A Catch Community Diversity Analysis of Purse Seine in the Tropical Western and Central Pacific Ocean

Abstract

Epipelagic fish communities dominate fish assemblages and are an important part of marine ecosystems due to their high abundance, vertical migration behavior, and global distribution. Purse seine fisheries are key components of marine fisheries in the tropical Western and Central Pacific Ocean (WCPO), primarily targeting skipjack tuna (Katsuwonus pelamis, SKJ), yellowfin tuna (Thunnus albacares, YFT), and bigeye tuna (Thunnus obesus, BET). In this study, WCPO purse seine fishery data from 2014 to 2022, combined with environmental factor data, were used, and Mantel tests and correlation analysis were employed to analyze the diversity, fish coexistence mechanisms, and environmental responses of catch communities under the following two different fishing strategies: free–swimming schools (FSCs) and drifting fish aggregating devices (DFADs). Mantel tests indicated that nitrate (NO_3⁻), the Oceanic Niño Index (ONI), and pH significantly impact the diversity of the FSCs community, whereas NO_3⁻ significantly affects the diversity of the DFADs community. Based on the correlation analysis results, in the FSCs community, yellowfin tuna was positively correlated with bigeye tuna, and yellowfin tuna was negatively correlated with skipjack tuna and black marlin (Istiompax indica, BLM). In the DFADs community, yellowfin tuna was only positively correlated with skipjack tuna and bigeye tuna. In addition, species with high correlations were also positively correlated. The results of this study provide a theoretical basis for the biodiversity conservation of catch communities under two different purse seine fishing strategies in the WCPO.

1. Introduction

Meso–pelagic fishes dominate the global total fish biomass and play a crucial role in marine ecosystems due to their impacts on carbon sequestration and trophic connectivity [1,2]. The largest tuna fishery in the world is found in the Western and Central Pacific Ocean (WCPO), where tuna fishery is highly diverse [3]. The purse seine fishery, as one of its key components, primarily targets species such as skipjack (Katsuwonus pelamis, SKJ), yellowfin (Thunnus albacares, YFT), and bigeye tuna (Thunnus obesus, BET) [4]. These large pelagic fishes are top predators in the marine food web and play a critical role in the structure of pelagic communities [5].

The tuna purse seine industry increasingly deploys free–drifting man–made floating objects, so–called drifting fish aggregating devices (DFADs) [6], to take advantage of the tendency of pelagic fish to associate with natural objects (such as logs or debris) or artificial structures (such as buoys or rafts) and thereby enhance catchability as fish schools are more stable under such conditions [7]. Such devices, which can concentrate randomly distributed pelagic fish schools, are important tools for assessing pelagic fish diversity [8]. Approximately 40% of the purse seine catch in the WCPO is taken in association with DFADs [3]; other catches are derived from free–swimming schools (FSCs) and dolphin–associated aggregations [9]. Before the extensive use of DFADs over the past 30 years, FSCs tuna schools were the main targets of purse seines [10]. Various purse seine fishing strategies lead to varying species and group compositions in catch. Understanding the composition of catch communities under different purse seine strategies in the WCPO is crucial for conserving fishery resources, achieving ecosystem–based fishery management, and maintaining sustainable fisheries.

The co–occurrence and –existence of species can be reflected by changes in resource diversity, environmental or habitat heterogeneity, interspecific interactions, biological invasions, and external disturbances [11]. Any change in these characteristics can alter biodiversity by changing the relative species abundance [12]. However, the mechanisms of species coexistence in the WCPO are understudied.

Climate change is rapidly altering the structure and functioning of marine ecosystems [13]. Changes in environmental parameters due to climate change, such as temperature, dissolved oxygen, and carbon dioxide levels in the ocean, may result in variations in fish mortality, growth, reproduction, and distribution [14], consequently impacting biodiversity. Against the background of a changing climate, the degree of the impact of environmental parameters on community diversity may vary among communities. However, the ecological processes of organisms within WCPO purse seine catch communities and comparisons among different communities remain largely unknown. In this context, focusing on the processes and mechanisms of structural changes in WCPO purse seine catch communities under different fishing strategies and the impacts of environmental parameters on the diversity of different communities is the key to achieving the protection of pelagic marine ecosystems.

In this study, data from the WCPO purse seine fishery, along with key environmental factors such as surface sea temperature (SST), surface sea salinity (SSS), and dissolved oxygen (DO), were used to investigate the variations in catch community diversity and their responses to environmental factors. The objectives are as follows: (1) to explore the diversity of FSCs and DFADs communities; (2) to assess the impacts of environmental factors on the diversity of FSCs and DFADs communities; and (3) to reveal the mechanisms underlying species coexistence and symbiosis within FSCs and DFADs communities.

2. Materials and Methods

2.1. Data Sources

Daily fishery data of seven purse seine vessels for 2014–2022 were provided by Shanghai Kaichuang Marine International Co., Ltd., Shanghai, China. This dataset includes information regarding fishing date, fishing location (longitude and latitude), fishing vessel name, catch, and school association codes. The operational range of these vessels was from 155° E to 180° E, 180° W to 150° W, and 10° S to 10° N (Figure 1). For community data analysis, the fishing sets of data on the FSC and DFAD community were filtered based on the school association code recorded in the logbook.

Marine environmental data, including surface sea temperature (SST), surface sea salinity (SSS), chlorophyll–a (Chl–a), nitrate (NO_3⁻), dissolved oxygen (DO), and pH, were sourced from the Copernicus Marine Environment Monitoring Service website [15]. These data cover a spatial range of 150° E to 180° E, 180° W to 150° W, 10° S to 10° N. with a spatial resolution of 0.5° × 0.5° and a temporal resolution of 1 month. Ocean Niño Index (ONI) data were sourced from the Climate Prediction Center of the National Oceanic and Atmospheric Administration, providing monthly average sea surface temperature anomalies (SSTAs) for the Niño3.4 region (120°–170° W and 5° N–5° S) for corresponding months. The data can be accessed at https://origin.cpc.ncep.noaa.gov/products/analysis_monitoring/ensostuff/ONI_v5.php (accessed on 15 December 2023).

2.2. Data Processing

2.2.1. Community Composition and Dominant Species

The index of relative importance (IRI) [16] was employed to determine the dominant species (IRI ≥ 1000), important species (1000 > IRI ≥ 100), common species (100 > IRI ≥ 10), general species (10 > IRI ≥ 1), and rare species (IRI < 1) in the fish community. The IRI is expressed as follows:

$$IRI=\left(N_i+W_i\right)\times F_i\times {10}^4$$

(1)

where $N_i$ is the proportion of the number of individuals of species i to the total number of individuals, $W_i$ is the proportion of the weight of species i to the total weight, and $F_i$ is the frequency percentage of the occurrence of species i, reflecting its spatial distribution, rather than predator feeding preferences.

2.2.2. Analysis of Community Diversity

Community diversity was assessed using the Shannon–Wiener diversity index (H′) [17], the Margalef richness index (D) [18], and Pielou’s evenness index (J) [19]. The Shannon–Wiener diversity index (H′) is an important indicator of community structure stability, quantifying the uncertainty in predicting species identity through the combined effects of richness and evenness. Higher values (H′ > 3) indicate complex communities with balanced species distributions, while lower values (H′ < 1) suggest dominance by a few taxa. The richer the community in terms of species, the higher the H′ value. When the ecological environment of a water area is subjected to stressors, both species and individual numbers decrease, leading to a reduction in H′. The Margalef richness index (D) reflects the abundance of fish resources; the higher the D value, the richer the fish resources. The Pielou’s evenness index (J) is commonly used to measure the uniformity of fish distribution, ranging from 0 (complete dominance) to 1 (perfect evenness). In fisheries, low J–values (<0.5) may indicate the selective overexploitation of target species [20]. The formulas are as follows:

$$H^{\prime }=-\sum \left[\left(n_i∕Na\right)\ast \ln \hspace{0.33em}\left(n_i/Na\right)\right]$$

(2)

$$D=\left(S-1\right)/\ln \hspace{0.33em}{N}_a$$

(3)

$$J=H^{\prime }/\ln \hspace{0.33em}S$$

(4)

where $n_i$ is the number of individuals of species; $Na$ is the total number of individuals in the catch; and $S$ is the number of species in the catch.

The Mantel test [21] was performed to assess the influence of environmental distance on the Shannon–Wiener diversity index, Marglef richness index, and Pielou’s evenness index. Individual and environmental factors were transformed into distance matrices as follows: First, for each environmental variable, pairwise dissimilarities between sampling sites were calculated using Euclidean distance, resulting in single–factor environmental distance matrices. To mitigate bias caused by scale discrepancies among variables, all environmental data were standardized before integration. A comprehensive environmental distance matrix was then constructed from the standardized dataset. Subsequently, the correlations between these matrices and the diversity matrix were analyzed. This analysis was performed in R using the vegan package, and the “linkET” package was used to visualize the Mantel results. All analyses were conducted using monthly aggregated data.

2.2.3. Correlation Analysis of Fish Species

The mechanisms of symbiosis and coexistence in catch communities were mainly reflected through the correlations between fish species. Spearman correlation analysis [22] was performed to test the relationships between fish species, a non–parametric statistical method that assesses the strength and direction of monotonic relationships between paired variables without necessitating assumptions of normality or linearity. To reduce the impacts of extreme values of occasional and dominant species, fish species with an occurrence frequency greater than 1.0% were selected for data processing.

3. Results

3.1. Differences in the Composition of Catch Species

From 2014 to 2022, a total of 66 species of catch were collected by purse seine vessels in the WCPO, belonging to 1 phylum, 4 classes, 23 orders, 37 families, and 58 genera. The FSCs catch included 15 species from 9 families, and the DFADs catch included 65 species from 37 families. Compared to the FSCs community, the DFADs community had a higher species richness. A comparison of the IRI values of different fish species in the two communities (

Table 1. Species composition and IRI values of purse seine catch communities in the Western and Central Pacific. IRI = index of relative importance, FSCs = free–swimming schools, DFADs = drifting fish aggregating devices.

Marine Species Classifications	Family	Scientific Name	English Name	Code	FSCs	DFADs
Billfish and Barracudas	Istiophoridae	Istiompax indica	Black marlin	BLM	2.04 × 10−2	3.46 × 10−2
		Kajikia audax	Striped marlin	MLS		6.72 × 10−6
		Makaira nigricans	Blue marlin	BUM	4.46 × 10−4	1.05 × 10−2
		Tetrapturus angustirostris	Shortbill spearfish	SSP		2.46 × 10−6
	Xiphiidae	Xiphias gladius	Swordfish	SWO		5.04 × 10−7
Bottom-Dwelling Fish	Arcidae	Anadara spp.	Anadara clams nei	BLS		2.02 × 10−7
	Bothidae	Monolene antillarum	Slim flounder	BML		3.07 × 10−6
	Gobiidae	Cryptocentrus filifer	/	YTF		1.81 × 10−6
	Macrouridae	Coryphaenoides nasutus	Largenose grenadier	MHN		3.17 × 10−7
	Percophidae	Percophis brasiliensis	Brazilian flathead	FLA		2.34 × 10−6
	Soleidae	Dicologlossa cuneata	Wedge sole	CET		2.96 × 10−7
Dolphins and Whales	Delphinidae	Globicephala macrorhynchus	Short-finned pilot whale	SHW		5.76 × 10−7
		Globicephala macrorhynchus	Short-finned pilot whale	SHW		5.76 × 10−7
		Grampus griseus	Risso’s dolphin	DRR		3.84 × 10−7
		Grampus griseus	Risso’s dolphin	DRR		3.84 × 10−7
		Pseudorca crassidens	False killer whale	FAW		5.40 × 10−4
		Pseudorca crassidens	False killer whale	FAW		5.40 × 10−4
		Stenella longirostris	Spinner dolphin	DSI		2.30 × 10−6
		Stenella longirostris	Spinner dolphin	DSI		2.30 × 10−6
		Steno bredanensis	Rough-toothed dolphin	RTD	2.05 × 10−4	8.59 × 10−5
		Steno bredanensis	Rough-toothed dolphin	RTD	2.05 × 10−4	8.59 × 10−5
		Tursiops truncatus	Bottlenose dolphin	DBO		1.47 × 10−5
		Tursiops truncatus	Bottlenose dolphin	DBO		1.47 × 10−5
Others	Loricariidae	Otocinclus affinis	Golden otocinclus	OCA		1.34 × 10−6
	Megalopidae	Megalops cyprinoides	Indo-Pacific tarpon	TAI		2.54 × 10−7
	Molidae	Masturus lanceolatus	Sharptail mola	MRW		2.02 × 10−7
		Mola mola	Ocean sunfish	MOX		9.75 × 10−6
		Ranzania laevis	Slender sunfish	RZV		2.13 × 10−7
	Stomiidae	Eustomias lipochirus	/	FAI		4.52 × 10−7
Rays and Mantas	Dasyatidae	Pteroplatytrygon violacea	Pelagic stingray	PLS		4.23 × 10−5
	Mobulidae	Mobula birostris	Giant manta	RMB	5.09 × 10−3	1.02 × 10−2
		Mobula spp.	Mobula nei	RMV	3.94 × 10−5	5.48 × 10−4
		Mobulidae	Mantas, devil rays nei	MAN	4.46 × 10−5	4.27 × 10−5
	Urolophidae	Urolophus aurantiacus	Sepia stingray	RUU		9.60 × 10−7
Reef and Coastal Fish	Balistidae	Balistidae	Triggerfishes, durgons nei	TRI		1.21 × 10−2
		Canthidermis maculata	Rough triggerfish	CNT		1.11 × 10−1
	Carangidae	Caranx sexfasciatus	Bigeye trevally	CXS		1.28 × 10−5
		Decapterus macarellus	Mackerel scad	MSD		8.58 × 10−1
		Elagatis bipinnulata	Rainbow runner	RRU		2.09
		Gnathanodon speciosus	Golden trevally	GLT		7.84 × 10−7
		Selar crumenophthalmus	Bigeye scad	BIS		1.62 × 10−5
		Seriola lalandi	Yellowtail amberjack	YTC		4.44 × 10−6
		Uraspis secunda	Cottonmouth jack	USE		2.02 × 10−7
	Coryphaenidae	Coryphaena hippurus	Common dolphinfish	DOL	6.99 × 10−4	2.03 × 10−2
	Ephippidae	Platax teira	Longfin batfish	BAO		1.82 × 10−6
	Kyphosidae	Kyphosus cinerascens	Blue sea chub	KYC		1.85 × 10−4
	Monacanthidae	Aluterus monoceros	Unicorn leatherjacket filefish	ALM		1.85 × 10−6
Sea Turtles	Cheloniidae	Caretta caretta	Loggerhead turtle	TTL		9.24 × 10−6
		Chelonia mydas	Green turtle	TUG	2.68 × 10−4	9.76 × 10−7
		Lepidochelys olivacea	Olive ridley turtle	LKV		1.35 × 10−5
		Natator depressus	Flatback turtle	FBT		1.73 × 10−6
Sharks	Alopiidae	Alopias pelagicus	Pelagic thresher	PTH		1.92 × 10−7
		Alopias superciliosus	Bigeye thresher	BTH		3.86 × 10−6
	Carcharhinidae	Carcharhinus falciformis	Silky shark	FAL	4.57 × 10−1	2.89
		Carcharhinus longimanus	Oceanic whitetip shark	OCS	3.57 × 10−4	5.90 × 10−3
	Galeocerdonidae	Galeocerdo cuvier	Tiger shark	TIG		1.92 × 10−7
	Lamnidae	Isurus oxyrinchus	Shortfin mako	SMA		4.54 × 10−6
	Rhincodontidae	Rhincodon typus	Whale shark	RHN	1.79 × 10−2	1.57 × 10−4
	Sphyrnidae	Sphyrna lewini	Scalloped hammerhead	SPL	8.92 × 10−5
		Sphyrna mokarran	Great hammerhead	SPK		2.22 × 10−6
Tunas	Scombridae	Acanthocybium solandri	Wahoo	WAH		7.27 × 10−3
		Auxis thazard	Frigate tuna	FRI		2.13 × 10−7
		Euthynnus affinis	Kawakawa	KAW		1.18 × 10−6
		Katsuwonus pelamis	Skipjack tuna	SKJ	8.02 × 103	1.85 × 104
		Thunnus albacares	Yellowfin tuna	YFT	1.92 × 102	3.01 × 102
		Thunnus obesus	Bigeye tuna	BET	4.97 × 10−2	3.6

) revealed that the target species skipjack tuna (Katsuwonus pelamis, SKJ) was the absolute dominant species in both FSC and DFAD communities (IRI >> 1000), with yellowfin tuna (Thunnus albacares, YFT) being an important species. In the FSCs community, the rare species included the silky shark (Carcharhinus falciformis, FAL), giant manta (Mobula birostris, RMB), whale shark (Rhincodon typus, RHN), bigeye tuna (Thunnus obesus, BET), black marlin (Istiompax indica, BLM), and blue marlin (Makaira nigricans, BUM), among 13 other species. The general species in the DFADs community included rainbow runner (Elagatis bipinnulata, RRU), silky shark (Carcharhinus falciformis, FAL), and bigeye tuna (Thunnus obesus, BET), whereas rare species were common dolphinfish (Coryphaena hippurus, DOL), mackerel scad (Decapterus macarellus, MSD), oceanic whitetip shark (Carcharhinus longimanus, OCS), giant manta (Mobula birostris, RMB), great barracuda (Sphyraena barracuda, GBA), black marlin (Istiompax indica, BLM), blue marlin (Makaira nigricans, BUM), triggerfishes (Balistidae, TRI), rough triggerfish (Canthidermis maculate, CNT), and various species of the Balistidae family, among 60 other species [23].

3.2. Community Diversity Analysis

The diversity indices of the catch communities in the WCPO purse seine fishery varied among different communities (Figure 2). The Margalef richness index of the FSCs community was considerably lower than that of the DFAD community from January to December (Figure 2b), whereas the Shannon–Wiener diversity index of FSCs was higher than that of the DFADs community just in November (Figure 2a). For the FSCs community, the corresponding annual averages of the Shannon–Wiener diversity index, Margalef richness index, and Pielou’s evenness index were 0.120, 0.281, and 0.112, respectively. The annual averages of the Shannon–Wiener diversity index, Margalef richness index, and Pielou’s evenness index for the DFADs community were 0.074, 0.812, and 0.036, respectively. In contrast, the DFADs community exhibited significantly higher Margalef richness but lower Shannon diversity and evenness, suggesting that the DFADs community had higher species richness on an annual basis. However, the monthly analysis conducted in November and December revealed an increase in the Shannon–Wiener diversity index (H′) and Pielou’s evenness index for the FSCs community, with distinct peaks observed in November.

Figure 3 illustrates the monthly average variation trends of environmental factors in the WCPO. Chl–a concentrations peaked during the early months (January to March) and declined to their lowest values from October to December. DO levels exhibited a sharp increase in December. NO_3⁻ concentrations displayed a bimodal pattern, with a primary peak in May and a secondary peak in October. A significant decline of 99.9% occurred from October to November. Intermediate fluctuations were observed between these extremes, with a gradual increase from July to August and a gradual decrease from August to September. The pH remained stable at 8.02 to 8.04. A notable spike in SSS occurred in November, likely due to mesoscale hydrographic events. SSTs were generally stable at 28 to 29 °C, with slight cooling towards the end of the year.

Figure 4 shows the correlations between the six environmental factors, the ONI, and the diversity of catch communities (FSC and DFADs communities) in the WCPO. The environmental factors significantly affecting the diversity of the FSC community were NO_3⁻, pH, and ONI. Both NO_3⁻ (r = 0.164, p = 0.029) and pH (r = 0.224, p = 0.006) were significantly positively correlated with the Margalef richness index (D), and the ONI (r = 0.377, p = 0.001) was positively correlated with the Shannon–Wiener diversity index (H′), Margalef richness index (D), and Pielou’s evenness index (J) (Figure 4a). The environmental factor significantly affecting the diversity of the DFADs community was NO_3⁻ (r = 0.161, p = 0.001), which was positively correlated only with the Shannon–Wiener diversity index (H′) and Margalef richness index (D) (Figure 3b). The diversity of the FSC community was significantly positively correlated with environmental distance (r = 0.213, p = 0.015), whereas the diversity of the DFADs community was not correlated with environmental distance (r = 0.055, p = 0.160).

3.3. Correlations Between Community Species

The correlations between different mid–upper layer species in the communities based on Spearman’s correlation analysis (p < 0.05) are shown in Figure 4. In the FSC community, there was a significant negative correlation between YFT and the species SKJ and BLM and a significant positive correlation between YFT and the species BET and FAL. Additionally, there was a positive correlation between FAL and the species BUM and RMB (Figure 5a).

For the DFADs community, there were multiple species correlations (Figure 5b). A positive correlation was found between YFT and other species such as BET, SKJ, and FAL. Additionally, based on the figure, species with high correlation values were primarily RRU and CNT. There were positive correlations between RRU and CNT as well as between RRU and other species such as DOL, MSD, and TRI.

4. Discussion

The Western and Central Pacific Ocean (WCPO) is one of the most heavily fished regions in the world and provides habitat to a variety of creatures, including highly migratory species such as tuna and mackerel [24]. This region is primarily influenced by the North Equatorial Current system, with surface currents in the south being affected by prevailing monsoons, whose direction changes with the seasons [25]. The northern part is mostly influenced by the Kuroshio Current, which provides rich productivity and suitable growth conditions, attracting abundant fish resources [26]. Climatic changes, anomalies, and oscillations drive changes in oceanic environmental factors, thereby affecting life activities and processes in marine ecosystems [27]. The dominant mode of year–to–year (interannual) climate variability in the WCPO and globally is the El Niño–Southern Oscillation (ENSO) [28], characterized by interannual oscillations of sea surface temperature (SST) and atmospheric pressure. This oscillating ENSO cycle triggers alterations in the structure of primary and secondary producers, generating complex and unique fish communities in this region [29]. This study on the fish composition from purse seine fishing vessel surveys in the WCPO showed a total of 66 species caught, belonging to 4 classes, 23 orders, 37 families, and 58 genera, with a greater species richness compared to those observed in previous studies in the same survey area [30]. Among them, the FSCs community contained 15 species, the DFADs community contained 65 species, and 14 species were common to both. While the primary target species in both WCPO purse seine fishing methods were similar, the DFADs community exhibited greater species richness, likely attributable to the extensive deployment of DFADs devices. The deployment of floating objects in the WCPO is the largest in the world [31]. Compared to the FSCs, DFADs have several advantages: (1) in searching reduction of searching times; (2) a reduction in the number of null sets (i.e., sets where the vessel is not successful in encircling the fish school); and (3) an increase in success rates and the achievement of more stable catches [32,33]. Considering purse seine fishing at tuna aggregations associated with DFADs, the catch rate of non–tuna species is higher, and the species are more diverse [34]. This indicates that DFADs not only excel in increasing catch rates but also significantly promote species diversity.

As shown in Figure 2a,c, the Margalef species richness index and Pielou’s evenness index of the FSCs community were higher in November and December, which may be related to coastal upwelling near the equatorial region. These warm, high–yield, and low–salinity waters are located within the warm pool, where the thermocline depth is lower. In winter, the influence of wind–driven jets causes colder and nutrient–rich water to upwell to the surface, giving rise to cool and highly productive surface waters [33].

In this study, comprehensive environmental distance (all environmental variables) had a considerable impact on the FSCs community but not on the DFADs community. This could be explained with the associative behavior of tropical tuna species with floating objects, forming large multi–species aggregations around them. Fisheries exploit this associative behavior by deploying drifting artificial fish aggregating devices, which may “trap” mid–upper layer species [6]. Compared with floating–object–associated schools, free–swimming tuna schools are more mobile and less predictable regarding their vertical and horizontal positions, which are generally influenced by environmental conditions [10]. This pattern aligns with the ecological drivers of the two communities: the FSCs community exhibited significant associations with NO_3⁻, pH, and the ONI, indicating their environmental sensitivity in open–water habitats. In contrast, the environmental factor related to the diversity of the DFADs community was solely NO_3⁻.

Based on the results of the Mantel test analysis, NO_3⁻ significantly influenced the richness index of purse seine catch communities. The scale and productivity of ecosystems play crucial roles in driving primary producer species richness, with similar patterns observed at higher trophic levels. The bioavailability of fixed inorganic nitrogen primarily regulates marine primary production [35]. In the equatorial Pacific region, trade winds cause a divergence of equatorial surface waters, leading to continuous upwelling areas. The advection of upwelling nitrates towards the poles and the continuous sinking of phytoplankton can create a north–south gradient of primary production, thereby impacting top predators in the marine food web [36]. Nagatomo et al. [37] highlight that in the mid–subtropical marine communities, all indicator species are positively correlated with integrated primary production and upper–layer nitrogen fixation rates. These communities adapt to highly productive environments, partly supported by nitrogen nutrition. In the present study, NO_3⁻ was more closely associated with the Shannon–Wiener diversity index of the DFADs community and not linked to the FSCs community. This difference may be due to the fact that catches related to DFADs are typically smaller in size [4], mainly comprising juvenile skipjack, yellowfin tuna, and bigeye tuna, whereas fishing on FSCs mainly targets large adult yellowfin and bigeye tuna [38]. Sinopoli et al. [8] suggest that small–sized fishes typically reside a short distance away from DFADs, and the survival and growth of juveniles are strongly influenced by nutrient concentrations.

Based on the findings of the current study, the ONI is closely related to the Shannon–Wiener diversity index, the Margalef richness index, and Pielou’s evenness index of the FSCs community. Large–scale climate changes such as El Niño and the Southern Oscillation can cause widespread temperature changes, directly affecting the habitat of mid–upper layer fish [1]. During El Niño events, the thermal and feeding habitats of fish in the western Pacific are vertically compresses, increasing the formation of dense fish schools [25]; this, in turn, impacts the diversity, richness, and evenness of the FSCs community. Studies have shown that under El Niño conditions, the sea surface temperature in the Central Pacific Ocean is anomalously increased, significantly affecting the surrounding ocean currents, upwelling, water temperature, chlorophyll concentrations, and other factors and leading to significant changes in the biomass of free–swimming organisms [39,40].

The pH is positively correlated with the Margalef richness index of the FSCs community. As one of the chemical properties of the ocean, pH is an important abiotic factor affecting fish communities. Mohammadi [41] states that pH serves as a controlling, loading, guiding, and even lethal factor for marine species, thus exerting a significant impact on the catch community.

Species interactions are crucial components of ecological and evolutionary challenges. Due to their interdependence, they are key elements in holistic ecosystem conservation strategies [42]. Most research has traditionally focused on the impacts of negative interactions (competition, predation, and parasitism) on species distribution. However, recent studies suggest that positive interactions (facilitation and mutualism) can also play significant roles [43]. The results of this investigation (Figure 4) show that the relationships among species in WCPO tuna purse seine catches differ between FSCs and DFADs communities. This indicates that there are not only strong competitive relationships but also mutualistic or symbiotic relationships among the catch species.

In both the FSCs and DFADs communities, yellowfin tuna was positively correlated with silky shark and bigeye tuna, indicating that these species have similar ecological requirements and behavioral patterns in the WCPO marine ecosystem. Yellowfin tuna primarily inhabits the top of the thermocline and the mixed surface layer, whereas bigeye tuna resides in deeper water layers [44]. However, around fish aggregating devices and floating objects, bigeye tuna tends to be closer to the surface, overlapping vertically with yellowfin tuna [45]. The almost identical feeding peak times of yellowfin tuna and bigeye tuna (from 8 AM to 12 PM) also indirectly confirm the overlap in their distribution [46]. Hutchinson et al. [47] observed that during most silky shark habitat behaviors, the daytime depth range is narrower and shallower, and the nighttime vertical oscillation range is between the surface and the bottom of the thermocline, overlapping with the yellowfin tuna distribution activities.

The correlation with skipjack tuna showed opposite trends in FSCs and DFADs communities (significantly negative in FSCs and significantly positive in DFADs). The daily variation pattern of purse seine catches was generally consistent with the feeding patterns of the fish schools. According to a previous study, free–swimming tuna schools disperse to feed during the day and regroup at night [48]. As a diurnal predator, yellowfin tuna is usually actively feeding during the day, especially at sunrise and sunset [49]. Actually, purse seine operations on FSCs (from 4 PM to 7 AM the next day) also cover the active feeding period of yellowfin tuna. The yellowfin tuna species mainly feed on crustaceans and squid larvae [50], with a feeding ecology similar to that of skipjack tuna, facilitating food competition between them. However, there are highly diverse animal groups near fish aggregating devices and floating objects, making tuna feeding habits more complex [51]. Additionally, the size and morphology within DFADs schools vary, with yellowfin tuna and skipjack tuna often being caught as juveniles, making them more vulnerable to predators [52,53]. These juvenile individuals exhibit positive correlations through behaviors of mutual predator avoidance.

In FSCs communities, compared to DFADs ones, there was a negative correlation between yellowfin tuna and black marlin. This difference may be related to the depth of the thermocline upper boundary. In areas with a shallower thermocline upper boundary, the success rate of catching free–swimming schools is higher, whereas fish associated with DFADs are more likely to enter the thermocline to forage, making their catch rate less related to the thermocline depth [54]. The habitat of black marlin is largely confined to the surface mixed layer [55]; they rarely pass through the thermocline and do not stay long in water 5–10 °C colder than the mixed layer. As apex predators in the food chain, black marlin mainly feed on mackerel species [56]. Thus, in free–swimming schools, black marlin and yellowfin tuna have a predator–prey relationship.

There were also other positive correlations in FSCs and DFADs communities. In FSCs communities, there was a positive correlation between silky shark, and blue marlin, and giant manta, most likely because silky shark and giant manta form groups in various ways and for a variety of reasons. Passive aggregation is the main driving factor for the spatial distribution of elasmobranch species, which also explains the correlation between silky shark and other species [57]. In DFADs communities, rainbow runner and rough triggerfish were positively correlated to mackerel scad, common dolphinfish, and TRI. As an opportunistic generalist, rough triggerfish consumes almost any available food; this species mainly lives in the open ocean but is occasionally found near the continental shelf [58]. Another distinct characteristic is that they form large schools under floating objects in tropical oceans worldwide [59], which explains the high correlation between rough triggerfish and durgons nei as both are triggerfish species. Forget et al. [60] performed an analysis of the average visitation time of rough triggerfish and rainbow runner to DFADs and pointed out that rainbow runner and rough triggerfish are usually associated with the same floating object; individuals of these two species do not venture far from the FAD. The correlation between common dolphinfish, rainbow runner and rough triggerfish is mainly reflected in foraging. The “bait supply” hypothesis suggests that the motivation for individual or small groups of predators (such as common dolphinfish) to gather under floating objects is to prey on other associated fish [48]. Quantifying the aggregations or social interactions of marine fish, compared with those of smaller freshwater bony fish species, presents significant challenges. Future research should further explore the mechanisms of coexistence among marine fish.

Although the relative importance index (IRI) is used in this study to assess dominant species, its limitations as a composite index are evident. The IRI integrates different dimensions the—proportion of the number ($N_i$), proportion of the weight ($W_i$), and frequency of occurrence ($F_i$)—into a single index, introducing significant ecological ambiguity. The multiplicative effect of $F_i$ disproportionately amplifies the importance of rare species with high $W_i$ (such as large fish) [61]. Furthermore, in this study, frequency ($F_i$) is defined as the “frequency of occurrence across the entire catch sample”. While it reflects the spatial distribution of species, it fails to distinguish their actual feeding selection [62], this further weakens the reliability of the IRI in understanding trophic relationships. Future research should incorporate stomach content data to explore more about the feeding patterns of specific species.

5. Conclusions

This study reveals significant diversity differences between free–swimming schools (FSCs) and drifting fish aggregating devices (DFADs) catch communities in the WCPO purse seine fisheries. Changes in fishing strategies have resulted in variations in species diversity across different habitats. In this study, species coexistence and co–occurrence mechanisms are primarily demonstrated through species correlations. In both FSC and DFADs communities, yellowfin tuna is positively correlated with bigeye tuna and silky shark. In the FSCs community, yellowfin tuna is negatively correlated with skipjack tuna and black marlin, whereas in the DFADs community, yellowfin tuna is positively correlated with skipjack tuna and shows no correlation with black marlin. This suggests that fishing methods influence species’ feeding behavior and vertical distribution. Moreover, the diversity indices are significantly influenced by sample size, sampling effort, and natural environmental variability, leading to varying impacts of environmental factors on the diversity of different fish communities. The NO_3⁻ significantly impacts the diversity of both FSCs and DFADs communities. The pH is closely linked to the Margalef richness index of the FSCs community, and the ONI affects all three indices of the FSCs community, emphasizing the higher mobility of free–swimming tuna schools and their susceptibility to environmental conditions. Whilst in this study, data from the purse seine fishery in the WCPO were used, the simulation framework, conclusions, and general recommendations drawn from this study are also relevant to other epipelagic fish community monitoring projects. Moving forward, it is essential to conduct comprehensive monitoring and assessment of the biodiversity, population dynamics, and interspecies interactions of epipelagic fish communities in the WCPO.