Assessment of Short-Term Sediment Deposition Patterns Along the Palamós Submarine Canyon (NW Mediterranean) Using ²³⁴Th

Abstract

Sedimentary dynamics in the Palamós Canyon are influenced by river inputs and storm resuspension, as well as by bottom trawling on the canyon flanks. In this study, we estimate recent sediment deposition patterns along the canyon axis using the excess activity concentration of the short-lived radiotracer ²³⁴Th (half-life of 24.1 days). Sediment cores were obtained at various locations along the canyon axis from a depth of approximately 800 m to 2100 m in June 2023 and August 2024. Excess ²³⁴Th (²³⁴Th_xs) was detected in all sampled sites with variable penetration depths (0.5–3.5 cm). ²³⁴Th_xs-derived estimations of mixing rates decreased downcanyon from up to 15.6 cm² y⁻¹ at the canyon head (~800 m) to negligible mixing at the canyon mouth (~2100 m). ²³⁴Th_xs inventories, a proxy of recent sediment deposition, were high (1800–3490 Bq m⁻²) at the canyon head and at the upper canyon (~1400 m) close to fishing grounds and decreased downcanyon (82–694 Bq m⁻²) at the lower canyon (~1800 m) and canyon mouth. Inventories varied 2-fold across years presumably attributed to enhanced riverine and bottom trawling sediment fluxes. Similar ²³⁴Th-derived sediment deposition patterns can be found in submarine canyons worldwide, highlighting the value of this radiotracer for sedimentary dynamics studies in such complex environments.

1. Introduction

Submarine canyons incise the continental margins and can favor transport of particles from continental shelves and upper slope regions towards the deep sea [1]. While near 6000 canyons have been mapped globally [2], they differ in topography and location, which has implications for the connectivity and efficiency of these canyons as across-margin conduits of particulate matter [3,4]. Although particle fluxes in canyons are variable, a general decreasing trend of particle flux and sedimentation rates, with increasing distance from coastal particle sources and water depth is expected. Particle transfer into submarine canyons can be intensified by high-energy hydrodynamic natural events, such as seasonal storms and dense shelf water cascading [1], or through sediment resuspension by anthropogenic activities, such as bottom trawling (e.g., [5]). Depending, among others, on the frequency and strength of such sediment transport events, particulate matter is preferentially deposited in the upper- and middle-canyon regions or remobilized and transported further downslope along the canyon axis, towards the adjacent basin [1].

Canyons are among the most biologically productive ecosystems on the planet (e.g., [6,7]) and can sequester large quantities of particulate organic carbon (POC), primarily due to the high fluxes of particulate matter occurring in these systems [8,9]. Vertical export of POC through the water column, from surface waters to the canyon seabed, can be enhanced during marine algae blooms [10]. Also, terrigenous POC discharged by rivers can reach the canyon by lateral advection from the shelf during energetic sediment transport events [11,12]. Benthic communities within canyons may benefit from increased POC fluxes, depending on the quality of the organic fraction reaching the seafloor [13,14].

Subject of this study, the Palamós Canyon, is located on the northwestern Mediterranean margin (Figure 1). Sediment transport is active in this submarine canyon and importantly influenced by natural and anthropogenic resuspension events [15,16]. Shelf sediment resuspension is efficiently induced during easterly storms occurring in this region in autumn and winter [15], which can be enhanced by the presence of dense shelf waters loaded with particles, which can cascade into the Palamós canyon head [17,18]. Bottom trawling on the Palamós canyon flanks also produces sediment gravity flows channelized through canyon tributaries towards the canyon axis, although the magnitude of trawling-derived sediment gravity flow generally decreases as the fishing season advances [16,18]. In this way, sediment resuspension by bottom trawling has increased sedimentation rates within the Palamós canyon axis on a decadal scale, as provided by lead-210 (²¹⁰Pb) chronology [19]. However, the spatio-temporal variations in recent sediment deposition in Palamós Canyon have not yet been studied on seasonal time scales and will be addressed in this study using the naturally occurring radionuclide thorium-234 (²³⁴Th).

Figure 1. Bathymetric map showing the sampling locations along the axis of the Palamós Canyon. The coring sites are shown as white circles. The Ter River gauging site and Cap de Begur buoy are indicated with blue and yellow markers, respectively. The main bottom trawling fishing grounds along both canyon flanks are shown as red-outlined areas, with grid cells based on vessel monitoring system (VMS) fishing effort data. Inset: The study area within the northwestern Mediterranean is indicated by a red marker. Map created using QGIS 3.34.

²¹⁰Pb (T_1/2 = 22.3 years) has been extensively used to study sedimentation processes in the marine environment, providing average sedimentation rates over decadal timescales [20]. However, short-term sediment fluxes to the seafloor and seasonal and interannual variability in sediment deposition needs to be resolved using radionuclides with a shorter half-life, such as ²³⁴Th (T_1/2 = 24.1 days). This radionuclide allows to estimate recent sedimentation processes over the last months, such as depositional pulses due to tidal influence, river flooding, or environmental pollution among others [21,22,23].

²³⁴Th is produced by the decay of the ubiquitous, long-lived uranium-238 (²³⁸U; T_1/2 = 4.5·10⁹ years). Given its high particle-reactivity, ²³⁴Th in the water column is adsorbed onto sinking particles and reaches the seafloor via particle transport, leading to an excess ²³⁴Th (²³⁴Th_xs) signal in surface sediment with respect to ²³⁴Th in secular equilibrium with ²³⁸U found naturally in marine sediments [24,25]. Due to its relatively short half-life, the ²³⁴Th_xs signal limits the time frame in which the particles forming the sediment layer were deposited on the seabed to approximately the last 120 days (five half-lives of ²³⁴Th). Analyzing ²³⁴Th_xs inventories can be used to assess relative changes in seasonal or episodic transport and deposition of sediment, where higher ²³⁴Th_xs inventories relate to higher recent sediment accumulation by vertical and lateral inputs [25,26]. With time, the ²³⁴Th_xs activity concentration in sediments decreases due to radioactive decay, as well as by diffusive mixing of particles across sediment layers [25,27]. Thus, ²³⁴Th_xs should mainly be present in recently deposited surface sediments. The detection of ²³⁴Th_xs deep in the sediment therefore indicates diffusive mixing of sediment caused by biological [24,28] or physical [26,29] processes. Based on the ²³⁴Th_xs concentration profile, the mixing intensity (D_b), also referred to as a bioturbation coefficient when benthic bioturbating organisms are present, can be estimated [27,28].

In this study, we used ²³⁴Th_xs to assess short-term sediment deposition patterns in the Palamós Canyon. We report surface ²³⁴Th_xs activity concentrations, penetration depths, inventories and mixing rates based on ²³⁴Th_xs profiles from sediment cores collected along the canyon axis in the summer of 2023 and 2024. Excess ²³⁴Th inventories are interpreted as a proxy for the magnitude of recent sediment deposition and further evaluated considering the magnitude of both natural and trawling-derived sediment transport events that may have influenced sediment distribution during the spring-summer season. Additionally, we present a compilation of studies that have used ²³⁴Th to investigate short-term sedimentary deposition in submarine canyons, collectively enhancing our understanding of particle transport on continental margins.

2. Materials and Methods

2.1. Study Area

The Palamós canyon head has an approximate N-S orientation and cuts into the continental shelf at a distance of 3 km from the coastline (Figure 1). The canyon axis is oriented in a WNW-ESE direction with the Liguro-Provençal-Catalan Current (or Northern Current) flowing almost perpendicular to the canyon in a mean NE-SW direction [30]. The canyon axis spans up to 40 km, being characterized by two distinct domains: an inner domain incising the continental margin until the canyon axis reaches approximately 1200 m water depth, experiencing higher sediment load than the outer domain below 1200 m water depth [15]. The mouth of the Ter River is located 10 km north from the Palamós canyon head and represents an important source of sediment into the canyon during events of flooding [15,31]. The mean water discharge was 9 ± 30 m³ s⁻¹ over the five most recent years (2019–2024) including the sampling periods, with maximum annual flows of 9 m³ s⁻¹ in 2023 and 18 m³ s⁻¹ in 2024 (hydrological data available on the ACA website [32]). However, peak flows of >100 m³ s⁻¹ during flood events have been recorded for Ter River (e.g., [31,33]). Furthermore, this canyon supports an industrialized fleet of trawl fishing vessels targeting almost exclusively the coveted red-and-blue deep-sea shrimp (Aristeus antennatus) trawling a depth range of 400–800 m along the canyon flanks, namely the Sant Sebastià fishing ground on the northern flank and the Rostoll fishing ground on the southern flank [19] (Figure 1).

2.2. Sample Collection

Sediment cores preserving the sediment–water interface were collected along the Palamós canyon axis on 29 June 2023 (ARCO-1 cruise) and from 16 to 22 August 2024 (ARCO-2 cruise) using a KC Denmark 6-tube multicorer (KC Denmark A/S, Silkeborg, Denmark) (Figure 1). Sediment sampling took place along the canyon axis with station 1 at the canyon head (~800 m), station 2 at the upper canyon in the vicinity of the fishing grounds (~1400 m), station 3 at the lower canyon (~1800 m) and station 4 at the canyon mouth (~2100 m). Station 4 could not be sampled in 2023 due to unfavorable weather conditions and limited cruise time, whereas all four stations were sampled in 2024. Within 24 h of retrieval, the cores were sliced at 0.5 cm intervals from surface until 10 cm depth on deck using a core extruder. Each section was sealed in plastic bags and frozen at −20 °C until further processing. Sediment samples were weighed before (wet weight), as well as after drying (dry weight) in an oven at 60 °C to a constant weight (ARCO-1) or via freeze-drying (ARCO-2). Dry weight was corrected for salt content assuming a pore-water salinity of 38. Dry bulk density (in g cm⁻³) was calculated as dry weight divided by total sample volume, estimated from seawater and sediment grain densities of 1.025 g cm⁻³ and 2.65 g cm⁻³, respectively [34].

2.3. Gamma Spectrometry and ²³⁴Th Quantification

Approximately 5.5 g of sediments was ground and homogenized with a ceramic mortar and pestle, and then sealed in vials for gamma measurements. ²³⁴Th gamma emissions (63.3 keV) were measured within 10 weeks (<3 half-lives of ²³⁴Th) of sample collection using well-type high-purity germanium detectors for 8 h to 4 days until reaching a minimum of 400 counts. The gamma-spectra were analyzed using the Genie 2000 software (v2.1A) to calculate net peak areas while accounting for background and associated counting uncertainties (Canberra Industries, Inc., Meriden, CT, USA). Detectors were calibrated for the vial geometry by measuring a commercial standard solution (MCR-2009-018) of known gamma activities (60–1836 keV). In addition, the laboratory participates regularly in IAEA intercomparisons obtaining successful results for ²³⁴Th in solid materials. For each station, sediment samples from surface (0–0.5 cm), intermediate (4.5–5.0 cm) and deep (9.5–10 cm) layers were remeasured >5 months after sample collection to obtain a mean supported activity concentration of ²³⁴Th (²³⁴Th_eq; Supplementary Table S1) equivalent to its parent radionuclide ²³⁸U after reaching secular equilibrium.

2.4. Excess ²³⁴Th-Derived Calculations of Inventories and Mixing Rates

The excess ²³⁴Th activity concentration (²³⁴Th_xs) of each sediment core section was determined by subtracting ²³⁴Th_eq from the measured ²³⁴Th activity concentration (²³⁴Th_meas), and correcting for the decay occurred during the time (t) elapsed between sampling and the midpoint of counting, with $\mathsf{\lambda }$ referring to the decay constant of ²³⁴Th:

$${}^{234}\mathrm{T}\mathrm{h}_{\mathrm{x}\mathrm{s}}\left[\mathrm{B}\mathrm{q}{\mathrm{k}\mathrm{g}}^{-1}\right]=({}^{234}\mathrm{T}\mathrm{h}_{\mathrm{meas}}-{}^{234}\mathrm{T}\mathrm{h}_{\mathrm{e}\mathrm{q}})\times e^{(-\lambda t)}$$

(1)

The ²³⁴Th_xs penetration depth is defined as the deeper boundary of the core section where negligible ²³⁴Th_xs activity concentrations were observed when considering their uncertainties. ²³⁴Th_xs inventories were calculated as the sum of products of ²³⁴Th_xs activity concentrations in the core sections, its corresponding dry bulk density and sampling section height:

$${}^{234}\mathrm{T}\mathrm{h}_{\mathrm{x}\mathrm{s}}\mathrm{i}\mathrm{n}\mathrm{v}\mathrm{e}\mathrm{n}\mathrm{t}\mathrm{o}\mathrm{r}\mathrm{y}\left[\mathrm{B}\mathrm{q}{\mathrm{m}}^{-2}\right]={}^{234}\mathrm{T}\mathrm{h}_{\mathrm{x}\mathrm{s}}\times \mathrm{d}\mathrm{r}\mathrm{y}\mathrm{b}\mathrm{u}\mathrm{l}\mathrm{k}\mathrm{d}\mathrm{e}\mathrm{n}\mathrm{s}\mathrm{i}\mathrm{t}\mathrm{y}\times \mathrm{s}\mathrm{a}\mathrm{m}\mathrm{p}\mathrm{l}\mathrm{i}\mathrm{n}\mathrm{g}\mathrm{s}\mathrm{e}\mathrm{c}\mathrm{t}\mathrm{i}\mathrm{o}\mathrm{n}\mathrm{h}\mathrm{e}\mathrm{i}\mathrm{g}\mathrm{h}\mathrm{t}$$

(2)

²³⁴Th_xs of sections excluded from measurement were extrapolated as the mean of the adjacent upper and lower measured sections (Supplementary Table S1).

Although sedimentation rates in the Palamós canyon axis can be high (2.4 cm y⁻¹; [19]), they are still sufficiently low relative to the exponential decrease of ²³⁴Th_xs activity concentration in the sediment so that the ²³⁴Th_xs profile should be driven by diffusive mixing [35]. Hence, mixing rates (D_b) were calculated for each core as described by Schmidt et al. [28], which assumes negligible influence of sedimentation rates in the ²³⁴Th_xs activity concentration profile:

$${}^{234}\mathrm{T}\mathrm{h}_{\mathrm{x}\mathrm{s}}={}^{234}\mathrm{T}\mathrm{h}_{\mathrm{x}\mathrm{s}}^0\times {\mathrm{e}}^{\left(-\mathrm{z}\sqrt{{\displaystyle \frac{\mathsf{\lambda }}{{\mathrm{D}}_{\mathrm{b}}}}}\right)}$$

(3)

where z is the midpoint depth of each sediment section, ²³⁴${\mathrm{T}\mathrm{h}}_{\mathrm{x}\mathrm{s}}^0$ refers to the ²³⁴Th_xs activity concentration at the sediment–water interface, and $\mathsf{\lambda }$ denotes the decay constant of ²³⁴Th. The data were analyzed and visualized in Python (version 3.10.) using the packages pandas (v1.5.2), NumPy (v1.23.5), SciPy (v1.10.1), and Matplotlib (v3.6.2). Weighted least squares regression was performed on natural log-transformed ²³⁴Th_xs activity concentrations versus z, with weights assigned based on the associated uncertainties of the measurements using the statsmodels package (v0.14.4) (see regression metrics in Supplementary Table S1). D_b values for each station were calculated using Equation (2) with propagated slope uncertainties.

2.5. Evaluation of Riverine Discharge, Storm Occurrence, and Fishing Effort

Discharge data from the nearby Ter River (see gauging site in Figure 1), monitored daily and published online by the Agència Catalana de l’Aigua [32], was used to compare timings and intensity of potential riverine sediment input to the Palamós canyon head region during the four months (~120 days) prior to the midpoint of sampling (1 March 2023–29 June 2023 for ARCO-1; 22 April 2024–20 August 2024 for ARCO-2).

Significant wave height (Hs) data was retrieved from hourly logging at the Cap de Begur buoy (3.65° E, 41.90° N; Figure 1) collected and made publicly available within the Puertos del Estado monitoring network [36]. Storms within the same timeframe as the river discharge data were identified by prolonged sustained Hs of >2 m for a duration of >6 h [37]. Based on previous studies [17,18], easterly storms influence sediment transport in the study region more than the more frequent northerly storms. Therefore, storms were divided into easterly (60–120°) and northerly (330–30°) storm events based on their mean wave direction.

The cumulative fishing effort during the ~120 days preceding the midpoint of sampling, analogous to the time periods analyzed for river discharge and Hs, was obtained over the same region for 2023 and 2024, and used to evaluate the resuspension impacts by trawling activities along the canyon flanks (Figure 1). Monthly fishing effort (in h km⁻²) in the surroundings was analyzed and provided by ICATMAR based on vessel monitoring system (VMS) data of bottom trawlers of Palamós harbor for 2023 and 2024 [38]. A speed filter between 1.5 and 4.5 knots was applied to identify fishing activities, and their tracks were interpolated to 10 min using 2-hourly VMS data, considering the direction of navigation. Monthly fishing effort was obtained from 1 × 1 km grid cells on each flank (see grid cell size in Figure 1). The cells with the highest total fishing effort over the four months prior to sampling (>92 h km⁻²) on both fishing grounds revealed the main trajectory of fishing vessels. Seven cells per fishing ground, following these trajectories, were selected across years to calculate the cumulative effort for both flanks, accounting for the fraction of monthly data corresponding to the respective time period.

2.6. Global Compilation of ²³⁴Th_xs Parameters from Submarine Canyons

A global compilation was generated to explore the application of ²³⁴Th as a tracer for recent sedimentation in submarine canyons [39]. We conducted a search across Google Scholar up until 08 September 2025 for periods encompassing 1979–2025, and obtained 123 search results using the terms: (“Excess 234Th” OR “Excess 234 Th” OR “Excess Th-234” OR “Excess Thorium-234”) AND (“submarine canyon” OR “canyon” OR “off-shelf”) AND (“sediment core” OR “sediment samples” OR “core” OR “gamma spectrometry” OR “gamma spectroscopy” OR “gamma counting” OR “radiochemical analysis” OR “radioisotopic”). After duplicate removal, 121 publications were thoroughly inspected for content. Publications reporting sedimentary ²³⁴Th data in canyon environments were selected, resulting in a compilation of data from 25 publications. Data on ²³⁴Th_xs parameters, sampling methodology, and contextual information of sediment cores obtained in submarine canyon environments were carefully extracted using the information given in the main text, tables, figures, and supplementary files. Latitude, longitude, and sampling dates were assigned to the midpoint or the sampling month when not explicitly stated. Gamma spectrometry was applied as the counting method, with one exception measured by beta counting. In 12 studies, data was also provided from the shelf, slope, or abyssal plain near the canyon. If the data were not provided in tables, a web-tool was used for manual extraction [40]. The compilation includes surface ²³⁴Th_xs activity concentrations, ²³⁴Th_xs penetration depths, ²³⁴Th_xs inventories, and D_b from canyon studies with coring sites inside canyons and, additionally, near those canyons. The surface ²³⁴Th_xs activity concentrations correspond to mean values of the surface sediment section of a given thickness (0.1–2 cm) as reported in the original publication. If full profiles of ²³⁴Th_xs activity concentrations were given, but not the ²³⁴Th_xs penetration depth, this value was approximated using the criteria described above (see Section 2.4).

3. Results

3.1. Excess ²³⁴Th Profiles, Inventories and Mixing Rates

Surface ²³⁴Th_xs activity concentrations, penetration depths, inventories, and mixing rates (D_b) for all the collected cores are provided in Table 1 and described below in detail. Data show a general downslope decrease for all these variables. Comparing the cores from stations 1 to 3 across years revealed interannual variations in the ²³⁴Th_xs parameters, particularly at station 2 at the upper canyon (except for surface activity concentration), whereas the cores from station 1 at the canyon head and station 3 at the lower canyon showed comparatively smaller differences between years.

Table 1. Sampling information and ²³⁴Th data of sediment cores, including surface ²³⁴Th_xs activity concentration (0–0.5 cm section), ²³⁴Th_xs penetration depth, ²³⁴Th_xs inventories and mixing rates (D_b).

Sampling	Station	Water Depth	Surface 234Thxs Activity Conc.	234Thxs Penetration Depth	234Thxs Inventory	Db
		(m)	(Bq kg−1)	(cm)	(Bq m−2)	(cm2 y−1)
June 2023	1	764	378 ± 8	3.5	2232 ± 91	15.6 ± 5.5
	2 1	1368	521 ± 28	3.5	3140 ± 140	15.5 ± 2.4
	3	1779	258 ± 18	1.5	694 ± 53	0.8 ± 0.1
August 2024	1	785	908 ± 25	3.5	3490 ± 120	4.1 ± 2.2
	2	1364	620 ± 63	1.5 2	>1800 ± 190 2	1.4 ± 0.1
	3	1739	116 ± 12	1.5	522 ± 87	3.0 ± 1.0
	4	2141	23 ± 8	0.5	82 ± 26	N/A 3

¹ The ²³⁴Th_eq activity concentration in this core is the mean of the constant ²³⁴Th activity concentrations in the deepest sediment sections during the first measurements. ² Although extrapolation from adjacent sediment sections suggests a ²³⁴Th_xs penetration depth of 2 cm, this value is not considered because the 1.5–2 cm section was not measured. Therefore, a greater-than sign (>) indicates that the ²³⁴Th_xs inventory should be considered a lower-bound estimate. ³ This sediment core only presented detectable ²³⁴Th_xs activity concentration in the upper 0.5 cm, hindering the calculation of mixing rate. N/A, not applicable.

²³⁴Th_xs activity concentration was detected in the surface layer of all collected sediment cores and ranged between 908 ± 25 and 23 ± 8 Bq kg⁻¹ with the highest value recorded at station 1 and the lowest at station 4, located at the canyon mouth (Table 1; Figure 2). The decrease in surface ²³⁴Th_xs activity concentrations of the cores from station 1 to station 3 was 30% in 2023 and almost 90% in 2024. Comparing the two cruises, higher surface ²³⁴Th_xs activity concentrations were recorded in 2024 at station 1 (2.4-fold increase) and station 2 (1.2-fold increase) in comparison to 2023, whereas station 3 showed a lower surface ²³⁴Th_xs activity concentration in 2024 (2.2-fold decrease) compared to 2023.

Figure 2. Depth profiles of ²³⁴Th_xs activity concentrations for sediment cores collected (a) in 2023 and (b) in 2024.

Following the general downslope decreasing pattern, ²³⁴Th_xs penetration depths in the cores were greatest at station 1, reaching 3.5 cm depth, and then became shallower in the cores located further downcanyon, with depths of 1.5 cm at station 3 and 0.5 cm at station 4 (Table 1). The ²³⁴Th_xs penetration depth at station 2 varied between years, reaching 3.5 cm in 2023 and 1.5 cm in 2024. ²³⁸U activity concentrations (equal to supported ²³⁴Th activity concentrations; ²³⁴Th_eq), remeasured after the decay of ²³⁴Th_xs, varied between 18 ± 2 and 25 ± 3 Bq kg⁻¹ across all cores (Supplementary Table S1). In the 2023 station 2 core, first measurements of ²³⁴Th activity concentrations stabilized below 3.5 cm at a mean value of 41 ± 6 Bq kg⁻¹, which was used as ²³⁴Th_eq for this core (Supplementary Table S1).

Inventories of ²³⁴Th_xs showed a clear downcanyon decrease, ranging from 3490 ± 120 Bq m⁻² at station 1 to 82 ± 26 Bq m⁻² at station 4 (Table 1; Figure 3). Similar to surface ²³⁴Th_xs activity concentrations, ²³⁴Th_xs inventories were highest at stations 1 or 2 and lowest at stations 3 and 4. Higher ²³⁴Th_xs inventories were recorded in 2024 at station 1 (1.6-fold increase), whereas cores at station 2 (1.7-fold decrease) and station 3 (1.3-fold decrease) showed lower ²³⁴Th_xs inventories that year compared to 2023.

Figure 3. ²³⁴Th_xs inventories at stations 1 to 4 in 2023 (open bars) and 2024 (hatched bars). Error bars represent the inventory uncertainty of individual core measurements.

In both sampling cruises, stations 1 to 3 exhibited an exponential downcore decrease in ²³⁴Th_xs activity concentrations, which was used to calculate D_b (Table 1; Figure 4 and Figure 5; R² values provided in Supplementary Table S1). At station 4, no ²³⁴Th_xs signal was detected below 0.5 cm, and therefore D_b could not be determined. Across cores in both years, the highest D_b values were recorded in 2023 at station 1 (15.6 ± 5.5 cm² y⁻¹) and station 2 (15.5 ± 2.4 cm² y⁻¹). Compared to these, the D_b value at station 3 that same year was an order of magnitude lower (0.8 ± 0.1 cm² y⁻¹), mirroring the general downslope decreasing trend observed in the other ²³⁴Th_xs parameters. In 2024, D_b values at stations 1 and 2 (4.1 ± 2.2 and 1.4 ± 0.1 cm² y⁻¹, respectively) exhibited a 4- to 11-fold decrease compared to 2023. In contrast, at station 3, D_b increased 4-fold compared to 2023, reaching 3.0 ± 1.0 cm² y⁻¹ in 2024. That year, the along-canyon pattern of D_b was marked by its lowest mixing rate at station 2.

Figure 4. Depth profiles of exponential decrease of ²³⁴Th_xs activity concentration on log-scale used for the calculation of mixing rates (D_b) for 2023 and 2024 sampling campaigns showing (a) station 1, (b) station 2, and (c) station 3. Station 4 could not be sampled in 2023 and the sediment core collected in 2024 only presented detectable ²³⁴Th_xs activity concentration in the upper 0.5 cm, hindering the calculation of mixing rate.

Figure 5. ²³⁴Th_xs-derived mixing rates at stations 1 to 3 in 2023 (open bars) and 2024 (hatched bars). Error bars represent the rate uncertainty of individual core measurements. Station 4 could not be sampled in 2023 and in 2024 only presented detectable ²³⁴Th_xs activity concentration in the upper 0.5 cm, hindering the calculation of mixing rate.

3.2. River Discharge, Storms and Fishing Effort

Within a timeframe of ~120 days, equivalent to approximately five half-lives of ²³⁴Th prior to sampling, the Ter River discharge in 2023 reached maximum values of approximately 9 m³ s⁻¹ at the end of May and approximately 8 m³ s⁻¹ during June, in the weeks before sampling (Figure 6a). In 2023, numerous northerly storms were recorded from March to June, with maximum Hs of 4.2 m recorded at the beginning of March (Figure 6a). No easterly storms occurred during the studied period in 2023. During the same relative period in 2024, a maximum Ter river discharge of approximately 18 m³ s⁻¹ was registered at the beginning of May (Figure 6b); twice as high than the year before. More peaks of river discharge of approximately 7 m³ s⁻¹ occurred mid-May and at the end of July, whereas river discharge peaked again (~13 m³ s⁻¹) in the week leading up to sampling. Prolonged high easterly Hs were observed shortly before the highest peak in river discharge at the end of April 2024, comprising 10 h of elevated Hs of 2.0–2.5 m within a total time span of 13 h (Figure 6b). Northerly storms were observed from the end of April through August, with maximum Hs of 3.9 m at the end of April.

Figure 6. Ter River daily discharge and the occurrence of storms during the ~120 days prior to sampling (a) in June 2023 and (b) in August 2024. Storm events are indicated according to their duration by vertical shadowed areas in the timelines: easterly storms in green, northerly storms in light blue. A striped horizontal gray bar shows the timeframe of five half-lives of ²³⁴Th prior to sampling to differentiate more recent events.

The mean fishing effort over the ~120 days before sampling in 2023 (437 ± 6 h km⁻²) was considerably higher than that before sampling in 2024 (293 ± 105 h km⁻²). Also, the cumulative fishing effort along the canyon flanks prior to sampling was 33% higher in 2023 than 2024 (Figure 7). Comparing fishing effort in May and June, shared months within the studied periods and fishing seasons, shows that the fishing pressure was about twice as high on both flanks in 2023 as in 2024 (Figure 7c,d).

Figure 7. Fishing effort on the main trawling grounds along the flanks of the Palamós Canyon, shown as the sum of computed effort at 1 × 1 km resolution prior to samplings: (a) March–June 2023 and (b) May–August 2024. Sediment coring sites (1–3) are marked as white circles. Cumulative monthly fishing effort over the ~120 days prior to sampling (c) in June 2023 and (d) in August 2024, summed over cells outlined in blue in (a,b). Black shading indicates uncertainty from error propagation of sums.

4. Discussion

4.1. Spatial and Temporal Variations in Short-Term Sediment Deposition in Palamós Canyon

²³⁴Th_xs activity concentrations were detected in the surface sediment layer from all sampled sites along the canyon axis at water depths ranging from ~800 to ~2100 m, which confirms the recent deposition of sediment along the entire canyon axis and an active sediment supply to the canyon. The surface ²³⁴Th_xs activity concentrations and ²³⁴Th_xs inventories generally decreased downcanyon in both cruises (Figure 2 and Figure 3; Table 1) with the cores of the inner domain (stations 1 and 2) having higher values than those of the outer domain (stations 3 and 4; Figure 1). Unfortunately, some of the inherent fine spatial scale variability could not be accounted for in the reported ²³⁴Th_xs results due to the lack of analysis in the replicate cores of the multicorer deployments. According to the ²³⁴Th_xs inventories, the accumulation at the canyon mouth (station 4; ~2100 m) constitutes only 2% of the recent deposition registered at the canyon head (station 1; ~800 m) (Figure 3; Table 1). At the lower canyon (station 3; ~1800 m), the ²³⁴Th_xs inventories were lower by factors of up to 5 and 7 in 2023 and 2024, respectively, compared to the inner domain (Figure 3; Table 1). A previous study [41] has shown that the average total mass flux near the seafloor in the Palamós Canyon decreased downcanyon, agreeing with the data presented here.

While all sampled stations may receive particles both from horizontal advective fluxes through along-canyon transport and from vertical fluxes of particle settling through the water column, the intensity of these fluxes could impact recent sediment deposition. At the lower canyon, the ²³⁴Th_xs inventories showed lower interannual variability than the shallower stations of the inner domain. Interannual variability was observed at station 1, with a 36% lower ²³⁴Th_xs inventory in 2023 than in 2024 (Figure 3). This difference in ²³⁴Th_xs inventory between years at the head of the canyon, could have been caused by changes in riverine discharge or storms, but not by bottom trawling since the fishing grounds are found further offshore from station 1 (Figure 1). The maximal Ter river discharges within the four months before samplings were similar (9 m³ s⁻¹ in 2023; 18 m³ s⁻¹ in 2024; Figure 6) to the average annual mean value of 9 ± 30 m³ s⁻¹ over 2019–2024 [32], hence we consider that no major flooding occurred within those time periods. An easterly storm was recorded solely in the studied time period in 2024 (maximum Hs of 2.5 m), coinciding within a window of 5 days with the maximum peak in riverine discharge in that time period (~18 m³ s⁻¹). This combined storm and fluvial discharge peak event occurred approximately five half-lives of ²³⁴Th before sampling (Figure 6) and could have enhanced shelf resuspension and sediment transport to the canyon, which may have provided higher surface ²³⁴Th_xs activity concentrations and inventories in 2024 compared to 2023, particularly at station 1 due to its relatively short distance from the coast (Table 1). Yet, it remains unclear whether this event occurred too far back in time with respect to sampling to have had a measurable impact on ²³⁴Th_xs inventories, and whether shelf sediment transport into the canyon was efficient during the time before sampling. However, the second-highest peak of river discharge before sampling in 2024 (~13 m³ s⁻¹), which still exceeded the maximum discharge recorded during the previous year, occurred in the week leading up to our sampling cruise in mid-August, coinciding with a northerly storm event (Figure 6). This combination of increased riverine discharge and northerly storms could have provided a much larger measurable impact on ²³⁴Th_xs inventories at station 1, despite the absence of a concurrent easterly storm.

The upper canyon (station 2, ~1400 m) also showed interannual variability with a ²³⁴Th_xs inventory 1.7-fold higher in 2023 than in 2024 (Table 1). Contrary to the generally decreasing deposition with distance to the shelf observed in 2024, ²³⁴Th_xs inventories in 2023 suggest that sediment deposition was higher at station 2 than at station 1, thus disrupting the general along-axis decreasing trend in 2023 (Figure 3). In this occasion, the potential causes of interannual variability include bottom trawling activities, since in the Palamós Canyon this practice takes place in fishing grounds distributed along the northern and southern canyon flanks adjacent to station 2 (Figure 1). Fishing activities in the area have been shown to trigger nearly daily sediment gravity flows during the spring and summer months [16,19], thus potentially overruling the typically calm downslope sediment transport characterized by a lack of highly energetic hydrodynamic events. Indeed, these trawling-derived sediment gravity flows have led to an increase in sedimentation rates in the canyon axis at mid-canyon in the 1970s and the early 2000s [19]. This is especially evidenced at approximately 1200 m, at depths overlying station 2, given the presence of tributaries that can channel sediment into the canyon [16]. The higher cumulative fishing effort four months prior to sampling in 2023 in comparison to 2024 (Figure 7) could have transferred higher sediment fluxes from the fishing grounds towards the canyon axis, leading to higher ²³⁴Th_xs inventories in 2023 than in 2024 at station 2.

Based on VMS data, the fishing effort was higher by ~570 h km⁻² (1.5-fold higher) in the four months before sampling in 2023 than during the respective time period in 2024 (Figure 7), similar to the 1.8-fold enhanced ²³⁴Th_xs inventory at station 2 in 2023 compared to 2024 (Figure 3). The enhanced sediment deposition recorded in 2023 at station 2 could have also been influenced by the time passed between the opening of the fishing season (early March) and sampling (June 2023 vs. August 2024), as the progressive erosion of sediments along the flanks due to trawling reduces the magnitude and intensity of sediment gravity flows as the fishing season advances [16,18]. In addition, the fishing effort indicates that, during the four months before each sampling, the same area was consistently more intensely trawled in 2023 than in 2024 (Figure 7c,d).

4.2. Spatial and Temporal Variation in Mixing Rates in Palamós Canyon

The sedimentation rate in the Palamós canyon axis derived from ²¹⁰Pb measurements in a sediment core retrieved in 2011 nearby station 3 at 1820 m has been previously reported to range between 1.1 and 2.4 cm y⁻¹ [19]. Despite these relatively high sedimentation rates, given the short half-life of ²³⁴Th (24.1 days), a ²³⁴Th_xs signal should only be detectable at the surface (<1.0 cm) in the absence of mixing. If the ²³⁴Th_xs penetration depths of 1.5–3.5 cm at stations 1 to 3 were an indicator solely of sediment accumulation, the sedimentation rate over the last four months would be higher (4.5–10.5 cm y⁻¹) than the sedimentation rate derived from ²¹⁰Pb. Hence, the ²³⁴Th_xs profiles mainly reflect downward vertical mixing of varying intensity, potentially generated by the activity of burrowing organisms in the surface sediment [25]. Indeed, meiofaunal analyses [42], as well as burrowing tracks detected in the X-radiographs from that same sediment core retrieved nearby station 3 in 2011 [19] indicate that diverse burrowing fauna, capable of mixing the sediment, is inhabiting the Palamós canyon axis surface sediments.

In 2023, ²³⁴Th_xs penetration depths of 3.5 cm and mixing rates of ~16 cm² y⁻¹ were equally high at stations 1 and 2, and then lowered to ~1 cm² y⁻¹ at station 3 (Table 1). This downslope reduction in mixing rates may reflect a gradient of benthic activity along the canyon axis, aligning with earlier reports, which observed a decrease in meiofaunal assemblages with increasing water depth [42].

As observed with the ²³⁴Th_xs inventories, mixing rates also presented interannual variability. Compared to 2023, mixing rates in 2024 were lower by factors of 4 and 11 at stations 1 and 2, respectively, regardless of showing similarly deep ²³⁴Th_xs penetration depths (Table 1). Located below active fishing grounds, the 11-fold variability of mixing rates at station 2 between years may be influenced by variable inputs of sediment originating from the flanks’ regions and remobilized through trawling. At station 3, similar inventories were recorded in both years, while the mixing rate was approximately 4-fold higher in 2024 than in 2023 (Table 1; Figure 3 and Figure 5). Indeed, the quality of organic matter delivered into the canyon strongly impacts the magnitude of surface mixing [43]. In Cap-Ferret Canyon (Bay of Biscay), an elevated mixing rate (68.7 cm² y⁻¹) was observed at mid-canyon at 1035 m water depth, more than four times higher than the values reported here at similar depths, which was attributed to the input of fresh organic material from a spring bloom [44].

4.3. Comparison of Short-Term Deposition in the Canyon Axis with the Adjacent Flanks

²³⁴Th_xs inventories have been studied on the flanks of the Palamós Canyon in a recent study [33], suggesting that their magnitude is driven primarily by seasonal variations, rather than by the effect of bottom trawling activities. To assess the capacity of this canyon to focus sediment within its axis based on differences in ²³⁴Th_xs inventories between the axis and the flanks, it is therefore necessary to exclude the influence of seasonality. We hereafter compare our along-axis ²³⁴Th_xs data with previously published data from flank coring sites at ~500 m water depth, sampled slightly northward from our station 2 in June 2017 [33]. The flank cores were collected at approximately 15 km from the coast, at a similar distance from the coast as our station 2. Although the water depth in the axis site is about 900 m deeper than on the adjacent flank, the ²³⁴Th_xs inventories reported on the flanks (from 2320 ± 70 to 3182 ± 110 Bq m⁻² [33]) are within the range of values observed at ~1400 m water depth at station 2 in the canyon axis (from 1800 ± 190 to 3140 ± 140 Bq m⁻², Table 1). The deeper water depth and location below fishing grounds, would allow settling particulate matter to scavenge more ²³⁴Th during its downward transport through the water column at station 2, as well as additional scavenging of ²³⁴Th when particles are transported from the flank to the axis. However, considering the interannual variability in ²³⁴Th_xs inventories observed in the present study, quantitative comparisons of inventories at nearby locations may not be reliable, when measured in different years. In contrast to the seemingly similar inventories in this canyon in comparison with the flank, Schmidt et al. [28] found enhanced inventories inside Nazaré Canyon compared to just outside the canyon in the same sampling year.

Interestingly, in a study across the Poverty margin (New Zealand), Alexander et al. [45] found lower ²³⁴Th_xs inventories in surface sediments than expected based on the overlying water-column concentrations of the parent nuclide, ²³⁸U, once the water depth exceeded 600 m. While ²³⁴Th scavenging by particles depends largely on the amount of particulate matter available in the water column, other processes, such as lateral fluxes, may modulate the relationship between the water column ²³⁴Th flux and the ²³⁴Th_xs found in the seabed, and therefore ultimately influence the net amount of ²³⁴Th_xs in the sediment.

4.4. ²³⁴Th Studies in Submarine Canyons

Despite the numerous submarine canyons that incise continental margins on a global scale, the ²³⁴Th tracer has seldomly been used to improve our understanding of short-term sedimentary processes in and near canyons around the globe. A global compilation of studies that have used this tracer to assess recent sedimentation processes in submarine canyons identified a total of 26 studies since 1979 [39] including the present work. The compiled studies focus on canyons located along the continental margins of the Arctic, North America (Pacific Ocean), Gulf of Mexico, northeastern Atlantic, northwestern Mediterranean, southern Australia, and western and southwestern Pacific (locations shown in Figure 8a [21,22,23,28,29,33,44,45,46,47,48,49,50,51,52,53,54,55,56,57,58,59,60,61,62]).

Figure 8. Overview of a global compilation of sedimentary ²³⁴Th data in canyons since 1979. (a) Location of canyons. (b) ²³⁴Th_xs inventories [23,28,44,45,50,52,54,55,61] and (c) ²³⁴Th_xs-derived mixing rates [28,44,45,58,61] reported from canyons. Note that the mixing rates y-axis in (c) has a scale break. The Palamós Canyon is highlighted in red. Compiled data set available under https://doi.org/10.1594/PANGAEA.987505 [39].

A total of 329 ²³⁴Th_xs data values have been collected across four ²³⁴Th_xs parameters: surface activity concentration, penetration depth, inventory, and mixing rate. Thereof, 243 values were obtained from sediments within the canyons, spanning a depth range from 120 to 4280 m, and 86 values were obtained from their immediate surroundings. In canyons, the most frequently provided parameter is the surface ²³⁴Th_xs activity concentration (19 of 26 studies) ranging from 20 to 4040 Bq kg⁻¹ with a mean value of 520 Bq kg⁻¹. 10 of 26 studies reported ²³⁴Th_xs inventories, showing high variability with a range of values between 10 and 50,700 Bq m⁻² and a mean value of 2860 Bq m⁻² (Figure 8b). 10 of 26 studies reported or provided enough data for extraction of ²³⁴Th_xs penetration depths (mean of 2 cm, ranging from 0.4 to 24 cm). The least frequently reported ²³⁴Th_xs parameter is mixing rate (6 of 26 studies) yet encompassing a large range of values from 0.2 to 68.7 cm² y⁻¹ with a mean of 6.9 cm² y⁻¹ (Figure 8c). The range of surface ²³⁴Th_xs activity concentrations (116–908 Bq kg⁻¹), ²³⁴Th_xs penetration depths (0.5–3.5 cm), ²³⁴Th_xs inventories (82–3490 Bq m⁻²) and mixing rates (0.8–15.6 cm² y⁻¹) recorded in Palamós Canyon at ~800–2100 m depth fall within the values reported in the compiled studies.

Canyons have different sedimentary settings due to their geomorphology, as well as their natural and anthropogenic influences. The accumulation of sediment in canyons or specific parts of canyons on the timescale of ²³⁴Th can be caused by various processes [1]. Important seasonal drivers of sedimentation were observed in the compiled studies and include flooding events, storms, typhoons and hurricanes, while anthropogenic influence, such as contamination, caused by a petrogenic blowout, and bottom trawling, was also reported [23,33]. Events leading to enhanced particle flux and greater deposition of sediments in the canyon are expected to produce an increase in ²³⁴Th_xs inventories due to thorium’s high particle affinity. Comparing inside-canyon ²³⁴Th_xs inventories with those from adjacent margin sites, when studied simultaneously, can provide estimates of the relative input of fresh particles into the canyon [28].

Studying sedimentary ²³⁴Th_xs in canyons may reveal patterns of recent sediment deposition across the diverse geomorphologies and sediment transport mechanisms that occur in canyons across the globe. Most studies do not examine in detail the different ²³⁴Th_xs parameters (surface activity concentration, penetration depth, inventory and mixing rate) and usually focus on only two of them, indicating that the applicability of ²³⁴Th_xs as a tracer has not been fully explored. Despite the limited number of studies that have used ²³⁴Th in submarine canyons, the dynamic nature of the sedimentary processes taking place in these deep-sea environments warrants its applicability and shows the value of this tracer in understanding short-term sedimentation patterns that could otherwise potentially be missed.

5. Conclusions

²³⁴Th radiotracer-based analysis in the Palamós Canyon indicated a downslope decreasing trend of recent sediment deposition and mixing rates from the canyon head toward the canyon mouth over the spring and summer of 2023 and 2024. Short-term sediment fluxes and mixing rates appear to be strongly modulated by bottom trawling activities around the canyon flanks, as well as by discharge from the Ter River and by shelf sediment resuspension during storm events, as evidenced by the observed interannual variability in recent sediment deposition. This pronounced interannual variability complicates direct comparisons of ²³⁴Th_xs inventories and mixing rates among sediment cores collected in different years. Nevertheless, the short half-life of ²³⁴Th makes it a powerful tracer for capturing temporal variability in sediment fluxes within submarine canyons, where hydrodynamic forcing is highly variable across event to interannual timescales. In a broader context, the global compilation of ²³⁴Th studies in submarine canyons presented here demonstrates the value of this tracer in assessing short-term sediment deposition patterns within them, capturing the wide range of dynamic fluxes that characterize these complex deep-sea canyon environments.