Inter-stock congruence of long-term changes in macroparasite communities of a small pelagic fish ¿An effect of climate change?

Abstract

The impact of climate change on the distribution and sustainability of pelagic fish remains poorly understood, even though it can cause drastic shifts in species distribution, reproduction, migratory behavior, and recruitment. In turn, parasites can reveal important details about the biology of their hosts and the changes they undergo. The aim of this study was to evaluate whether changes occurred in the parasite community structure of the Argentine anchovy, Engraulis anchoita, after three decades, during which environmental changes have been documented in the Argentine Sea. Parasite assemblages of anchovies captured throughout their distribution in the Argentine Sea in 2022 were examined and compared with those of anchovies collected in the same regions between 1993 and 1995. Multivariate analyses were performed by assigning fish to three spatial groups (belonging to two different stocks), identified from the first sampling period (Northern Buenos Aires, Southern Buenos Aires, and Patagonian). The dataset included 42,290 metazoan parasites belonging to 13 species. Samples differed significantly among groups and between periods. A decrease in parasite abundances was observed for most species across all three groups, resulting in similar patterns of dissimilarity between periods. The congruence of these changes across three groups over a broad geographic scale suggests that such variations are likely driven by large-scale processes, possibly climate change, rather than by random, short-term, cyclical, or local variability.

Supplementary Information

The online version contains supplementary material available at 10.1007/s00436-026-08653-x.

Introduction

Marine ecosystems are expected to be affected by anthropogenic climate change, which is projected to lead to ocean warming, acidification, deoxygenation, and reductions in near-surface nutrients (Kwiatkowski et al. 2020; Marcogliese 2025). Accordingly, the pelagic zone, and its mid-trophic level fauna are expected to be impacted (Ariza et al. 2022) as a direct consequence of changes in their metabolic needs and their interspecific interactions such as competition or predation (Lefort et al. 2015), but also indirectly through the influence of environmental variables on phytoplankton and zooplankton assemblages (Benedetti et al. 2021).

Despite the world importance of pelagic fisheries, the environmental drivers of spatial distribution limits and population fluctuations remain not well understood for many species (Muhling et al. 2017). This is because because many species have high migratory capabilities and mobility, being able to move away from temporarily unsuitable habitats. Therefore, projecting climate impacts has been particularly difficult for small coastal species (such as anchovies and sardines), characterized by population fluctuations spanning orders of magnitude due to periods of continuous high and low spawning stock biomass (Swalethorp et al. 2022), short generation times and large interannual fluctuations of abundance in a boom-and-bust dynamics. Consequently, the direction and magnitude of the impact of climate change on pelagic fishes is still uncertain and still not corroborated by observation-based statistical studies (Ariza et al. 2022).

The Argentine anchovy is one of the most important pelagic fishery resources inhabiting the Argentine Sea. It has a wide distribution on the continental shelf, reaching latitudes south of the San Jorge Gulf, in Patagonian waters (47°30’ S) (Angelescu and Cousseau 1967). Two stocks have been historically considered in the region, a Northern and a Southern or Patagonian stock, separated at approximately 41° S (Hansen 1994; Timi 2003). Recently the presence of a third stock in the San Matias Gulf has been proposed, based on parasitological evidence (Marcotegui et al. 2024) and daily length growth of anchovies juveniles (Brown et al. 2014). This is the most abundant pelagic species in the southwest Atlantic Ocean (Defeo and Vasconcellos 2020), where also represents a key role in the regional trophic web as the main prey of other valuable fishery resources, as well as for marine birds and mammals (Angelescu 1982; Yorio et al. 2017; Ciancio et al. 2018, 2021; Machovsky-Capuska et al. 2020). Therefore, and due to their biomass, anchovies play an important role in the ecosystem, exerting major control on the dynamics of upper and lower trophic levels (del Favero et al. 2017).

Like other small pelagic fish species, theArgentine anchovy exhibits considerable year to year variability in abundance (Parietti et al. 2025), which affects other trophic levels in the food chain (Cury et al. 2000) and its populations are likely affected by long-term regime shifts (Auad and Martos 2012).

In the Southwestern Atlantic, the Brazil Current, composed by subtropical waters influencing the oceanography north of the Argentine Sea, has been reported as one of the most extensive and intense surface warming hotspots in the global ocean (Hobday and Pecl, 2014; Franco et al. 2020a). The Brazil Current has intensified and shifted southward, causing significant warming in the Southern Brazilian Bay and the Río de la Plata (Franco et al. 2020b). Between 2003 and 2017, warming reached up to 1 °C per decade, affecting both coastal waters and deeper layers (Franco et al. 2020a). The entire Río de la Plata water column is warming (> 0.4 °C/decade), and surface isotherms are moving south (Franco et al. 2020b), and marine heatwaves in the area have also become more frequent, longer, and more intense, particularly since 2000 (Luz Clara et al. 2025). This rise of sea temperatures in northern Argentina has affected the distribution of some fisheries and the composition of the captures in the region (Franco et al. 2020b), dealing with pronounced tropicalization of demersal species since 2013 (Gianelli et al. 2019; Alvarez Perez and Sant’Ana 2022). Similarly, in Patagonian waters, two opposite trends in sea surface temperature have been recorded in the last decade: a warming area between 42° − 45° S, and a cooling area between 49° − 52° S (Risaro et al. 2022). Additionally, a breakpoint with a constant increase in positive temperature anomalies for the region, has been detected in the area (Parietti et al. 2025). All these environmental changes could be affecting the anchovy populations, generating the need to assess the effects of future environmental changes on resource availability and fishery management.

Parasites with complex life cycles are particularly susceptible to environmental threats (Marcogliese 2001; Sures et al. 2023),with those species infecting ectothermic hosts being at greatest risk(Harvell et al. 2002a, b; Cizauskas et al. 2017). Therefore, parasitic organisms offer excellent systems for measuring and monitoring such effects (Marcotegui et al. 2025). Unfortunately, the impacts of climate change on parasite diversity remains mostly undocumented (Carlson et al. 2017; Wood 2025) with variable responses beingrecorded (Fiorenza et al. 2020; Welicky et al. 2021; Preisser et al. 2022; Wood and Vanhove 2023). However, alterations in the composition and structure of parasite assemblages are a likely outcome influenced by global change factors (Marcotegui et al. 2025). A remarkableexample in the study area is a digenean parasite of anchovies as larvae and of penguins as adults, Cardiocephaloides physalis, that has shown a significant decline in its abundance and become extinct in some areas in recent decades (Marcotegui et al. 2025). This is not the first change in parasitism recorded regionally, as several parasites of a coastal fish, Cynoscion guatucupa, have either increased or declined in abundance during the same period (Lanfranchi et al. 2024).

Ideally, to identify the ongoing changes in parasitism and to understand their causes, the temporal trajectory of multiple parasite species should be addressed (Wood 2025). Therefore, in this work, other parasites of E. anchoita are analyzed, at different latitudes, with the aim to assess possible changes on the parasite communities by comparing with previous data in different stocks along the Argentine Sea.

Materials and methods

Anchovy samples were obtained from both research cruises and commercial catches between August and November 2022 at different latitudes in the Argentine Sea (Fig. 1; Table 1). A total of 207 anchovies with total length ≥ 120 mm were sampled, prior assignment to the two known stocks in the Argentine Sea, the North and the Patagonian (P) (Hansen et al. 1984; Hansen 2004) from where previous samples were available. Following Timi (2003), samples from the Northern stock were in turn divided into two groups: North Bonaerense (NB) and South Bonaerense (SB) which resulted in parasitologically homogeneous groups during the first period. Samples of the recently proposed stock from San Matías Gulf and adjacent waters (Marcotegui et al. 2024) were not considered. Fish were deep frozen on board and preserved at −18 °C until examination. After thawing, fish were measured for total length (TL) in mm. Data of 520 anchovies caught between 1993 and 1995 from a previous work (Timi 2003) were used for comparative purposes, following the same geographical design. For most analyses, samples from NB and SB groups, captured in different years, were pooled together (Fig. 1; Table 1).

Fig. 1.

Distribution map of Engraulis anchoita samples. Samples colored by stock, period and year. NB: North Bonaerense, SB: South Bonaerense, P: Patagonia. First period: reddish circles; second period: bluish circles

Table 1.

Composition of samples of Argentine anchovy, Engraulis anchoita from three samples and two temporal periods (data of pooled samples of different years in bold)

Stock	Sample	Date	Latitude S	Longitude W	Capture	N	Lt (mm) ± S.D.
North Bonaerense	NB1	1993–1995	37°57’00’’−38° 55’00’’	56°47’0’’−59°01’0’’	RC	188	145.6 ± 14.3
	NB-93	Nov. 1993	38°51’00’’−38°55’00’’	58°21’0’’−59°01’0’’	RC	105	148.3 ± 11.0
	NB-94	Aug. 1994	37°57’00’’	56°47’0’’	RC	48	140.8 ± 16.0
	NB-95	Sep. 1995	38°51’00’	58°43’0’’	RC	35	143.9 ± 18.7
South Bonaerense	SB1	1994–1995	40°15’06’’−40°26’03’’	61°03’02’’−61°03’12’’	RC	144	157.48 ± 14.9
	SB1-94	Aug. 1994	40°15’06’’	61°03’02’’	RC	102	163.3 ± 12.4
	SB1-95	Oct. 1995	40°26’03’’	61°03’12’’	RC	42	143.3 ± 10.4
Patagonian	P1	Nov./Dec. 1993	43°22’36’’−45°22’00’’	63°13’00’’−66°43’00’’	RC	188	161.7 ± 14.8
North Bonaerense	NB2	Aug. 2022	38°49’03’’	58°07’28’’	CO	38	146.5 ± 14.6
South Bonaerense	SB2	Aug. 2022	39°81’00’’	60°42’00’’	RC	34	143.5 ± 14.8
Patagonian	P2	Jul/Aug. 2022	43°07’00’- 43°07’24’S	64°22’24’−64°28’0’S	RC/CO	135	145.8 ± 13.8

CO: commercial fishery; RC: research cruise; S.D.: Standard Deviation

Each fish was dissected under a stereoscopic microscope. The body surface, eyes, gills, branchial and body cavities, viscera (stomach, intestine, liver, heart, gonads and mesenteries), gas bladder, kidneys, and musculature were examined separately. Macroparasites were recovered from gills, body cavities, stomach, pyloric caeca, and intestine, counted and identified following standard procedures for each taxon (Pritchard and Kruse 1982; Berland 1984).

Fish length was compared across samples by a Kruskal-Wallis One Way Analysis of Variance on Ranks (1 × 6 factorial design, ‘sample’ as fixed factor) (Zar 1999), followed by post-hoc Dunn’s pairwise tests (Dunn 1964). Prevalence and mean abundance (sensu Bush et al. 1997) were calculated for each parasite species in each sample. No microparasites (myxozoans and coccidians), previously reported from anchovies (Timi and Sardella 1998), were considered for the analysis because they can not be counted and only presence-absence data are available.

The species richness (S) was calculated for each individual fish (infracommunity sensu Bush et al. 1997) and the Brillouin´s indices of diversity (HB) and evenness (E) for those anchovies harboring at least 2 parasite species (Magurran 1988). For each group, the abundance of those species with prevalence higher than ten in at least one sample, as well as the infracommunity descriptors were compared between periods by one-way permutational multivariate analysis of the variance (PERMANOVA, Anderson et al. 2008) based on Euclidean distances and including TL as a co-variable and “sample” as a factor (1 × 2 factorial design). Following Anderson et al. (2008) a permutation of residuals under a reduced model was used as a method of permutation. A sequential sum of squares (Type I SS) was applied because TL was introduced as a covariable.

Multivariate analyses were used to assess if parasite communities showed similar or disparate temporal patterns across the three groups. For this, Bray-Curtis similarity indices (Magurran 1988) on parasite abundance were calculated between all possible pairs of hosts. The similarity of parasite infracommunities across samples was evaluated, previous square-root transformation of data to down-weight the importance of very abundant species so that the less dominant species played some role in determining similarity among samples (Clarke and Gorley 2015).

A non-parametric multidimensional scaling (nMDS, Clarke and Gorley, 2015) was performed to assess potential geographic or temporal variations across samples. The distances between the centroids of each sample were plotted by bootstrap averaging the original dataset (50 iterations with replacement) to show differences among samples more clearly (Clarke and Gorley 2015). For this, mean values were visualised in nMDS plots, using as many dimensions as necessary in order to match the original data matrix as closely as possible (setting a ρ = 0.99 correlation coefficient), and their fitting was quantified by means of a stress value. The analysis was carried out considering the annual samples from the first period both individually and combined.

The influence of the origin of assemblages was further analyzed by means of canonical analysis of principal coordinates (CAP) (Anderson and Willis 2003), testing for significant differences by a permutation “trace” test, after 9999 permutations. Over-parameterization was prevented by choosing the number of PCO axes (m) that maximized success on a “leave-one-out” allocation to groups cross-validation method (Anderson and Robinson 2003).

As both nMDS and CAP analyses do not take into account the influence of host size on the significance of the differences among samples, the differences in infracommunity structure between samples were tested using a one-way PERMANOVA on the Bray-Curtis dissimilarity matrix, in a similar way than in the case of infracommunity descriptors. Because PERMANOVA is sensitive to differences in multivariate dispersion between groups (sensu homogeneity of variances, which can inflate Type I error even when centroids have identical locations), the same models were tested for differences in dispersion using the routine PERMDISP (Anderson et al. 2008). Dispersions were measured as distance to the centroid, and each term in the analysis was tested using 9999 permutations.

At the component community level (Bush et al. 1997), an nMDS was performed on the Bray-Curtis similarity index on square root-transformed data of prevalence. A hierarchical agglomerative clustering was applied to the component communities using group-average linking, and resemblance levels were overlaid on the nMDS plot (Clarke and Gorley 2015).

All similarity and distance measures, as well as all multivariate analyses, were implemented in PRIMER V7 and PERMANOVA + for PRIMER package (Anderson et al. 2008; Clarke and Gorley 2015).

Results

Significant differences in fish length were observed across samples (H₍₅₎: 149.8, P < 0.01), due to pairwise differences between anchovies from P1 and SB1 regarding all the other samples (all P < 0.05) (Table 1).

A total of 10 parasite species were found in the whole sample, totalizing 2898 individuals, with no parasite-free anchovy individuals observed. Variable values of prevalence (Table 2) and mean abundance (Table 3) were observed across samples. Considering the seven species selected (Table 4), thirteen of the 20 comparisons between periods showed significant differences in abundance. Three species showed significant changes in their abundance across the two periods analysed in all anchovy populations. Specifically, the abundance of P. heterocotyle and S. polymorphus decreased, whereas that of Anisakis sp. increased during the same time frame. The highest number of differences were observed for anchovies from NB and P samples (5 species), with only three species differing in SB. Hysterothylacium aduncum was the only species with similar abundance between periods in all the samples. The lack of influence of fish size was noteworthy for adult, short-lived, parasites. Burdens of larval parasites, on the other hand, depended largely on fish length in most cases, evidencing significant interaction with the period of sampling. The results of all PERMANOVAs are given as Online Resource (Table S1).

Table 2.

Microhabitat and prevalence (%) of macroparasites of Engraulis anchoita across six samples from the Argentine Sea caught at two periods (1993–1995 and 2022)

Parasite	Microhábitat	NB1	SB1	P1	NB2	SB2	P2
Monogenea
Pseudanthcotyloides heterocotyle	Gills	19.2	19.4	9.6	-	-	1.5
Mazocraes australis	Gills	5.8	5.6	5.3	-	2.9	3.0
Digenea
Lecithochirum microstomum	Stomach	62.8	59.7	25.5	2.6	-	1.5
Parahemiurus merus	Stomach	14.4	11.1	-	-	2.9	-
Cestoda
Diphlyllobothriidae gen sp.	Body cavity	-	0.7	-	-	-	-
Bothriocephalus sp.	Pyloric caeca	-	0.7	1.1	-	-	2.2
Scolex polymorphus	Intestine	94.7	95.1	85.1	52.6	20.6	29.6
Nematoda
Hysterothylacium aduncum	Body cavity	49.5	78.5	84.6	94.7	97.1	89.6
Contracaecum sp.	Body cavity	68.1	32.5	81.9	28.9	38.2	48.9
Anisakis sp.	Body cavity	3.7	11.1	44.1	10.5	17.7	65.2
Terranova galeocerdonis	Body cavity	0.5	-	-	-	-	-
Acanthocephala
Corynosoma australe	Body cavity	5.8	2.1	0.5	-	2.9	-
Copepoda
Nothobomolochus cresseyi	Gills	1.1	0.7	0.5	-	-	-

NB: North Bonaerense; SB: South Bonaerense; P: Patagonian; 1: first period (1993–1995); 2: second period (2022)

Table 3.

Mean abundance ± standard deviation of macroparasites of Engraulis anchoita across six samples from the Argentine Sea caught at two periods (1993–1995 and 2022)

Parasite	NB1	SB1	P1	NB2	SB2	P2
Pseudanthcotyloides heterocotyle	0.36 ± 0.97	0.28 ± 0.69	0.15 ± 0.54	-	-	0.21 ± 0.12
Mazocraes australis	0.11 ± 0.61	0.07 ± 0.30	0.11 ± 0.74	-	0.09 ± 0.51	0.04 ± 0.27
Lecithochirum microstomum	4.40 ± 11.15	7.34 ± 38.48	2.65 ± 9.88	0.03 ± 0.16	-	0.02 ± 0.19
Parahemiurus merus	0.20 ± 0.60	0.17 ± 0.527	-	-	0.03 ± 0.17	-
Diphlyllobothriidae gen sp.	-	0.01 ± 0.08	-	-	-	-
Bothriocephalus sp.	-	0.02 ± 0.25	0.07 ± 0.81	-	-	0.04 ± 0.26
Scolex polymorphus	30.18 ± 37.01	29.21 ± 32.99	13.57 ± 14.71	9.47 ± 11.36	3.62 ± 8.056	6.26 ± 15.96
Hysterothylacium aduncum	4.44 ± 15.59	15.06 ± 35.60	29.90 ± 50.07	4.79 ± 4.33	8.71 ± 10.81	8.83 ± 12.94
Contracaecum sp.	1.78 ± 2.40	3.87 ± 3.26	2.81 ± 2.53	1.24 ± 2.58	2.56 ± 4.86	2.44 ± 5.22
Anisakis sp.	0.05 ± 0.28	0.18 ± 0.72	0.96 ± 1.57	0.24 ± 0.79	0.35 ± 1.04	3.14 ± 5.18
Corynosoma australe	0.07 ± 0.298	0.03 ± 0.20	0.01 ± 0.07	-	0.06 ± 0.34	-
Nothobomolochus cresseyi	0.01 ± 0.10	0.01 ± 0.08	0.01 ± 0.07	-	-	-

NB: North Bonaerense; SB: South Bonaerense; P: Patagonian; 1: first period (1993–1995); 2: second period (2022)

Table 4.

Significance levels (Pperms) of comparisons between two periods (1993–1995 and 2022) of abundance of selected macroparasites of Engraulis anchoita across three groups from the Argentine Sea caught

Variable	NB			SB			P
	TL	Period	TLxPeriod	TL	Period	TLxPeriod	TL	Period	TLxPeriod
Pseudanthcotyloides heterocotyle	0.052	0.020↓	0.329	0.944	0.016↓	0.667	0.715	0.002↓	0.466
Lecithochirium microstomum	0.461	0.027	0.700	0.543	0.164	0.392	0.056	0.0001	0.003
Parahemiurus merus	0.434	0.037	0.707	0.292	0.248	0.488	-	-	-
Scolex polymorphus	0.331	0.001↓	0.611	0.021	0.001↓	0.628	0.0002	0.003↓	0.131
Hysterothylacium aduncum	0.0001	0.998	0.041	0.0001	0.167	0.036	0.0001	0.435	0.0001
Contracaecum sp.	0.0001	0.164	0.070	0.0001	0.974	0.128	0.0001	0.011	0.018
Anisakis sp.	0.013	0.016↑	0.037	0.018	0.038↑	0.320	0.503	0.0001↑	0.333

NB: North Bonaerense; SB: South Bonaerense; P: Patagonia. Upward-pointing arrows indicate increases, while downward-pointing arrows signify decreases

At the infracommunity level (Fig. 2), the species richness was significantly affected by anchovy size and differed significantly between periods, being higher in the first period (Fig. 2a), with no interaction between both factors, in all the three samples (all P_perm<0.01, Online Resource Table S2). PERMDISP analyses showed no differences in multivariate dispersions between samples from SB (F_1,176: 1.24; P_perm >0.05) and P (F_1,321: 0.87; P_perm >0.05), being different those from NB (F_1,224: 4.85; P_perm =0.04). The Brillouin diversity index (Fig. 2b), on the other hand, was also affected by anchovy size in the three samples (all P_perm<0.01) but only differed between periods in SB (P_perm=0.03), with no interaction between both factors (all P_perm>0.01, Online Resource). Multivariate dispersions also varied for comparisons in SB (F_1,160: 1.24; P_perm =0.02), not differing in NB (F_1,202: 0.55; P_perm >0.05) nor in P (F_1,293: 3.31; P_perm >0.05). In the case of evenness (Fig. 2c), the anchovy size had an effect in comparisons from NB and P (P_perm=0.04 and P_perm<0.01, respectively), but not in SB (P_perm>0.05). Multivariate dispersions did not vary for comparisons in NB (F_1,201: 0.06; P_perm=0.81) and SB (F_1,160: 3.31; P_perm =0.10), but differed in P (F_1,293: 12.95; P_perm <0.01).

Fig. 2.

Comparative parasitic community metrics of Engraulis anchoita between Period 1 (1993–1995) and Period 2 (2022) across the three spatial groups: Northern Buenos Aires (NB), Southern Buenos Aires (SB), and Patagonian (P). Panels show (a) Brillouin diversity, b Brillouin evenness, and (c) species richness

When the annual samples from the first period were analyzed individually, the bootstrap-average-based nMDS ordination showed an apparent separation of samples according to their period of capture (Fig. 3), with a low value of stress (0.03). Samples from the first period were located at the left of the biplot, with an ordered north-south sequence along the first axis. Those of different years from NB clustered close to each other and together to SB from 1995. In the new period, the most different anchovies were those from P. A similar picture was observed when the samples of different years were pooled in both north Bonaerense samples (Fig. 4).

Fig. 3.

Non-metric multi-dimensional scaling plot (nMDS) of bootstrap averages (50 repetitions) of parasite infracommunities in nine samples of Engraulis anchoita caught in the Argentine Sea, based on Bray-Curtis dissimilarity of square root-transformed data of abundance. Black full symbols represent the overall centroids across all repetitions. Samples identified by Locality Code-Period/Year. Locality Code: NB: North Bonaerense, SB: South Bonaerense, P: Patagonia

Fig. 4.

Non-metric multi-dimensional scaling plot (nMDS) of bootstrap averages (50 repetitions) of parasite infracommunities in six samples of Engraulis anchoita caught in the Argentine Sea, based on Bray-Curtis dissimilarity of square root-transformed data of abundance. Black full symbols represent the overall centroids across all repetitions. Samples identified by Locality Code-Period. Locality Code: NB: North Bonaerense, SB: South Bonaerense, P: Patagonia

Indeed, higher differences were observed for samples of the same stocks between different periods, than between stocks of the same period, judging by the distances between centroids (Fig. 5), although both P samples were more similar to each other than their B counterparts.

Fig. 5.

Distance between centroids treemap based on bootstrap averaging (50 iterations with replacement) of nMDS analyses on parasite abundance in six samples of Engraulis anchoita in the Argentine Sea. Blue: between periods for each group; brown: between groups from the first period; green: between groups from the second period. The bigger the square, the larger the distance

The CAP analysis on parasite abundance showed a relatively low percentage of correct allocations (48.69%), although it was considerably higher than the value expected by chance alone (21.24%). Cross-validation results (Table 5) showed that the lowest percentages of allocation were between both B samples of each period, but the percentages improved considerably for all samples when anchovies were allocated to their period of capture rather than to their sample site.

Table 5.

Results of cross validation (leave-one-out allocation of individual anchovies to one of six samples) for the canonical analysis of principal coordinates (CAP) of parasite assemblages of Engraulis anchoita in six samples from the Argentine Sea. Rows correspond to group memberships, including the percentage of correctly classified fish to their own stock or period of capture. In bold the higher number of allocated anchovies to each sample and the highest percentage of correct allocations

Origin group							% correct allocations to same
	NB1	SB1	P1	NB2	SB2	P2	Sample	Stock	Period
NB1	128	35	12	8	4	1	68.08	72.34	93.08
SB1	62	41	32	5	2	2	28.47	31.25	93.75
P1	34	28	82	11	6	27	43.62	57.98	76.60
NB2	4	9	2	4	15	4	10.53	21.05	60.53
SB2	2	2	3	2	21	4	61.76	67.65	79.41
P2	2	12	12	8	23	78	57.78	59.26	80.74

The results of PERMANOVA analyses (Table 6) showed a significant effect of host length on the response variables and an interaction between the host length and sample, indicating that the nature of the relationship between the covariate and the multivariate response differed within different levels of the factor. Furthermore, taking into account the variations among samples due to fish size, significant differences were detected among the parasite assemblages. Post-hoc pairwise comparisons showed significant differences for all comparisons (t: 2.21–13.07; all P_perm< 0.01). Significant differences in the multivariate dispersions were observed across samples (F_5,721: 5.06; P_perm <0.001), with five pairs of samples differing from each other, namely NB1-P2 (t: 3.37; P_perm <0.01), SB1-P1 (t: 2.78; P_perm <0.01), SB1-NB2 (t: 2.23; P_perm =0.03), SB1-P2 (t: 5.14; P_perm <0.001) and P1-P2 (t: 2.36; P_perm =0.03).

Table 6.

One-factor PERMANOVA results comparing square-root transformed abundance data of parasites of Engraulis anchoita in six samples from the Argentine Sea, based on the Bray-Curtis dissimilarity measure, with host length as co-variable. P-values obtained after 9999 permutations

Source	d.f.	SS	MS	Pseudo F	P (perm)
Host length	1	1.02e5	1.02e5	82.72	< 0.001
Sample	5	3.65e5	72,951	59.37	< 0.001
Host length x Sample	5	25,786	5157.3	4.20	< 0.001
Residual	715	8.78e5	1228.7
Total	726	1.37e6

The nMDS and cluster analyses carried out at the level of component communities (Fig. 6) showed a similar pattern to those of infracommunities. Indeed, the sampling period prevailed over the sample sites in terms of similarity across samples, with a stress value of 0. When vectors corresponding to Pearson correlations of selected species were superimposed with the nMDS, H. aduncum and Anisakis sp. were associated to recent samples, whereas all the rest were indicators of parasite assemblages of the first period.

Fig. 6.

Non-metric multi-dimensional scaling plot (nMDS) and cluster analysis of parasite component communities in samples of Engraulis anchoita caught in the Argentine Sea, based on Bray-Curtis dissimilarity of square root-transformed data prevalence. Stress value = 0

Discussion

Climate change is expected to influence multiple characteristics of the pelagic marine environment, such as temperature, circulation, chemistry (pH), stratification, productivity and oxygen availability (Muhling et al. 2017); whose shifts may produce changes in spatial redistribution and ecological interactions of pelagic fishes (Muhling et al. 2017; Sarre et al. 2024). The observed temporal patterns in the populations and communities of parasites de E. anchoita, are likely a consequence of these processes. Indeed, the loss of biodiversity is generally considered to lead to decreases in parasite diversity (Lafferty 2012; Marcogliese 2023), which, however, would depend also on differential characteristics of parasites, particularly their specificity and the complexity of their life cycles (Lafferty 2012; Wood 2025).

In the present paper, except for the diphyllobothriid and copepod species, which were found at very low prevalence during the first period, the rest of species were present in the recent samples. However, like C. physalis (see Marcotegui et al. 2025), their abundance often varied over time. Among the ectoparasites with direct life cycle, only P. heterocotyle decreased in abundance in the three sample sites. This can not be attributed to changes in host density, an expected driver of parasitism (Kamiya et al. 2014), because no evidence of drastic changes in the density of E. anchoita has been recorded during the last decades (Orlando et al. 2024a, 2024b). Thus, environmental factors arose as the most probable cause of decreasing abundance for this parasite. In the case of heteroxenous parasites, by having complex life cycles that require multiple hosts, environmental changes can have an even stronger impact, because there is a greater chance that one or more of required host species will be affected, interrupting the parasite life cycle (Rohr et al. 2011; Lafferty 2012; Wood and Vanhove 2023). Furthermore, beyond influencing host density, changes in temperature and other abiotic parameters can affect life-history characteristics of both hosts and parasites, repercussing on parasite transmission and, potentially, virulence (Marcogliese 2001). Despite the absence of recurrent patterns across all stocks, most changes reflected decreases in abundance, suggesting less favorable conditions in recent years for the majority of parasites. The exception was Anisakis sp., who consistently increased in abundance in all stocks. Larval Anisakis in anchovies seem to follow the general trend of the genus recorded at global scale. Indeed, larval Anisakis have shown notable increases in abundance in both intermediate (Fiorenza et al. 2020) and definitive (Mastick et al. 2024) hosts during last decades around the world, which was attributed mostly to protection measures on marine mammals that allowed the recovery of many cetacean populations.

Because of the changes underwent by the parasite species combined, the structure of their assemblages was also affected. Indeed, the decrease in parasitism by several species in all stocks led to a lower species richness in the infracommunities, which, however, was reflected in a decrease in the diversity only for SB, but not in the other two stocks due to an increase in their respective evenness. As in the case of parasite abundance, changes in host density are not attributable as causatives of the observed patterns. Additionally, and except for the monogenean, all other parasite species are widely distributed among fish in the region (Navone et al. 1998; Timi et al. 1999a, b, 2001), which suggests that the observed decreases in parasitism are unlikely to be attributed to changes in host density in a single species.

In addition to changes in host density, variations in the ecophysiology of host-parasite interactions and shifts in behavior, movement and phenology of both parasites and hosts have been identified among the causes of alterations in parasitism induced by environmental change (Altizer et al. 2007). Indeed, ENSO (El Niño Southern Oscillation) events, which are becoming more intense and frequent, produce changes in wind and sea surface temperature anomalies, contributingto altering the pelagic ecosystem structure and communities in theSouthwestern Atlantic Ocean (Morley et al. 2025). These distributional shifts may alter the match-mismatch dynamics between pelagic fish and their parasites, leading to declines in parasitism with climate change, a trend that has already been observed in this host-parasite system (Marcotegui et al. 2025). Although there is no information about the sensitivity of Argentine anchovy parasites to environmental variations, anchovies have been identified as having low sensitivity to climate change, displaying a high potential to adjust its distribution (Gianelli et al. 2023) in response to changes in the regional warming hotspots (Alvarez Perez and Sant’Ana 2022). This ability may facilitate phenological mismatches involving parasites present in other host species, probably accounting for the decreasing burdens for most parasites.

We acknowledge that, with data obtained only from both ends of a given period, it cannot be reliably established whether the causes of the observed changes are due to large-scale processes, such as climate change, or to variations occurring at more limited spatiotemporal scales, for instance, seasonal, interannual, or even random fluctuations. The interannual similarity observed for the first three years of study in the northern regions, however, demonstrates that the results obtained are unlikely due to short-term variations. On the other hand, variations occurring simultaneously in all three stocks support that processes acting on parasite assemblages occur at broad spatial scales, to produce higher differences between periods of samples from the same stocks than those from the same period between stocks, according to the centroid distances, nMDS and CAP analyses. Although these temporal differences were less pronounced in P, all pairwise comparisons of multivariate abundance were significantly different from each other. Component communities also revealed that sampling period, rather than stock, was the primary factor influencing similarity among samples.

In brief, major changes were detected in both parasite populations and assemblages, most occurring at broad geographical scales and in congruent patterns across anchovy stocks in the last decades. Sea temperatures have increased in the Southwestern Atlantic Ocean over this period (Risaro et al. 2022; Gianelli et al. 2023; Parietti et al. 2025). It is known that such extended temperature anomalies can produce changes in the distribution of pelagic fishes (Zhai et al. 2024), leading to a decoupling of phenological relationships and producing trophic mismatches that also alter food-web structures (Edwards and Richardson 2004), reshaping marine communities and trophic interactions across decades (Alheit and Niquen 2004). Therefore, due to their vulnerability to climate variability, small pelagics have been proposed as valuable indicators for detecting regime shifts in fish stocks (Vasconcelos et al. 2024). Nevertheless, their boom-and-bust population dynamics, tightly tied to climate variability, complicate the detection and measurements of climate change (Ariza et al. 2022; Swalethorp et al. 2022). Consequently, their parasites, highly reliant on environmental factors, host ecology, and food webs structure for transmission, constitute reliable indicators and even earlier sentinels of such changes.

Supplementary Information

Below is the link to the electronic supplementary material.

Author contributions

All authors contributed to the study conception and design. Material preparation, data collection and analysis were performed by PM, MM, MP and JT. The first draft of the manuscript was written by PM and all authors commented on previous versions of the manuscript. All authors read and approved the final manuscript.

Data availability

The data that support the findings of this study are available from the corresponding author upon reasonable request.

Declarations

Ethical approval

All procedures involving animals were conducted in accordance with relevant national and international guidelines for the care and use of animals in scientific research. All fish specimens used in this study were obtained through the National Institute of Fisheries Research, under the Secretariat of Agriculture, Livestock and Fisheries of the Ministry of Economy. Importantly, no fish were sacrificed specifically for the purposes of this study.

Consent to participate

Not applicable.

Consent to publish

Not applicable.

Competing interests

The authors declare no competing interests.

Clinical trial number

not applicable.