Abstract
Introduction:
Reducing fuel consumption, hydrodynamic drag and seabed impact without compromising catch performance and size selectivity remains a key challenge for demersal trawl fisheries. This study presents a comparison of demersal trawl configurations designed to reduce energy demand and improve the environmental performance of a Mediterranean bottom trawl fishery.
Methods:
The analysis focused on two configurations of the four-panel otter trawl design known as the “Americana”: a traditional commercial configuration and a newly optimised experimental design developed through targeted modifications in mesh size distribution, netting material and belly geometry. The two gears were first evaluated through flume-tank trials using 1:10 scale models. These tests showed that the experimental trawl achieved a wider horizontal geometry and substantially lower towing load than the traditional design, supporting the expected reduction in drag associated with the decrease in twine surface area and projected resistance. Full-scale sea trials were then conducted in the central Adriatic Sea, Italy, under comparable operational conditions. A total of 19 comparative hauls were carried out, alternating between the traditional and experimental configurations. Fuel consumption, vessel speed, towing loads at the winch and wing-ends, net geometry and GPS data were collected using a harmonised instrumentation system. Catch data were also recorded to assess changes in fishing performance, catch composition and size-dependent catch efficiency. A dedicated lift test was used to derive a vessel-specific speed -fuel consumption model and to partition total fuel use during towing into vessel-related and gear-related components.
Results and conclusions:
Compared with the traditional gear, the experimental trawl showed greater horizontal net spread and markedly reduced vertical opening. These changes resulted in a 26.7% reduction in hydrodynamic drag and a 15% decrease in fuel consumption during towing. When fuel use was partitioned, gear-related fuel demand showed an even larger relative reduction of 29.7%. Total catch rates did not differ significantly between gear configurations. Catch comparison analyses indicated species-specific responses, with higher catch efficiency for pelagic and semi-pelagic species and reduced retention of smaller individuals for some demersal species. Overall, the results demonstrate that targeted trawl modifications can improve energy efficiency and potentially reduce seabed interaction while maintaining commercially acceptable catch performance.
Introduction
The transition towards sustainable fisheries management requires the integration of ecological, economic, and technological innovations capable of reducing environmental impacts while maintaining the viability of fishing operations. Demersal trawl fisheries remain central to global seafood production, accounting for a large share of marine landings and sustaining thousands of coastal communities (, ). However, mobile bottom-contact gears are also among the most energy-intensive fishing methods, contributing to significant greenhouse gas (GHG) emissions, high fuel dependency, and physical disturbance of seabed habitats. Recent global assessments further highlight that trawl fisheries exhibit some of the highest fuel requirements per unit of landed biomass (, ).
Fuel use is not just an environmental issue; it is also a key factor shaping the economic resilience of demersal fleets. In many demersal fleets, fuel can represent up to 20–40% of total operational costs, and exposure to fuel-price volatility can substantially undermine fleet profitability and long-term stability (–). Periods of fuel price spikes have historically resulted in reduced fishing effort, fleet restructuring, and, in some cases, temporary inactivity of smaller-scale vessels (, ). Consequently, improving fuel efficiency through reductions in hydrodynamic drag has become both an environmental priority and a strategic business objective.
Hydrodynamic drag in demersal trawls results from the combined effects of several components, including trawl length, cross-section area of the trawl body, netting area, twine diameter, mesh configuration, groundgear, and otterboards. Both experimental and modelling work show that modest reductions in twine diameter or netting area can yield measurable decreases in drag and fuel demand (–). However, many of these investigations have been conducted under controlled conditions or at model scale, while fewer studies have quantified the performance of modified gears under full-scale commercial conditions.
Beyond fuel consumption, bottom trawling has been widely discussed in relation to seabed disturbance, sediment resuspension, and potential impacts on benthic communities and biogeochemical processes. Meta-analyses have shown that mobile bottom fishing can alter benthic habitat structure and reduce biomass of sensitive taxa, with effects varying according to gear configuration and sediment type (, ).
More recently, growing attention has turned to how sediment disturbance may influence carbon cycling, including the risk of remineralisation of stored organic carbon under certain environmental conditions (, ). In this context, gear designs that reduce seabed contact and penetration may help limit benthic disturbance and biodiversity impacts, while also reducing the contribution of demersal fishing activities to climate-related pressures.
Despite growing recognition of these issues, a persistent gap remains between experimental gear development and demonstrable, auditable improvements under commercial fishing conditions.
Empirical evidence quantifying the proportion of total fuel use directly attributable to the gear, rather than baseline propulsion, remains limited (, ). Disentangling these components is essential for accurately assessing the environmental benefits of gear modifications and for enabling transparent reporting frameworks that could support eco-labelling, incentive schemes, or carbon accounting in fisheries.
In this study, we evaluate the performance of an experimentally optimised demersal trawl developed within this framework under commercial fishing conditions. The redesign aimed to reduce hydrodynamic drag, fuel consumption, and seabed interaction while maintaining fishing performance compatible with commercial requirements, through targeted modifications in mesh size distribution, netting material, and belly geometry intended to decrease projected netting area and twine surface exposed to flow.
Specifically, the experimental trawl was compared with a traditional commercial trawl typical of the Mediterranean bottom trawl fishery. This was first done in the flume tank with small-scale models of the experimental and traditional gears to validate the design modifications and to assess changes to drag and gear geometry. Subsequently, full-scale trials at sea were carried out with the aim of quantifying differences in towing drag and fuel consumption and assessing whether reductions in hydrodynamic resistance translate into measurable energy savings without compromising commercially relevant catch performance, and evaluating associated changes in catch comparison and size structure.
Materials and methods
Flume tank and scale models
Experimental trials with 1:10 scale models of the traditional and experimental gears were carried out at Flume Tank North Sea A/S, Hirtshals, Denmark. The flume tank is a circulating water tank equipped with four 55 kW propellers, which can drive 1200 m³ of water up to a speed of 1 m s−1. The working section is 21.3 m long, 8.0 m wide and 2.7 m deep. A schematic representation of the experimental setup is provided in Supplementary Figure 1.
During the tests, the scale model was held in position while water flow was circulated through the tank at model-scale velocities corresponding to the full-scale towing speeds. In Supplementary Figure 1, the flow direction is shown from left to right. The water flow acted on the model trawl and otterboards, generating the horizontal gear spread and vertical headline height. The tank floor consists of a textile-based nylon conveyor belt, which can be driven at speeds up to 0.7 m s−1. The belt was operated at the same speed as the water flow in order to simulate the relative movement between the gear and the seabed and to reproduce seabed-contact conditions at model scale.
Gear geometry was measured using the movable camera system installed in the flume tank. Two top-view cameras were used to measure door spread and groundgear spread, whereas a side-view camera was used to measure headline height. Camera positions were adjusted according to the geometry of each model before testing and were then kept fixed throughout each test.
Each gear was tested at four model-scale current velocities: 0.40, 0.50, 0.55, and 0.65 m s−1, corresponding to full-scale towing speeds of approximately 1.26, 1.58, 1.74, and 2.06 m s−1 (2.45–4.00 knots), respectively, where the velocity scale = √(model scale) (here = 1/√10) (). The bridle tensions, measured just aft of the doors, were recorded for a 5-minute period using two Futek load cells (1070091 100lb 2024.01, 200 Hz), which were calibrated each day. Average trawl drag was estimated by summing the resolved bridle tensions.
Study area and survey design
The full-scale study was conducted in the central Adriatic Sea, within FAO Geographical Sub-Area 17 (GSA 17), off the coast of Civitanova Marche (Italy) (Figure 1). This area is characterised by extensive sand-muddy seabed habitats that support some of the most important demersal bottom-trawl fisheries in the Mediterranean Sea. The fishery targets mixed assemblages of fish and invertebrates and is representative of typical Mediterranean otter-trawl operations on soft bottoms.
Figure 1
Sea trials were carried out during a five-day survey between 25 June and 2 July 2025 on board the commercial bottom otter trawler Albatros Selvaggio (length overall: 22 m; beam: 4 m; engine power: 536 kW), home-ported in Civitanova Marche (Supplementary Figure 2).
Two additional preparatory days were devoted to the installation, calibration, and verification of all onboard monitoring systems prior to the start of fishing activities. Environmental conditions during the survey were stable and typical of early summer in the central Adriatic, allowing consistent towing speeds and comparable seabed contact conditions across hauls.
A total of 19 trawl hauls were conducted, alternating between the vessel’s conventional commercial trawl (hereafter referred to as “traditional” gear) and an experimentally optimised trawl.
Comparative fishing trials were conducted over five survey days, during which 2–5 hauls were completed per day. Detailed information on each fishing operation, including survey day, haul number, gear type, mean fishing depth, haul duration, sea state and weather conditions, is provided in Supplementary Table 1.
The experimental design followed an unpaired comparative approach, with hauls conducted on the same fishing grounds, at comparable depths (14–32 m), and under similar operational conditions and in subsequent survey days in order to minimise environmental and behavioural bias between gear types. Each haul lasted approximately 60–90 minutes (mean duration ≈ 1.5 h). This approach reflects the constraints of full-scale commercial trials, while ensuring comparability through consistent fishing grounds, depth ranges, and towing practices.
For each haul, a harmonised suite of operational, mechanical, and catch-related variables was recorded, including fuel consumption, both instantaneous (l h−1) and cumulative (l); navigation data, including GPS position, propulsion engine speed (RPM), speed over ground (kn), and course (degrees); towing loads measured at the winch and at the wing ends/sweeps (kg); net geometry, including horizontal and vertical net opening (m); haul descriptors, including start and end time, duration (min), depth (m), and coordinates (decimal degrees); and catch metrics, including total catch weight and species-specific weights (kg).
This sampling design ensured that differences in performance could be attributed primarily to gear configuration rather than to external environmental or operational variability.
Fishing gears
The gears tested were configured as “Americana” bottom otter trawls (FAO gear code: OTB4), which represents the most widespread trawl design used in Mediterranean demersal fisheries. These four-panels trawls are typically constructed from polyethylene netting, rigged with short sweeps and bridles (). The groundgear commonly includes a fishing line and a tickler chain to ensure effective seabed contact and to enhance the capture of mixed demersal species on sandy substrates.
The traditional gear corresponded to the vessel’s standard commercial configuration and served as the baseline for performance comparisons (Figure 2).
Figure 2
The experimental gear, in contrast, was developed within the RightFish project (Grant Agreement No. 817992) through an iterative process combining evidence synthesis, flume-tank experiments, and component-level measurements of drag and seabed penetration (Figure 3).
Figure 3
Key design modifications implemented in the experimental gear included mesh size redistribution, with larger meshes in selected sections of the net body to facilitate water flow and reduce hydrodynamic resistance; netting material optimisation, through the use of high-tenacity, thinner twines in the upper panels, decreasing the total twine surface area exposed to flow; and redesign of the belly geometry, resulting in a narrower and more streamlined configuration with reduced frontal area and improved flow alignment. The design modifications reduced the total twine surface area from 118 m2 in the traditional trawl to 89 m2 in the experimental trawl, corresponding to a 24.6% reduction.
These modifications are visually summarised in Figure 4.
Figure 4
Such design principles are consistent with previous studies demonstrating that reductions in netting area, twine diameter, and projected frontal surface can lead to substantial decreases in trawl drag and associated fuel consumption (–). However, full-scale validation of these concepts under realistic commercial conditions remains limited, particularly in Mediterranean fisheries, motivating the present study.
Instrumentation and data acquisition
A comprehensive, time-synchronised monitoring system was deployed to measure vessel performance, gear mechanics, and net geometry throughout all hauls (Supplementary Figures 3–S8).
Fuel consumption was measured using a KRAL BEM 500 fuel flow-metering system integrated with the vessel’s engine monitoring unit, providing high-resolution records of instantaneous and cumulative fuel use (Supplementary Figure 4).
Navigation data, including geographic position, speed over ground, and course, were acquired from the vessel’s GPS system at a temporal resolution of 10 s. Fishing depth refers to the water depth of the fishing ground along which the gear was towed; it was considered constant within each haul because trawling was conducted along isobaths, and was measured using the vessel’s Furuno echosounder. Towing forces were measured using calibrated load cells installed at two key locations: at the winch, to measure the total towing force (TTF), which includes the combined resistance of the trawl, otterboards, and towing warps (Supplementary Figure 5); and at the wing-ends (sweeps), using underwater load cells to quantify the drag generated by the rigged net alone (gear drag) (Supplementary Figure 6).
This configuration allowed a clear separation between the drag generated by the rigged net alone (gear drag) and the total towing resistance of the fishing system, including otterboards and towing warps (TTF).
Net geometry was continuously monitored using acoustic sensors. Horizontal net opening and headline height were measured using Simrad PX sensors (Supplementary Figure 7), while a Simrad TV80 gear-monitoring system provided real-time visualisation and logging of horizontal and vertical opening throughout each haul (Supplementary Figure 8).
All sensors were calibrated prior to the survey following manufacturer specifications and all instrumentation systems were synchronised in time, enabling precise alignment of fuel consumption, towing loads, navigation data, and net geometry along the haul timeline.
This integrated data-collection framework follows established best practices for full-scale assessments of trawl gear performance under commercial fishing conditions.
Fuel use intensity
Fuel consumption data for the traditional gear were recorded in time steps of 5–38 minutes (median of 15 minutes and mean value of 14.7 minutes). Fuel consumption data for the experimental gear were recorded in time steps of 4–28 minutes (median of 13 minutes and mean value of 14.1 minutes). We assumed that the fuel consumption value recorded at each data point remained constant until the next data point. Using these data in connection with round weight of catch for each haul, we explored changes in fuel use intensity (FUI; Litres of fuel per kg of catch). Haul 17 was excluded from the FUI analysis because abnormal catch conditions associated with mucilage presence affected the representativeness of catch weight.
In fisheries, FUI is commonly used as an indicator of energy efficiency: a high FUI indicates low efficiency, while a low FUI reflects better energy performance. When evaluating changes to fishing gear, it is therefore important to consider the effects on both fuel consumption and catch simultaneously. Focusing only on fuel use can be misleading since a reduction in fuel consumption is not necessarily beneficial if it also leads to a significant decrease in catch. FUI captures this balance by relating fuel use directly to output, making it a more meaningful indicator of overall performance (–).
We used Tukey’s boxplots to display the data distribution, with the boxes representing the lower quartile (Q1), median, and upper quartile (Q3) of values. The whiskers represent the lowest data within a 1.5 interquartile range [i.e., 1.5 (Q3-Q1)] of Q1 and the highest data within a 1.5 interquartile range of Q3. Data visualisation was performed in R [version 4.5.2; ()] using the ggplot2 package [version 4.0.2; ()].
Lift test and vessel fuel–speed model
To isolate the contribution of the fishing gear to total fuel consumption, a dedicated vessel “lift test” (speed test) was conducted. The test consisted of a series of navigation runs at progressively increasing speeds, performed in opposite directions to minimise the influence of wind and currents. During these trials, vessel speed, engine revolutions per minute (RPM), and fuel consumption were continuously recorded.
The resulting dataset was used to derive a vessel-specific relationship between speed and fuel consumption under navigation conditions without towing gear.
This relationship was modelled using polynomial regression, yielding a high coefficient of determination (R2 > 0.99). The resulting vessel speed–fuel consumption (VS–FC) model represents the baseline energy demand of the vessel and was subsequently used to partition fuel use during trawling into vessel-related and gear-related components. This partitioning approach enabled a direct comparison of the energetic performance of the two gear configurations independently of vessel hull resistance.
Vessel-specific fuel–speed models are widely applied in fisheries energy audits and gear-performance studies to disentangle propulsion-related energy use from resistance induced by towed fishing gears (, ).
Data processing and performance indicators
Gear configuration parameters and haul-averaged performance metrics from the 19 comparative hauls were summarised by gear type (traditional versus experimental). The primary variables considered in the analysis were:
trawling speed (TS, kn);
engine speed (RPM);
horizontal net opening (HNO, m);
vertical net opening (VNO, m);
fishing depth (m);
total fuel consumption (FC, L/h);
gear drag measured at the sweeps (kg);
total towing force at the winch (TTF, kg).
Fuel consumption was further partitioned into vessel-related and gear-related components. For each haul, fuel consumption attributable solely to the vessel at the observed towing speed (FC_vessel) was estimated using the VS–FC model derived from the lift test. Gear-related fuel consumption (FC_gear) was calculated as the difference between total observed fuel consumption during towing (FC_total) and the modelled vessel-only fuel demand at the corresponding speed (FC_vessel).
Corresponding CO2 emissions were estimated using a standard emission factor for marine diesel fuel (2.64 kg CO2 L−1), following established methodologies for carbon-footprint assessment in trawl fisheries (, 27).
Data visualisation was performed in R [version 4.5.2; ()] using the ggplot2 package [version 4.0.0; ()].
Catch data
Catch data were used to provide an operational comparison of fishing performance and to assess whether the observed reductions in drag and fuel consumption were achieved without compromising overall catch quantity or the capture of key commercial species.
Catch sampling focused on species composition, catch rates and size distributions. Discard fractions were not quantified separately during the sea trials and were therefore not included in the present analysis.
Catch composition was recorded for each haul and summarised as total catch weight (kg) by species for both the experimental and traditional gears. One experimental haul (haul 17) was excluded from the biological analyses due to abnormal catch conditions associated with mucilage presence, which compromised comparability with the remaining hauls. Consequently, catch analyses were performed on nine hauls per gear type. After each haul, the catch was processed according to standard commercial procedures. Researchers then sorted the retained commercial fraction by species, identifying all taxa to the lowest possible taxonomic level. For each species or genus, the total weight was recorded. Particular attention was paid to the retention of main target and most abundant species (e.g. red mullet, Mullus barbatus) for which individual length measurements were collected as total length, to the lowest 1 cm for fish, and as carapace length, to the lowest 1 mm, for crustaceans.
Counts and measurements were performed using subsampling in the case of the most abundant catches. In such cases, abundance estimates were derived from a randomly selected subsample, applying an appropriate raising factor.
To assess differences in catch rates between the two groups (e.g., gear types/experimental conditions), a non-parametric Wilcoxon rank-sum test (Mann–Whitney U test) was applied to CPUE values expressed both as biomass (kg/h) and abundance (n/h). The test was performed independently for each target species (Squilla mantis, Mullus barbatus, Trachurus spp., Scomber spp.) as well as for total catch. This approach was selected because CPUE data did not meet normality assumptions and sample sizes were limited, making a distribution-free test more appropriate. Statistical significance was assessed at α = 0.05.
Catch comparison and catch ratio analysis
Length-frequency distributions of the most commercially important species were analysed using catch comparison and catch ratio methods. The objective was to assess whether the modified gear configuration resulted in length-dependent differences in catch efficiency relative to the traditional gear. All analyses were performed using the software package SELNET (28). Because the two gear configurations were not alternated on a haul-by-haul basis during the experiment, the datasets were treated as unpaired, following the methodology described by Herrmann et al. (29), rather than as paired-haul observations. For each species analysed separately, the length-dependent catch comparison function, CC(l), was estimated. This function describes the probability that a fish of length l is retained by the experimental gear relative to the traditional gear. The catch comparison rate was calculated according to Equation 1 (29):
where nblj and ntlj are the numbers of fish of length l from haul j retained by the traditional gear (b) and the experimental gear (t), respectively. The parameters qbj and qtj represent the subsampling fractions, defined as the proportion of measured individuals relative to the total catch in each haul conducted with the traditional and experimental gears, respectively. Similarly, hb and ht denote the total number of hauls performed with the traditional and experimental gear configurations. The experimental catch comparison function, CC(l,v), defined in Equation 1, was estimated using a maximum-likelihood approach by minimising the negative log-likelihood function shown in Equation 2:
The outer summation is performed over length classes (l), while the inner summation accounts for the individual hauls of each gear type. The parameters v describes the catch comparison curve defined by CC(l,v). The function CC(l,v) was modelled using a logistic function (Equation 3):
where f is a polynomial function of order k, with coefficients v0 to vk. Polynomial orders up to 4 were considered. By excluding one or more polynomial terms, a total of 31 additional candidate models were generated. Model selection was performed using multi-model inference following (29, 30), resulting in a combined model representing the best-supported description of the data. Model performance was evaluated using the deviance-to-degrees-of-freedom ratio (deviance/DOF) and the corresponding p-value (31). A p-value greater than 0.05 was considered indicative of an adequate fit between the model and the observed data. In contrast, when poor goodness-of-fit statistics were obtained (p-value < 0.05 and deviance/DOF substantially greater than 1), residual patterns were inspected to determine whether the lack of fit resulted from model misspecification or from overdispersion in the data. The catch comparison function CC(l,v) expresses the probability that a fish of length l is retained by the experimental gear, given that it was retained by either of the two gears being compared. Since an equal number of valid hauls was conducted with each gear type (9 traditional and 9 experimental hauls), a value of CC(l,v) = 0.5 indicates identical retention probabilities for the two gear configurations. Although CC(l,v) provides information on the relative retention probability of the two gears, it does not directly quantify differences in catch efficiency. To facilitate interpretation, the catch ratio function, CR(l,v), was derived according to Equation 4:
Under this formulation, CR(l,v) = 1.0 indicates equal catch efficiency between the two gear configurations, whereas CR(l,v) = 0.25 indicates that the experimental gear retains only 25% of the fish of length l retained by the traditional gear. The uncertainty associated with both CC(l,v) and CR(l,v) estimates was assessed by calculating 95% confidence intervals using a double-bootstrap procedure with 1,000 resampling iterations. This approach accounts for variability arising both within and between hauls and follows the methodology described by Lomeli (32).
Results
Flume tank experiments
Flume tank experiments confirmed that the experimental trawl achieved a wider horizontal geometry and lower towing resistance than the traditional design (Figure 5). When raised to full scale, the experimental model showed a 7% increase in door spread and a 3% increase in horizontal net opening (wing spread), together with a 6% reduction in headline height. Towing load was on average 25.9% lower than for the traditional gear, closely matching the reduction in twine surface area introduced in the experimental design.
Figure 5
Gear performance, net geometry, and hydrodynamic drag
The differences in net geometry between the traditional and experimental trawl configurations were greater during the full-scale trials (Table 1).
Table 1
| Type | Haul nr. | TS[kn] | RPM | HNO[m] | VNO[m] | Gear Drag[kg] | TTF [kg] | FC.All [l/h] | FC.Vessel [l/h] | FC.Gear [l/h] | CO2.All [kg/h] | CO2.Vessel [kg/h] | CO2.Gear [kg/h] |
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| Traditional | 1 | 3.60 | 1154 | 24.70 | 1.54 | 3485.50 | 4259.17 | 62.73 | 27.00 | 35.73 | 165.60 | 71.27 | 94.33 |
| 2 | 3.79 | 1120 | 23.60 | 1.76 | 3750.16 | 4358.43 | 57.62 | 27.65 | 29.98 | 152.13 | 72.99 | 79.14 | |
| 3 | 3.60 | 1151 | 22.92 | 1.68 | 3854.65 | 4373.29 | 62.23 | 27.00 | 35.23 | 164.28 | 71.27 | 93.01 | |
| 4 | 3.79 | 1140 | 22.74 | 1.93 | 3749.12 | 4440.11 | 60.59 | 27.66 | 32.93 | 159.96 | 73.02 | 86.95 | |
| 5 | 3.42 | 1189 | 22.35 | 1.89 | 4314.42 | 4552.67 | 68.49 | 26.33 | 42.16 | 180.81 | 69.52 | 111.29 | |
| 6 | 3.63 | 1158 | 20.00 | 2.24 | 4530.02 | 4819.17 | 63.44 | 27.11 | 36.32 | 167.47 | 71.58 | 95.89 | |
| 7 | 3.50 | 1176 | 20.73 | 2.15 | 4113.95 | 4478.20 | 66.33 | 26.64 | 39.69 | 175.11 | 70.32 | 104.79 | |
| 8 | 3.50 | 1158 | 20.20 | 2.20 | 4106.94 | 4580.00 | 63.41 | 26.64 | 36.77 | 167.40 | 70.32 | 97.07 | |
| 9 | 3.50 | 1134 | 20.31 | 2.06 | 3966.37 | 4677.50 | 59.62 | 26.64 | 32.99 | 157.41 | 70.32 | 87.09 | |
| Experimental | 10 | 3.97 | 1068 | 27.40 | 0.80 | 2858.41 | 3862.00 | 50.42 | 28.27 | 22.15 | 133.11 | 74.62 | 58.48 |
| 11 | 3.97 | 1069 | 26.02 | 0.94 | 2615.97 | 3479.33 | 50.44 | 28.27 | 22.18 | 133.17 | 74.62 | 58.55 | |
| 12 | 3.91 | 1072 | 25.61 | 1.06 | 2883.77 | 3703.11 | 50.86 | 28.08 | 22.79 | 134.28 | 74.12 | 60.16 | |
| 13 | 3.99 | 1102 | 25.69 | 1.00 | 2951.05 | 3785.75 | 54.99 | 28.34 | 26.65 | 145.17 | 74.81 | 70.36 | |
| 14 | 4.03 | 1121 | 25.50 | 1.04 | 2981.83 | 3758.33 | 57.65 | 28.49 | 29.16 | 152.20 | 75.22 | 76.98 | |
| 15 | 3.83 | 1085 | 25.38 | 0.97 | 2672.01 | 3469.50 | 52.66 | 27.81 | 24.85 | 139.02 | 73.42 | 65.60 | |
| 16 | 3.83 | 1088 | 25.38 | 0.94 | 3162.98 | 3871.67 | 53.09 | 27.81 | 25.28 | 140.16 | 73.42 | 66.74 | |
| 17 | 3.93 | 1077 | 24.50 | 1.10 | 2917.40 | 3716.67 | 51.58 | 28.15 | 23.43 | 136.18 | 74.32 | 61.86 | |
| 18 | 4.06 | 1106 | 24.80 | 1.00 | 3091.40 | 4102.00 | 55.53 | 28.58 | 26.94 | 146.59 | 75.46 | 71.13 | |
| 19 | 3.96 | 1109 | 24.10 | 1.10 | 3096.76 | 3702.00 | 56.04 | 28.24 | 27.80 | 147.95 | 74.56 | 73.39 | |
| Average | |||||||||||||
| traditional | 3.59 | 1153 | 21.95 | 1.94 | 3985.68 | 4504.28 | 62.72 | 26.96 | 35.76 | 165.57 | 71.18 | 94.40 | |
| experimental | 3.95 | 1090 | 25.44 | 0.99 | 2923.16 | 3745.04 | 53.33 | 28.20 | 25.12 | 140.78 | 74.46 | 66.33 | |
| 9.9% | -5.5% | 15.9% | -48.7% | -26.7% | -16.9% | -15.0% | 4.6% | -29.7% | -15.0% | 4.6% | -29.7% | ||
Summary of data collected during sea trials.
TS, trawling speed; RPM, revolution speed of the propulsion engine; HNO horizontal net opening; VNO, vertical net opening; Gear Drag, drag measured at the sweeps of the trawl; TTF, total towing force (kg). FC.All, fuel consumption; FC.vessel, fuel consumption of the solely vessel at the specific speed; FC.gear, fuel consumption of the gear, as subtraction of the FC.vessel from FC.All; CO2.All, total CO2 emission; CO2.vessel, CO2 emissions of the vessel; CO2.gear, CO2 emissions proper of the gear.
Legend for the sea-trial variables in the lower section.
When averaged over the 19 comparative hauls, the experimental trawl operated with a consistently larger horizontal net opening, while vertical net opening was substantially reduced relative to the conventional gear. On average, horizontal spread increased by approximately 15.9%, whereas vertical opening decreased by about 48.7%.
These geometric differences, combined with the 24.6% reduction in twine area, translated into a marked reduction in towing resistance. Drag measured at the wing-ends (gear drag) was, on average, 26.7% lower for the experimental trawl compared with the traditional configuration, which corresponds very well with the reduction observed during the flume tank trials. A consistent reduction was also observed in total towing force measured at the winch, which decreased by approximately 16.9%, reflecting the combined effect of reduced net drag and overall system resistance.
The reduction in hydrodynamic drag resulted in a substantial decrease in fuel consumption during towing. Mean total fuel consumption was reduced by approximately 15% when operating the experimental gear.
This pattern is further illustrated in Figure 6, which shows the relationship between towing speed and fuel consumption for the two gear configurations, with the experimental trawl consistently operating at lower fuel consumption levels despite slightly higher towing speeds.
Figure 6
When fuel use was partitioned using the vessel speed–fuel consumption model, the fuel demand attributable to the gear itself showed an even larger relative reduction (29.7%), partially offset by a small increase in vessel-related fuel consumption associated with the slightly higher towing speed adopted with the experimental trawl.
Net-geometry readouts and drag distribution along the gear indicate reduced bottom contact and penetration for the experimental configuration. Although seabed interaction was not measured directly, the observed combination of lower vertical opening and reduced gear drag is consistent with a slightly higher off-bottom fishing geometry.
Fuel use intensity
Figure 7 uses Tukey’s boxplots to display the changes in total fuel consumption (for the vessel and gear), catch, and FUI for the hauls retained in the FUI analysis. Haul 17 was excluded from this analysis. By switching to the experimental gear, mean and median fuel consumption decreased by 8.41% and 3.17%, respectively. Mean and median round catch weight per haul were also lower for the experimental gear by 8.13% and 18.58%, respectively, although statistical tests on CPUE indicated no significant difference in total catch rates between gear configurations. The combined effect resulted in 1.65% lower mean FUI and 16.79% higher median FUI.
Figure 7
Vessel performance and lift test results
The lift test provided a vessel-specific reference for interpreting towing performance by characterising the relationship between vessel speed, engine load, and fuel consumption under non-towing conditions. Fuel consumption increased non-linearly with vessel speed, as expected, and polynomial regression models fitted to the lift-test data showed a high coefficient of determination (R2 > 0.99) (Figure 8).
Figure 8
Bidirectional runs conducted during the lift test reduced the influence of wind and currents, allowing a robust estimation of baseline vessel fuel demand. This vessel speed–fuel consumption relationship was subsequently used to partition fuel use during fishing operations into vessel-related and gear-related components, enabling a direct comparison of the energetic performance of the two gear configurations.
Catch rates
The results of the Wilcoxon rank-sum tests are presented in Table 2 (CPUE expressed as kg/h) and Table 3 (CPUE expressed as n/h).
Table 2
| Wilcoxon rank-sum test (Mann–Whitney) - CPUE (kg/h) | |||||
|---|---|---|---|---|---|
| Traditional | Experimental | ||||
| Species | n | Mean ± SD | n | Mean ± SD | p-value |
| Squilla mantis | 9 | 0.70 ± 0.89 | 9 | 0.23 ± 0.34 | 0.084 |
| Mullus barbatus | 9 | 6.36 ± 5.51 | 9 | 2.51 ± 4.05 | 0.051 |
| Trachurus spp. | 9 | 0.90 ± 0.59 | 9 | 0.89 ± 0.96 | 0.536 |
| Scomber spp. | 9 | 0.71 ± 0.91 | 9 | 2.10 ± 1.68 | 0.034 |
| Total | 9 | 8.66 ± 6.43 | 9 | 5.73 ± 4.59 | 0.387 |
Results of the Wilcoxon rank-sum (Mann–Whitney U) test comparing CPUE expressed as biomass (kg/h) between the two gear configurations for each species and for the total catch.
For each gear configuration, sample size (n) and mean ± standard deviation (SD) are also reported. P-values< 0.05 were considered statistically significant.
Table 3
| Wilcoxon rank-sum test (Mann–Whitney) - CPUE (n/h) | |||||
|---|---|---|---|---|---|
| Traditional | Experimental | ||||
| Species | n | Mean ± SD | n | Mean ± SD | p-value |
| Squilla mantis | 9 | 22.66 ± 30.83 | 9 | 8.48 ± 11.09 | 0.143 |
| Mullus barbatus | 9 | 148.69 ± 135.89 | 9 | 60.12 ± 95.93 | 0.063 |
| Trachurus spp. | 9 | 10.56 ± 7.17 | 9 | 8.99 ± 8.51 | 0.596 |
| Scomber spp. | 9 | 9.69 ± 13.25 | 9 | 31.71 ± 25.71 | 0.017 |
| Total | 9 | 191.60 ± 164.92 | 9 | 109.29 ± 97.98 | 0.387 |
Results of the Wilcoxon rank-sum (Mann–Whitney U) test comparing CPUE expressed as abundance (n/h) between the two gear configurations for each species and for the total catch.
For each gear configuration, sample size (n) and mean ± standard deviation (SD) are also reported. P-values< 0.05 were considered statistically significant.
When CPUE was analysed in terms of biomass (kg/h; Table 2), a statistically significant difference between the two gear configurations was found only for Scomber spp. (p = 0.034). For Mullus barbatus, the p-value was very close to the conventional significance threshold (p = 0.051), indicating a tendency toward a difference, although not formally significant at α = 0.05. No significant differences were detected for Squilla mantis (p = 0.084) or Trachurus spp. (p = 0.536). Importantly, the total CPUE in biomass did not differ significantly between the two configurations (p = 0.387).
A similar pattern emerged when CPUE was analysed in terms of abundance (n/h; Table 3). Again, Scomber spp. showed a significant difference between gear types (p = 0.017), whereas Mullus barbatus showed a p-value close to significance (p = 0.063). No significant differences were observed for Squilla mantis (p = 0.143), Trachurus spp. (p = 0.596), or for total catch in numbers (p = 0.387).
Overall, as shown in Tables 2, 3, the observed differences were species-specific and primarily driven by Scomber spp., whereas total catch rates remained statistically comparable between gear configurations under both biomass and abundance metrics. This is consistent with the descriptive patterns shown in Figure 9 (mean CPUE by species), where differences appear limited to a subset of taxa, and is further supported by the strong overlap in CPUE distributions illustrated in Figure 10 (boxplots). The substantial within-group variability and overlapping interquartile ranges shown in Figure 10 reinforce the absence of a significant difference in overall catch performance between the two gear configurations.
Figure 9
Figure 10
Catch comparison results
Figure 11, Table 4 summarise the catch comparison results for the four most abundant commercial species collected during the sea trials: red mullet (Mullus barbatus), mackerels (Scomber spp.), mantis shrimp (Squilla mantis) and horse mackerels (Trachurus spp.). Table 4 also reports the total number of measured individuals and the fit statistics of the combined model used in the catch comparison analysis.
Figure 11
Table 4
| Mullus barbatus | Scomber spp. | Squilla mantis | Trachurus spp. | |
|---|---|---|---|---|
| Nr Measured individuals | 709 | 345 | 255 | 234 |
| p-value | <0.001 | 0.94 | 0.02 | 0.76 |
| Deviance | 87.55 | 9.66 | 26.11 | 16.17 |
| DOF | 17 | 18 | 13 | 21 |
Total number of measured individuals for each species included in the catch comparison analysis, and fit statistics of the combined model used in the catch comparison between the experimental and the standard net.
DOF, degrees of freedom.
The fit statistics of the combined model used in the catch comparison analysis are reported in Table 4. We observed p-values < 0.05 for both M. barbatus and S. mantis, however, these results were likely influenced by overdispersion in the experimental rates, since the visual inspection of the modelled catch comparison curves against the observed data did not reveal systematic length-dependent deviations for these species.
The experimental net significantly increased the catch efficiency of mackerels (Scomber spp.) for all length classes above 20 cm. Also, the mean catch comparison and catch ratio curves revealed a higher catch efficiency, although not statistically significant, for the experimental net in the 18–20 cm length range. This pattern is consistent with the length-frequency distributions shown in Figure 11 (top right). Almost no individuals below the Minimum Conservation Reference Size (MCRS) of 18 cm were caught during the sea trials. For horse mackerels (Trachurus spp.), the results revealed no significant differences in catch efficiency between the two nets.
The experimental net seemed to perform better than the traditional net for larger individuals (>22 cm), however, the wide confidence intervals, likely resulting from the low number of individuals caught in each size class, prevented this trend from reaching statistical significance (Figure 11, middle). No horse mackerels below the MCRS of 15 cm were caught in the sea trials.
Regarding both red mullet (M. barbatus) and mantis shrimp (S. mantis), the experimental net significantly reduced the catch of smaller and undersized individuals compared to the traditional net (10–13.5 cm for M. barbatus; 15–20 mm for S. mantis). However, across most length classes, the experimental net showed slightly lower mean catch efficiency than the traditional configuration, although these differences were not statistically significant. This pattern is reflected in both the catch comparison curves and the length-frequency distributions (Figure 11, middle-bottom and bottom).
Discussion
This study provides a full-scale comparison between a conventional Mediterranean demersal trawl and an experimentally optimised configuration, with a focus on gear performance, energy demand, and operational feasibility under commercial fishing conditions. The results indicate that relatively targeted and technically feasible modifications to trawl design can substantially reduce hydrodynamic drag and fuel consumption while maintaining overall catch rates comparable to the conventional configuration.
This study contributes to the limited body of full-scale empirical evidence available under commercial Mediterranean conditions, where quantitative validation of gear innovations remains scarce.
The reduction in drag observed for the experimental trawl can be interpreted as the combined outcome of the design changes implemented in the net body and overall geometry. The redistribution of mesh sizes, the use of thinner high-tenacity netting in the upper panels, and the more streamlined belly shape are all likely to enhance water flow through and around the gear, thereby reducing resistance during towing. The simultaneous increase in horizontal spread and reduction in vertical opening suggests a different balance between spreading forces and net shaping, achieved with a lower overall energy demand.
These findings are consistent with previous experimental and modelling studies demonstrating that reductions in netting area, twine diameter, and projected frontal surface can significantly decrease trawl drag and associated fuel demand (–), and confirm that such design principles remain effective when scaled up to full-size gears operating under realistic commercial conditions.
We observed a reduction in total fuel consumption (i.e. for the vessel and gear) by modifying the gear; however, we also observed reductions in catch. The combined effect of these resulted in 1.65% less mean FUI and 16.79% more median FUI. When evaluating changes to fishing gear, it is important to consider the effects on both fuel consumption and catch. However, FUI is most informative when calculated using data collected over longer time periods that capture variability in weather conditions, fishing areas, catch composition and catch rates throughout the fishing season. Short-term measurements may not provide a representative picture. Another important consideration is how fuel consumption is measured. Statistics on fuel consumption of vessels typically are based on fuel consumption reported by fishers or data on fuel sales. This includes all operational stages: manoeuvring in harbour, steaming to and from fishing grounds, moving between grounds, shooting the gear, fishing, and hauling. In this study, only fuel consumption during the fishing stage was measured. Besides, the vessel conducted only 2–5 hauls per day, whereas commercial fishing trips usually last several days and involve around 9 hauls. This means that, in the trial, a relatively large share of fuel was spent on steaming compared to the number of hauls. As a result, including all operational stages would disproportionately penalize the gears used in the trial. For these reasons, we excluded fuel use from non-fishing operations and did not compare the results directly with other vessels in the fleet, as such comparisons would not be consistent or meaningful. Data collection on longer time periods, during various seasons and during all operational stages can enable better estimation of FUI in future studies and capture overall changes on fuel consumption and catch.
Focusing only on the gear, from an operational perspective, the reduction in fuel consumption during towing represents a meaningful gain in energy efficiency.
The observed decrease (≈15%) falls within the range reported in previous studies investigating gear optimisation and energy efficiency in trawl fisheries, where fuel savings between 10% and 30% have been documented depending on gear configuration and operating conditions (, 27).
Although the experimental gear required a slightly higher towing speed to maintain stable geometry, the associated increase in vessel-related fuel demand was more than offset by the reduction in gear-related resistance. This highlights the importance of separating vessel and gear contributions to fuel consumption when assessing the energetic performance of fishing gears (). The combined use of lift tests and full-scale towing measurements provides a transparent framework for attributing observed fuel savings to gear design rather than to vessel characteristics alone and may facilitate comparisons across vessels operating under similar conditions.
Although direct measurements of seabed contact and sediment disturbance were beyond the scope of this study, the lower vertical opening, reduced drag, and observed changes in species composition may be consistent with a modified interaction between the gear and the seabed, possibly reflecting a slightly higher off-bottom fishing geometry. However, the present study does not allow firm conclusions to be drawn regarding seabed contact or benthic impact.
Similar relationships between gear configuration, vertical opening, and benthic interaction have been reported in previous studies (, ). In addition, changes in seabed contact may have implications for sediment disturbance and associated carbon processes, which have recently gained attention in the context of climate impacts of bottom trawling (, ). Since these processes were not quantified here, the observed patterns may be compatible with reduced seabed disturbance but this interpretation should be considered as a hypothesis requiring dedicated investigation.
Catch results indicate that the experimental trawl maintained commercially relevant catch levels, while showing a moderate shift in species composition towards more pelagic and semi-pelagic taxa.
This pattern may be consistent with the observed changes in gear geometry; however, it cannot be used as direct evidence of reduced seabed interaction.
The lower retention observed for some demersal and benthic-associated species may be related to the modified geometry of the experimental trawl, particularly its reduced vertical opening and potentially slightly higher off-bottom configuration.
Comparable outcomes have been reported in studies investigating modified trawl designs, where reductions in bottom contact or drag do not necessarily result in losses of target species catch, although effects on species composition may occur depending on gear configuration and habitat ().
Importantly, the experimental configuration showed a tendency to reduce the capture of smaller and undersized individuals for certain species, while maintaining or increasing the proportion of larger size classes. From a management perspective, this pattern is ecologically relevant, as reduced retention of juveniles may contribute to lowering fishing mortality on early life stages and mitigating growth overfishing (33, 34). At the same time, the significantly increased retention probability for larger individuals of species as Trachurus spp. and Scomber spp. may enhance the economic value of the catch, given the higher market price generally associated with larger size classes. This dual ecological and economic effect strengthens the case for gear optimization strategies that improve selectivity without reducing overall catch performance.
The strength of the present study lies in its implementation under full commercial conditions, which ensures high operational relevance. Some limitations should nevertheless be acknowledged. The experimental design was unpaired at the haul level, reflecting practical constraints of commercial fishing operations, seabed interaction was inferred indirectly rather than measured explicitly and discard fractions were not quantified separately. Consequently, although changes in catch composition and size selectivity were assessed, the potential effects of the experimental gear on discard generation could not be evaluated. Future studies should explicitly quantify retained and discarded fractions to determine whether the observed modifications in gear performance and selectivity translate into reductions in unwanted catches. These aspects define the scope of inference rather than undermining the main findings. The analytical framework adopted, combining high-resolution monitoring of fuel consumption, towing loads, and net geometry with vessel-specific fuel–speed modelling, provides a robust basis for evaluating gear performance and may be transferable to other fisheries and vessel types.
Overall, this work demonstrates that modest, well-targeted modifications to demersal trawl design can deliver measurable energy savings and potential reductions in seabed interaction while maintaining statistically comparable catch rates. By improving size selectivity for certain species (35, 36) and reducing the capture of undersized individuals, the experimental configuration may contribute simultaneously to environmental sustainability and economic performance. By providing a transparent, full-scale comparative framework, the study contributes operational evidence supporting the development and adoption of low-impact trawl gears in Mediterranean demersal fisheries, with implications for fuel efficiency, carbon mitigation, and ecosystem-based fisheries management.
Statements
Data availability statement
The data supporting the conclusions of this article (i.e. fuel consumption records, towing loads, net geometry time series and catch composition data) are available from the corresponding author upon reasonable request, subject to project and institutional data sharing policies.
Ethics statement
Ethical approval was not required under applicable italian legislation and istitutional requirements because the trials were conducted under commercial fishing conditions and were authorized by the Italian Ministry of Agricolture Food Sovereignity and Forests and the Italian Coast Guard. No procedures beyond standard commercial fhishing operations were performed.
Author contributions
IM: Writing – original draft, Writing – review & editing. DL: Writing – review & editing. AP: Writing – review & editing. GV: Writing – review & editing. BH: Methodology, Validation, Writing – review & editing. RC: Writing – review & editing. AL: Writing – review & editing. AS: Writing – review & editing. SJ: Writing – review & editing. NH: Writing – review & editing. ES: Writing – review & editing. FO: Writing – review & editing. EN: Writing – review & editing, Writing – original draft.
Funding
The author(s) declared that financial support was received for this work and/or its publication. This work was supported by the RightFish project, ‘Reducing environmental impact and greenhouse gas emissions in commercial fisheries’, funded by the ERA-NET BlueBio programme under Grant Agreement No. 817992; CUP B53C22005410006.
Acknowledgments
This work was carried out within the RightFish project, focused on the development and validation of low-impact towed fishing gears under commercial operating conditions. The authors acknowledge all RightFish consortium partners (DTU Aqua, CNR-IRBIM, SINTEF Ocean, and Hampiðjan) for their collaboration. We are particularly grateful to the vessel owners, Attilio and Pasquale Fava, and to the crew of the Albatros Selvaggio for their support and cooperation during the sea trials.
Conflict of interest
The author(s) declared that this work was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.
The author AS declared that they were an editorial board member of Frontiers at the time of submission. This had no impact on the peer review process and the final decision.
Generative AI statement
The author(s) declared that generative AI was used in the creation of this manuscript. The authors used ChatGPT (OpenAI, GPT-5.3) for language refinement and rephrasing of selected sections. All content was reviewed and validated by the authors.
Any alternative text (alt text) provided alongside figures in this article has been generated by Frontiers with the support of artificial intelligence and reasonable efforts have been made to ensure accuracy, including review by the authors wherever possible. If you identify any issues, please contact us.
Publisher’s note
All claims expressed in this article are solely those of the authors and do not necessarily represent those of their affiliated organizations, or those of the publisher, the editors and the reviewers. Any product that may be evaluated in this article, or claim that may be made by its manufacturer, is not guaranteed or endorsed by the publisher.
Supplementary material
The Supplementary Material for this article can be found online at: https://www.frontiersin.org/articles/10.3389/frish.2026.1893921/full#supplementary-material
References
1
MuirJF. Fuel and Energy Use in the Fisheries Sector: Approaches, Inventories and Strategic Implications. (2015).
2
FAO. The State of World Fisheries and Aquaculture 2020: Sustainability in Action. Rome: Food and Agriculture Organization of the United Nations (2020). doi: 10.4060/ca9229en
3
JhaPNBaijuMVEdwinL. A Review of Energy Use and Carbon Footprint in Fishing With Special Reference to Life Cycle Assessment (LCA). Fishery Technology. (2025) 62(1):1–15. doi: 10.56093/ft.v62i1.151686
4
ParkerRWRBlanchardJLGardnerCGreenBSHartmannKTyedmersPHet al. Fuel use and greenhouse gas emissions of world fisheries. Nat. Clim Change. (2018) 8:333–7. doi: 10.1038/s41558-018-0117-x
5
CheilariAGuillenJDamalasDBarbasT. Effects of the fuel price crisis on the energy efficiency and the economic performance of the European Union fishing fleets. Mar. Policy. (2013) 40:18–24. doi: 10.1016/j.marpol.2012.12.006
6
SuuronenPChopinFGlassCLøkkeborgSMatsushitaYQueiroloDet al. Low impact and fuel efficient fishing-Looking beyond the horizon. Fish Res. (2012) 119–120:135–46. doi: 10.1016/j.fishres.2011.12.009
7
ThomsenBZachariassenKEliasenMHansenPHeinesenS. (2010). “ Fuel optimisation in the Faroese fishing fleet”, in: Proceedings of the First International Symposium on Fishing Vessel Energy Efficiency (E-Fishing). Vigo, Spain.
8
AbernethyKETrebilcockPKebedeBAllisonEHAbernethyNKDTrebilcockKEet al. Fuelling the decline in UK fishing communities? ICES J. Mar. Sci. (2010) 67(5):1076–85. doi: 10.1093/icesjms/fsp289
9
GuillenJMaynouF. Increasing fuel prices, decreasing fish prices and low productivity lead to poor economic performance and capacity reduction in the fishing sector: Evidence from the Spanish mediterranean. Turk J. Fish Aquat Sci. (2016) 16:659–68. doi: 10.4194/1303-2712-v16_3_20
10
ThierryNNBTangHLiuxiongXYouXHuFAchileNPet al. Hydrodynamic performance of bottom trawls with different materials, mesh sizes, and twine thicknesses. Fish Res. (2020) 221:105403. doi: 10.1016/j.fishres.2019.105403
11
SterlingDEayrsS. (2010). “ Trawl-gear innovations to improve the energy efficiency of Australian prawn trawling”, in: Proceedings of the First International Symposium on Fishing Vessel Energy Efficiency (E-Fishing). Vigo, Spain.
12
TangHHuFXuLDongSZhouCWangX. Variations in hydrodynamic characteristics of netting panels with various twine materials, knot types, and weave patterns at small attack angles. Sci. Rep. (2019) 9:1923. doi: 10.1038/s41598-018-35907-1
13
HiddinkJGJenningsSSciberrasMSzostekCLHughesKMEllisNet al. Global analysis of depletion and recovery of seabed biota after bottom trawling disturbance. PNAS. (2017) 114:8301–6. doi: 10.1073/pnas.1618858114
14
SciberrasMHiddinkJGJenningsSSzostekCLHughesKMKneafseyBet al. Response of benthic fauna to experimental bottom fishing: A global meta-analysis. Fish Fish. (2018) 19:698–715. doi: 10.1111/faf.12283
15
EpsteinGFullerSDGullageLBaumJK. Mobile bottom fishing in the Canadian Pacific and Atlantic causes disturbance and risk to remineralisation of seabed sediment carbon stocks. Can J Fish Aquat Sci. (2025) 82:1–22. doi: 10.1139/cjfas-2024-0325
16
SalaEMayorgaJBradleyDCabralRBAtwoodTBAuberAet al. Protecting the global ocean for biodiversity, food and climate. Nature. (2021) 592:397–402. doi: 10.1038/s41586-021-03371-z
17
ParkerRWRTyedmersPH. Fuel consumption of global fishing fleets: Current understanding and knowledge gaps. Fish Fish. (2015) 16:684–96. doi: 10.1111/faf.12087
18
TyedmersP. Fisheries and energy use. In: ClevelandCJ, editor. Encyclopedia of Energy. Vol. 2. San Diego, CA: Elsevier Academic Press (2004). p. 683–93. doi: 10.1016/B0-12-176480-X/00204-7
19
HuFMatudaKTokaiT. Effects of drag coefficient of netting for dynamic similarity on model testing of trawl nets. Fish Sci. (2001) 67(1):84–9. doi: 10.1046/j.1444-2906.2001.00203.x
20
MontgomerieM. Basic Fishing Methods: A Comprehensive Guide to Commercial Fishing Methods. Edinburgh, UK: Seafish (2022). 110 p. Publication Reference No. 35572.
21
GabiñaGBasurkoOCNottiESalaAAldekoaSClementeMet al. Energy efficiency in fishing: Are magnetic devices useful for use in fishing vessels? Appl. Therm Eng. (2016) 94:670–8. doi: 10.1016/j.applthermaleng.2015.10.161
22
JafarzadehSEllingsenHAanondsenSA. Energy efficiency of Norwegian fisheries from 2003 to 2012. J. Cleaner Prod. (2016) 112:3616–30. doi: 10.1016/j.jclepro.2015.06.114
23
ZieglerFJafarzadehSSkontorp HognesEWintherU. Greenhouse gas emissions of Norwegian seafoods: From comprehensive to simplified assessment. J. Ind. Ecol. (2022) 26:1908–19. doi: 10.1111/jiec.13150
24
R Core Team. R: A Language and Environment for Statistical Computing. Vienna: R Foundation for Statistical Computing (2025). Available online at: https://www.R-project.org/.
25
WickhamH. Ggplot2 Elegant Graphics for Data Analysis Second Edition (2016). doi: 10.1007/978-3-319-24277-4
26
SalaADe CarloFBuglioniGLucchettiA. Energy performance evaluation of fishing vessels by fuel mass flow measuring system. Ocean Eng. (2011) 38:804–9. doi: 10.1016/j.oceaneng.2011.02.004
27
SalaADamalasDLabanchiLMartinsonJMoroFSabatellaRet al. Energy audit and carbon footprint in trawl fisheries. Sci Data. (2022) 9:428. doi: 10.1038/s41597-022-01478-0
28
HerrmannBSistiagaMNielsenKNLarsenRB. Understanding the size selectivity of redfish (Sebastes spp.) in North Atlantic trawl codends. J. Northwest Atl Fish Sci. (2012) 44:1–13. doi: 10.2960/J.v44.m680
29
HerrmannBSistiagaMRindahlLTatoneI. Estimation of the effect of gear design changes on catch efficiency: Methodology and a case study for a Spanish longline fishery targeting hake (Merluccius merluccius). Fish Res. (2017) 185:153–60. doi: 10.1016/j.fishres.2016.09.013
30
BurnhamKPAndersonDR. Model Selection and Multimodel Inference: A Practical Information-Theoretic Approach. 2nd ed. New York, NY: Springer New York (2002). doi: 10.1007/b97636
31
WilemanDAFerroRSTFonteyneRMillarRB. Manual of Methods of Measuring the Selectivity of Towed Fishing Gears. ICES Cooperative Research Report No. 215. Copenhagen: International Council for the Exploration of the Sea (1996). 126 p.
32
LomeliMJM. Bycatch Reduction in Eastern North Pacific Trawl Fisheries. (PhD thesis). Tromsø: UiT The Arctic University of Norway (2019). 119 p.
33
BellidoJMSantosMBPenninoMGValeirasXPierceGJ. Fishery discards and bycatch: Solutions for an ecosystem approach to fisheries management? Hydrobiologia. (2011) 670:317–33. doi: 10.1007/s10750-011-0721-5
34
PetettaAHerrmannBLi VeliDCanduciGTatoneIBonanomiSet al. Vitality insights of fish escaping from a sorting grid installed on a bottom trawl net. Sci. Rep. (2025) 15:552. doi: 10.1038/s41598-024-84364-6
35
BrčićJHerrmannBSalaA. Can a square-mesh panel inserted in front of the cod end improve size and species selectivity in mediterranean trawl fisheries? Can. J. Fish AquatSci. (2018) 75:704–13. doi: 10.1139/cjfas-2017-0123
36
SalaAPriourDHerrmannB. Experimental and theoretical study of red mullet (Mullus barbatus) selectivity in codends of Mediterranean bottom trawls. Aquat Living Resour. (2006) 19:317–27. doi: 10.1051/alr:2007002
Summary
Keywords
catch comparison, demersal trawl, drag reduction, fuel consumption, low-impact gear, Mediterranean Sea
Citation
Martino I, Li Veli D, Petetta A, Vianson G, Herrmann B, Cacciamani R, Lucchetti A, Sala A, Jafarzadeh S, Huda N, Skaftason E, O’Neill FG and Notti E (2026) Energy savings in Mediterranean demersal trawl fisheries: full and model scale comparison of experimental and conventional gears on drag and fuel consumption. Front. Fish Sci. 4:1893921. doi: 10.3389/frish.2026.1893921
Received
28 May 2026
Revised
26 June 2026
Accepted
04 August 2026
Published
16 September 2026
Volume
4 - 2026
Edited by
Emanuela Fanelli, Marche Polytechnic University, Italy
Updates
Copyright
© 2026 Martino, Li Veli, Petetta, Vianson, Herrmann, Cacciamani, Lucchetti, Sala, Jafarzadeh, Huda, Skaftason, O’Neill and Notti.
This is an open-access article distributed under the terms of the Creative Commons Attribution License (CC BY). The use, distribution or reproduction in other forums is permitted, provided the original author(s) and the copyright owner(s) are credited and that the original publication in this journal is cited, in accordance with accepted academic practice. No use, distribution or reproduction is permitted which does not comply with these terms.
*Correspondence: Emilio Notti, emilio.notti@cnr.it
Disclaimer
All claims expressed in this article are solely those of the authors and do not necessarily represent those of their affiliated organizations, or those of the publisher, the editors and the reviewers. Any product that may be evaluated in this article or claim that may be made by its manufacturer is not guaranteed or endorsed by the publisher.