Mediterranean submarine canyon. Deep Mediterranean Submarine canyons: Ecology and governance. Ed. M. Würtz Sea Res. Part II Top. Stud. Oceanogr. 104, 174–183. doi: 10.1016/j.dsr2.2013.05.036 (Gland, Switzerland: International Union for Conservation of Nature), 45–50. Mendoza, E. T., and Jimé nez, A. A. (2009). “Vulnerability assessment to coastal Fernandez-Arcaya, U., Ramirez-Llodra, E., Aguzzi, J., Allcock, A. L., Davies, J. S., storms at a regional scale,” in Coastal engineering 2008, vol. vol 5. , 4154–4166. Dissanayake, A., et al. (2017). Ecological role of submarine canyons and need for Mortlock, R. A., and Froelich, P. N. (1989). A simple method for the rapid canyon conservation: A review. Front. Mar. Sci. 4. doi: 10.3389/fmars.2017.00005 determination of biogenic opal in pelagic marine sediments. Deep Sea Res. Part A. Goni, M. A., Aceves, H., Benitez-Nelson, B., Tappa, E., Thunell, R., Black, D. E., Oceanogr. Res. Pap. 36, 1415–1426. doi: 10.1016/0198-0149(89)90092-7 et al. (2009). Oceanographic and climatologic controls on the compositions and Nieuwenhuize, J., Maas, Y. E. M., and Middelburg, J. J. (1994). Rapid analysis of fluxes of biogenic materials in the water column and sediments of the cariaco basin organic carbon and nitrogen in particulate materials. Mar. Chem. 45, 217–224. over the late Holocene. Deep Sea Res. Part I Oceanogr. Res. Pap. 56, 614–640. doi: 10.1016/0304-4203(94)90005-1 doi: 10.1016/j.dsr.2008.11.010 Palanques, A., Garcı́a-Ladona, E., Gomis, D., Martı́n, J., Marcos, M., Pascual, A., Goñi, M. A., and Hedges, J. I. (1995). Sources and reactivities of marine-derived et al. (2005). General patterns of circulation, sediment fluxes and ecology of the organic matter in coastal sediments as determined by alkaline CuO oxidation. palamó s (La fonera) submarine canyon, northwestern Mediterranean. Prog. Geochim. Cosmochim. Acta 59, 2965–2981. doi: 10.1016/0016-7037(95)00188-3 Oceanogr. 66, 89–119. doi: 10.1016/j.pocean.2004.07.016 Goñi, M. A., and Montgomery, S. (2000). Alkaline CuO oxidation with a Palanques, A., Martı́n, J., Puig, P., Guillé n, J., Company, J. B., and Sardà, F. microwave digestion system: Lignin analyses of geochemical samples. Anal. (2006). Evidence of sediment gravity flows induced by trawling in the palamó s Chem. 72, 3116–3121. doi: 10.1021/ac991316w (Fonera) submarine canyon (northwestern Mediterranean). Deep Sea Res. Part I Goñi, M. A., O’Connor, A. E., Kuzyk, Z. Z., Yunker, M. B., Gobeil, C., and Oceanogr. Res. Pap. 53, 201–214. doi: 10.1016/j.dsr.2005.10.003 Macdonald, R. W. (2013). Distribution and sources of organic matter in surface Palanques, A., and Puig, P. (2018). Particle fluxes induced by benthic storms marine sediments across the north American Arctic margin. J. Geophys. Res. Ocean. during the 2012 dense shelf water cascading and open sea convection period in the 118, 4017–4035. doi: 10.1002/jgrc.20286 northwestern Mediterranean basin. Mar. Geol. 406, 119–131. doi: 10.1016/ Gordon, E. S., and Goñi, M. A. (2004). Controls on the distribution and accumulation j.margeo.2018.09.010 of terrigenous organic matter in sediments from the Mississippi and atchafalaya river Palanques, A., Puig, P., Durrieu de Madron, X., Sanchez-Vidal, A., Pasqual, C., margin. Mar. Chem. 92, 331–352. doi: 10.1016/J.MARCHEM.2004.06.035 Martı́n, J., et al. (2012). Sediment transport to the deep canyons and open-slope of Grinyó , J., Isla, E., Peral, L., and Gili, J.-M. (2017). Composition and temporal the western gulf of lions during the 2006 intense cascading and open-sea variability of particle fluxes in an insular canyon of the northwestern convection period. Prog. Oceanogr. 106, 1–15. doi: 10.1016/j.pocean.2012.05.002 Mediterranean Sea. Prog. Oceanogr. 159, 323–339. doi: 10.1016/j.pocean. Paradis, S., Goñi, M., Masqué , P., Durá n, R., Arjona-Camas, M., Palanques, A., 2017.11.005 et al. (2021b). Persistence of biogeochemical alterations of deep-Sea sediments by Hedges, J. I., and Mann, D. C. (1979). The characterization of plant tissues by bottom trawling. Geophys. Res. Lett. 48, e2020GL091279. doi: 10.1029/ their lignin oxidation products. Geochim. Cosmochim. Acta 43, 1803–1807. 2020GL091279 doi: 10.1016/0016-7037(79)90028-0 Paradis, S., lo Iacono, C., Masqué , P., Puig, P., Palanques, A., Russo, T., et al. Heussner, S., Durrieu de Madron, X., Calafat, A., Canals, M., Carbonne, J., (2021a). Evidence of large increases in sedimentation rates due to fish trawling in Delsaut, N., et al. (2006). Spatial and temporal variability of downward particle submarine canyons of the gulf of Palermo (SW Mediterranean). Mar. pollut. Bull. fluxes on a continental slope: Lessons from an 8-yr experiment in the gulf of 172, 112861. doi: 10.1016/j.marpolbul.2021.112861 lions (NW Mediterranean). Mar. Geol. 234, 63–92. doi: 10.1016/J.MARGEO. Paradis, S., Masqué , P., Puig, P., Juan-Dı́az, X., Gorelli, G., Company, J. B., et al. 2006.09.003 (2018a). Enhancement of sedimentation rates in the foix canyon after the renewal Heussner, S., Ratti, C., and Carbonne, J. (1990). The PPS 3 time-series sediment of trawling fleets in the early XXIst century. Deep Res. Part I Oceanogr. Res. Pap. trap and the trap sample processing techniques used during the ECOMARGE 132, 51–59. doi: 10.1016/j.dsr.2018.01.002 experiment. Continent. Shelf Res. 10, 943–958. doi: 10.1016/0278-4343(90)90069-X Paradis, S., Puig, P., Masqué , P., and Juan-Diá z, X. (2017). Bottom-trawling Hsu, R. T., Liu, J. T., Su, C.-C., Kao, S.-J., Chen, S.-N., Kuo, F.-H., et al. (2014). along submarine canyons impacts deep sedimentary regimes. Sci. Rep. 7, 43332. On the links between a river’s hyperpycnal plume and marine benthic nepheloid doi: 10.1038/srep43332 layer in the wake of a typhoon. Prog. Oceanogr. 127, 62–73. doi: 10.1016/ Paradis, S., Puig, P., Sá nchez-Vidal, A., Masqué , P., Garcı́a-Orellana, J., Calafat, j.pocean.2014.06.001 A., et al. (2018b). Spatial distribution of sedimentation-rate increases in blanes Lastras, G., Canals, M., Amblas, D., Lavoie, C., Church, I., De Mol, B., et al. canyon caused by technification of bottom trawling fleet. Prog. Oceanogr. 169, 241– (2011). Understanding sediment dynamics of two large submarine valleys from 252. doi: 10.1016/j.dsr.2018.01.002 seafloor data: Blanes and la fonera canyons, northwestern Mediterranean Sea. Mar. Pasqual, C., Goñi, M.A., Tesi, T., Sanchez-Vidal, A., Calafat, A., and Canals, M. Geol. 280, 20–39. doi: 10.1016/j.margeo.2010.11.005 (2013). Composition and provenance of terrigenous organic matter transported Liu, J. T., Wang, Y.-H., Yang, R. J., Hsu, R. T., Kao, S.-J., Lin, H.-L., et al. (2012). along submarine canyons in the gulf of lion (NW Mediterranean Sea). Prog. Cyclone-induced hyperpycnal turbidity currents in a submarine canyon. J. Oceanogr. 118, 81–94. doi: 10.1016/j.pocean.2013.07.013 Geophys. Res. Ocean. 117, C04033. doi: 10.1029/2011JC007630 Pasqual, C., Sanchez-Vidal, A., Zú ñiga, D., Calafat, A., Canals, M., Durrieu de Lopez-Fernandez, P., Bianchelli, S., Pusceddu, A., Calafat, A., Sanchez-Vidal, A., Madron, X., et al. (2010). Flux and composition of settling particles across the and Danovaro, R. (2013a). Bioavailability of sinking organic matter in the blanes continental margin of the gulf of lion: the role of dense shelf water cascading. canyon and the adjacent open slope (NW Mediterranean Sea). Biogeosciences 10, Biogeosciences 7, 217–231. doi: 10.5194/bg-7-217-2010 3405–3420. doi: 10.5194/bg-10-3405-2013 Pedregosa, F., Varoquaux, G., Gramfort, A., Michel, V., Thirion, B., Grisel, O., Lopez-Fernandez, P., Calafat, A., Sanchez-Vidal, A., Canals, M., Mar Flexas, M., et al. (2011). Scikit-learn: Machine learning in Python. J. Mach. Learn. Res. 12, Cateura, J., et al. (2013b). Multiple drivers of particle fluxes in the blanes submarine 2825–2830. canyon and southern open slope: Results of a year round experiment. Prog. Pedrosa-Pàmies, R., Sanchez-Vidal, A., Calafat, A., Canals, M., and Durán, R. (2013). Oceanogr. 118, 95–107. doi: 10.1016/j.pocean.2013.07.029 Impact of storm-induced remobilization on grain size distribution and organic carbon Martı́n, J., Palanques, A., and Puig, P. (2006). Composition and variability of content in sediments from the blanes canyon area, NW Mediterranean Sea. Prog. downward particulate matter fluxes in the palamó s submarine canyon (NW Oceanogr. 118, 122–136. doi: 10.1016/j.pocean.2013.07.023 Mediterranean). J. Mar. Syst. 60, 75–97. doi: 10.1016/j.jmarsys.2005.09.010 Porter, M., Inall, M. E., Hopkins, J., Palmer, M. R., Dale, A. C., Aleynik, D., et al. Martı́n, J., Palanques, A., and Puig, P. (2007). Near-bottom horizontal transfer of (2016). Glider observations of enhanced deep water upwelling at a shelf break particulate matter in the palamó s submarine canyon (NW Mediterranean). J. Mar. canyon: A mechanism for cross-slope carbon and nutrient exchange. J. Geophys. Res. 65, 193–218. doi: 10.1357/002224007780882569 Res. Ocean. 121, 7575–7588. doi: 10.1002/2016JC012087 Martı́n, J., Palanques, A., Vitorino, J., Oliveira, A., and de Stigter, H. C. (2011). Puig, P., Canals, M., Company, J. B., Martı́n, J., Amblas, D., Lastras, G., et al. Near-bottom particulate matter dynamics in the nazaré submarine canyon under (2012). Ploughing the deep sea floor. Nature 489, 286–289. doi: 10.1038/ calm and stormy conditions. Deep Sea Res. Part II Top. Stud. Oceanogr. 58, 2388– nature11410 2400. doi: 10.1016/j.dsr2.2011.04.004 Puig, P., Martı́n, J., Masqué , P., and Palanques, A. (2015). Increasing sediment Martı́n, J., Puig, P., Masqué , P., Palanques, A., and Sá nchez-Gó mez, A. (2014a). accumulation rates in la fonera (Palamó s) submarine canyon axis and their Impact of bottom trawling on deep-sea sediment properties along the flanks of a relationship with bottom trawling activities. Geophys. Res. Lett. 42, 8106–8113. submarine canyon. PloS One 9, e104536. doi: 10.1371/journal.pone.0104536 doi: 10.1002/2015GL065052 Martı́n, J., Puig, P., Palanques, A., Masqué , P., and Garcı́a-Orellana, J. (2008). Puig, P., Palanques, A., and Martı́n, J. (2014). Contemporary sediment-transport Effect of commercial trawling on the deep sedimentation in a Mediterranean processes in submarine canyons. Ann. Rev. Mar. Sci. 6, 53–77. doi: 10.1146/ submarine canyon. Mar. Geol. 252, 150–155. doi: 10.1016/j.margeo.2008.03.012 annurev-marine-010213-135037 Frontiers in Marine Science 19 frontiersin.org Paradis et al. 10.3389/fmars.2022.1017052 Puig, P., Palanques, A., Orange, D. L., Lastras, G., and Canals, M. (2008). Dense canyon system during an exceptional river flood: Effects of typhoon morakot on the shelf water cascades and sedimentary furrow formation in the cap de creus canyon, gaoping river–canyon system. Mar. Geol. 363, 191–201. doi: 10.1016/J.MARGEO. northwestern Mediterranean Sea. Cont. Shelf Res. 28 (15), 2017–2030. doi: 10.1016/ 2015.02.013 j.csr.2008.05.002 Tesi, T., Langone, L., Goñi, M. A., Turchetto, M., Miserocchi, S., and Boldrin, A. Pusceddu, A., Bianchelli, S., and Danovaro, R. (2015). Quantity and biochemical (2008). Source and composition of organic matter in the bari canyon (Italy): Dense composition of particulate organic matter in a highly trawled area (Thermaikos water cascading versus particulate export from the upper ocean. Deep Sea Res. Part gulf, Eastern Mediterranean Sea). Adv. Oceanogr. Limnol. 6, 21–32. doi: 10.4081/ I Oceanogr. Res. Pap. 55, 813–831. doi: 10.1016/j.dsr.2008.03.007 aiol.2015.5448 Tesi, T., Puig, P., Palanques, A., and Goñi, M. A. (2010). Lateral advection of Pusceddu, A., Bianchelli, S., Martin, J., Puig, P., Palanques, A., Masque, P., et al. (2014). organic matter in cascading-dominated submarine canyons. Prog. Oceanogr. 84, Chronic and intensive bottom trawling impairs deep-sea biodiversity and ecosystem 185–203. doi: 10.1016/J.POCEAN.2009.10.004 functioning. Proc. Natl. Acad. Sci. 111, 8861–8866. doi: 10.1073/pnas.1405454111 Tudela, S., Sardà, F., Maynou, F., and Demestre, M. (2003). Influence of Ribó , M., Puig, P., Palanques, A., and Lo Iacono, C. (2011). Dense shelf water submarine canyons on the distribution of the deep-water shrimp, Aristeus cascades in the cap de creus and palamó s submarine canyons during winters 2007 antennatus (Risso 1816) in the NW Mediterranean. Crustaceana 76, 217–225. and 2008. Mar. Geol. 284, 175–188. doi: 10.1016/j.margeo.2011.04.001 doi: 10.1163/156854003321824567 Román, S., Vanreusel, A., Ingels, J., and Martin, D. (2018). Nematode community Vetter, E. W., Smith, C. R., and De Leo, F. C. (2010). Hawaiian Hotspots: zonation in response to environmental drivers in blanes canyon (NW Mediterranean). Enhanced megafaunal abundance and diversity in submarine canyons on the J. Exp. Mar. Bio. Ecol 502, 111–128. doi: 10.1016/j.jembe.2017.08.010 oceanic islands of Hawaii. Mar. Ecol. 31, 183–199. doi: 10.1111/j.1439-0485. Rumı́n-Caparró s, A., Sanchez-Vidal, A., Gonzá lez-Pola, C., Lastras, G., Calafat, 2009.00351.x A., and Canals, M. (2016). Particle fluxes and their drivers in the avilé s submarine Wilson, A. M., Kiriakoulakis, K., Raine, R., Gerritsen, H. D., Blackbird, S., canyon and adjacent slope, central cantabrian margin, bay of Biscay. Prog. Allcock, A. L., et al. (2015). Anthropogenic influence on sediment transport in the Oceanogr. 144, 39–61. doi: 10.1016/j.pocean.2016.03.004 whittard canyon, NE Atlantic. Mar. pollut. Bull. 101, 320–329. doi: 10.1016/ Sanchez-Vidal, A., Canals, M., Calafat, A. M., Lastras, G., Pedrosa-Pàmies, R., j.marpolbul.2015.10.067 Mené ndez, M., et al. (2012). Impacts on the deep-Sea ecosystem by a severe coastal Wu, Y., Liu, Z., Hu, J., Zhu, Z., Liu, S., and Zhang, J. (2016). Seasonal dynamics storm. PloS One 7, e30395. doi: 10.1371/journal.pone.0030395 of particulate organic matter in the changjiang estuary and adjacent coastal waters Sanchez-Vidal, A., Pasqual, C., Kerhervé , P., Calafat, A., Heussner, S., Palanques, illustrated by amino acid enantiomers. J. Mar. Syst. 154, 57–65. doi: 10.1016/ A., et al. (2008). Impact of dense shelf water cascading on the transfer of organic j.jmarsys.2015.04.006 matter to the deep western Mediterranean basin. Geophys. Res. Lett. 35, L05605. Zhang, Y., Kaiser, K., Li, L., Zhang, D., Ran, Y., and Benner, R. (2014). Sources, doi: 10.1029/2007GL032825 distributions, and early diagenesis of sedimentary organic matter in the pearl river Sardà, F., Cartes, J. E., and Norbis, W. (1994). Spatio-temporal structure of the region of the south China Sea. Mar. Chem. 158, 39–48. doi: 10.1016/ deep-water shrimp aristeus antennatus (Decapoda: Aristeidae) population in the j.marchem.2013.11.003 western Mediterranean. Fish. Bull. 92, 599–607. Zú ñiga, D., Flexas, M. M., Sanchez-Vidal, A., Coenjaerts, J., Calafat, A., Jordà, G., Sparkes, R. B., Lin, I.-T., Hovius, N., Galy, A., Liu, J. T., Xu, X., et al. (2015). et al. (2009). Particle fluxes dynamics in blanes submarine canyon (Northwestern Redistribution of multi-phase particulate organic carbon in a marine shelf and Mediterranean). Prog. Oceanogr. 82, 239–251. doi: 10.1016/j.pocean.2009.07.002 Frontiers in Marine Science 20 frontiersin.org