Abstract
Introduction:
Small marine carnivores inhabiting fragmented rocky coastlines are increasingly exposed to anthropogenic pressures, yet mortality drivers in these systems remain poorly quantified.
Methods:
We analyzed 22 years of carcass records (2003–2024) and standardized necropsies to evaluate the contribution of anthropogenic causes and their spatial and temporal patterns in marine otters (Lontra felina) from north-central Chile.
Results:
We documented 173 mortality events involving 181 individuals. Space–time permutation analyses revealed three significant peri-urban clusters associated with areas of high human activity. Among cases with assigned cause, anthropogenic drivers predominated, particularly free-ranging dog attacks, followed by fisheries bycatch and non-specific trauma. Although seasonal differences were not statistically significant, anthropogenic mortality contributed proportionally more during summer months.
Discussion:
Our findings highlight terrestrial human-related pressures, particularly free-ranging dogs, as underrecognized mortality pathways in marine otters inhabiting human-dominated coastlines. Recurrent spatial clustering indicates predictable mortality risk hotspots and demonstrates the utility of spatio-temporal approaches for informing conservation and management strategies integrating terrestrial and marine systems.
1 Introduction
In recent decades, mortality events in marine mammals have increased, largely driven by anthropogenic pressures such as fisheries bycatch, ship strikes, harmful algal blooms, plastic ingestion, pollution, and emerging infectious diseases (). Understanding the causes of these mortalities is essential for identifying the factors driving these patterns and developing appropriate management strategies, particularly in human-dominated coastal systems (; ). Systematic quantification of mortality events, combined with standardized necropsies, allows for the detection of spatial and temporal trends, as well as the identification of proximate causes of death (, ; ). This approach has been applied to sea otters (Enhydra lutris), where common causes of death include white shark predation and infectious diseases. Moreover, demographic differences in mortality patterns have been observed (; ). “These data are especially relevant given that sea otters are key intertidal predators and ecosystem engineers whose presence influences the structure and function of coastal marine communities (; ; ; ; ; ; ). Consequently, increased mortality in otter populations may have cascading effects on ecosystem stability.
In the southeastern Pacific, the marine otter (Lontra felina), the second exclusively marine otter species and substantially smaller than the sea otter, occupies fragmented rocky intertidal habitats from northern Peru (9°S) to southern Chile (56°S) (; ). It inhabits narrow rocky intertidal and is typically found at low population densities along the Chilean coast. (; ). This species is currently listed as Endangered (), facing multiple anthropogenic threats including historical overhunting for fur, urban coastal development, habitat degradation and fragmentation, competition with fisheries, tidal surges, and the impact of invasive species, that are increasingly concentrated in peri-urban coastal environments (; ; ; ) and, more recently, mortality events suggestive of highly pathogenic avian influenza (H5N1) have been reported in marine fauna in Chile, including marine otters under emergency surveillance protocols (; ).
Despite its conservation status, relatively few mortality events of marine otters have been documented. Reported causes include poisoning (e.g., rodenticides in Vila-Vila, Peru; ), entanglement in fishing gear (e.g., nets and crab pots; ; ), and notably, attacks by free-ranging domestic dogs in Chile and Peru (, ; ; ). In addition, some mortalities have been associated with environmental phenomena such as the El Niño Southern Oscillation (ENSO) (), while others remain of unknown cause ().
Since 2017, increased public awareness, social media engagement, and support from non-governmental organizations have led to a rise in the reporting of marine otter strandings (MOS) and mortalities, particularly in north-central Chile, and around major urban areas. These trends are complemented by data collected by Chile’s National Fisheries and Aquaculture Service (SERNAPESCA), which maintains official records of marine fauna strandings nationwide. Using a 22-year mortality dataset integrating carcass records and necropsies, we evaluated the relative contribution of anthropogenic drivers and identified spatially explicit mortality risk patterns in an endangered marine mammal from north-central Chile. We hypothesized that anthropogenic causes would dominate assigned mortalities and that mortality events would exhibit non-random spatial clustering associated with human-dominated coastal areas.
2 Materials and methods
2.1 Data collection
Information on mortality events of marine otters from the north-central coast of Chile, encompassing regions from Valparaíso to Coquimbo, was collected from the National Fisheries Service (SERNAPESCA) and the Maritime Technological Directorate (DIRECTEMAR), including data on location, date, sex, and possible cause of death, covering the period from January 2003 to December 2024. Anthropogenic drivers were defined as causes of death directly or indirectly associated with human activities, including dog attacks, fisheries bycatch, and trauma occurring in human-dominated coastal areas. The study area includes peri-urban coastal zones characterized by tourism, artisanal fisheries, domestic dog presence, and coastal urban development, all of which may influence both mortality risk and detection probability. Additional records were obtained from (1) literature searches conducted in both English and Spanish, and (2) gray literature, including proceedings from past conferences, Chilean newspapers, magazines, and local reports.
2.2 Carcass collection and necropsy
From 2003 to 2024, we collected 62 carcasses from fishermen, SERNAPESCA officers, local communities, and researchers within the same geographical range covered by SERNAPESCA’s official marine fauna strandings database for north-central Chile, and researchers from the Lontra Foundation. Necropsies were performed by veterinarians and trained researchers with experience in marine mammal pathology and stranding response, at the National Museum of Natural History of Chile and the Universidad Santo Tomás veterinary laboratories in Viña del Mar, Chile, following and general recommendations for marine mammal mortality diagnostics (). For each necropsy we recorded sex, age class, body condition, morphometrics, and macroscopic lesions, and we annexed ancillary information on possible causes of death. Causes of death, or probable causes of death, were inferred from gross necropsy findings, field reports, and available photographic or audiovisual evidence. Cases were classified into the following categories: (1) dog attack, characterized by canine puncture wounds and extensive subcutaneous hemorrhage; (2) bycatch/drowning, characterized by net marks, abrasions on the rostrum or pectoral region, froth in the airways, and pulmonary congestion or edema; (3) non-specific trauma, defined as blunt-force polytrauma without pathognomonic lesions; and (4) undetermined cause, when the available evidence was inconclusive. In a small number of cases, additional information was obtained from marine otters that died during rehabilitation. For these animals, clinical history, signs observed during treatment, and rehabilitation records were incorporated into the mortality assessment. Lung flotation (docimasia) was performed when needed to support drowning diagnosis, though not definitive on its own ().
2.3 Spatial visualization and spatio-temporal analyses
Event locations were visualized in QGIS v3.8.2 () using WGS 1984. These analyses were conducted to identify spatially explicit mortality risk hotspots associated with human activity (). Global spatial autocorrelation was summarized with Moran’s I (). Spatio-temporal clusters were evaluated with the space–time permutation scan statistic implemented in SaTScan v10.0.2 (; ; ; ). The model used only event locations and start dates under a null of random spatio-temporal distribution, scanning for windows with higher-than-expected case counts. A “case” was a single stranding event (regardless of the number of individuals) at one location/date. The maximum temporal window was set to 3 months (capturing the four seasons), and the maximum spatial window to a 4-km radius, consistent with marine otter home-range scales ().
2.4 Monthly etiological mortality composition
We estimated monthly etiological composition as the conditional probability P(cause | month). Values were then standardized within each month (column-standardized). Low-frequency categories were pooled as “Other”, and “Unknown” denotes cause not determined. Plots were produced in R () with the tidyverse () and ggplot2 (), using viridis () and scales () for 0–1 heatmaps with months annotated; Cramér’s V was obtained via vcd ().
2.5 Descriptive and time series analyses
Descriptive analyses characterized MOS events from January 2003 to December 2024. We quantified monthly counts and decomposed an additive time series because random fluctuations appeared approximately homoscedastic. Decomposition used decompose (R package TTR; ) and functions from tseries () to estimate seasonal and trend components (see also ). We inspected ACF/PACF to assess temporal independence. All analyses were conducted in R version 4.5.3 ().
To assess the association between month and cause of death, we applied a chi-square test with Monte Carlo p-values (10,000 replicates), reporting Cramér’s V as an effect size. We used standardized Pearson residuals with a Benjamini–Hochberg false discovery rate adjustment (BH method; ) to control for multiple comparisons. These tests were implemented in R (v4.x) using the functions chisq.test(simulate.p.value = TRUE, B = 10000), assocstats() from the vcd package, and p.adjust().
3 Results
3.1 Marine otter mortality events
Between February 2003 and December 2024, a total of 173 MOS was recorded, involving 181 individual marine otters. Of all documented events, 98.3% (n = 170) were single-individual events. There were three multi-individual events: two with two individuals each (Damas Island, Coquimbo Region; Colcura Beach, Biobío Region) and one with seven individuals at Punta de Choros Cove (northern zone). Although the study focuses on north-central Chile, a small number of records from other regions of the Chilean coast were also included.
Across years, 2023 had the highest number of events (n = 44). No events were recorded in 2008, 2011, or 2015. Monthly counts peaked in February (n = 25), followed by January (n = 23) and September (n = 22). This corresponds to an average minimum marine otter mortality of 8.5 individuals per year over 22 years. Since 2016, when systematic recording of carcass findings began, the average annual minimum marine otter mortality increased to 17.8 individuals per year. Grouping by season, counts were highest in summer (n = 58) and lowest in winter (n = 34); however, no statistical differences were found among seasons (H = 4.11, p = .25).
The spatial distribution of mortality events shows that the central region (Valparaíso Region) recorded the highest number of events (81; 47.4%), followed by the north-central region (Coquimbo Region) with 44 events (25.7%). In the southern zone, Los Ríos and Los Lagos Regions recorded three and two events, respectively (2.9% combined). These records represent a small proportion of the dataset. No mortality events were recorded in the central-south and southernmost regions (including O’Higgins, Maule, Ñuble, La Araucanía, Aysén, and Magallanes) during the study period.
3.2 Demographic patterns
Among the individual variables recorded from the mortality events, 65.4% (n = 117) of the individuals corresponded to adult specimens, 17.3% (n = 31) to juveniles, and 8.9% (n = 16) to calves. In 15 (8.4) cases, it was not possible to determine the age class. Regarding sex, 26.8% (n = 48) of the individuals were males and 16.8% (n = 30) females. However, sex could not be determined in 56.4% (n = 101) of the cases. Additionally, one of the female specimens was found to be pregnant with two fetuses (Table 1).
Table 1
| Sex | Male | Female | Unknown | |||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|
| Age class | Adult | Juvenile | Calf | Unknown | Adult | Juvenile | Calf | Unknown | Adult | Juvenile | Calf | Unknown |
| Year | ||||||||||||
| 2003 | 1 | – | – | – | – | – | – | – | – | – | – | 1 |
| 2004 | – | – | – | – | – | – | – | 1 | – | – | – | – |
| 2005 | – | – | – | – | – | – | – | – | – | – | – | 1 |
| 2006 | 2 | – | – | – | – | – | – | 1 | – | – | – | 1 |
| 2007 | 1 | – | – | – | – | – | – | – | – | – | – | – |
| 2009 | – | – | – | – | – | – | – | – | – | – | – | 2 |
| 2010 | 1 | – | – | – | – | 2 | – | – | – | – | – | 1 |
| 2012 | – | – | – | – | – | 1 | – | – | – | – | – | – |
| 2013 | 3 | – | – | – | – | – | – | – | – | – | – | – |
| 2014 | – | – | – | – | – | – | – | – | 1 | – | – | – |
| 2016 | 4 | – | – | – | 5 | 1 | – | – | 1 | 1 | – | 1 |
| 2017 | 2 | – | – | – | – | – | – | – | 1 | 1 | – | 1 |
| 2018 | – | – | – | – | 2 | – | – | – | 3 | 3 | 1 | – |
| 2019 | 3 | 1 | 1 | – | 3 | – | 1 | – | 1 | 2 | 1 | – |
| 2020 | 5 | – | 1 | – | – | – | – | – | 6 | 3 | 1 | – |
| 2021 | 4 | – | – | – | 1 | 2 | – | – | 4 | – | – | 3 |
| 2022 | 5 | 2 | 2 | – | 3 | – | – | – | 9 | 2 | – | 2 |
| 2023 | 1 | 1 | – | – | 1 | – | – | – | 31 | 5 | 4 | – |
| 2024 | 5 | 3 | – | – | 2 | 2 | 2 | – | 6 | 1 | 2 | – |
| Total | 37 | 7 | 4 | 0 | 17 | 8 | 3 | 2 | 63 | 18 | 9 | 13 |
Annual marine otter (Lontra felina) mortality (2003–2024), classified by sex and age class.
“Unknown” indicates cases where sex or age could not be determined.
3.3 Spatial cluster analyses
Consistent with the overall distribution described above, MOS events occurred throughout the Chilean coast but showed a clear spatial structure (Figure 1a). Given the limited international familiarity with Chile’s administrative divisions, for descriptive purposes, the coastline was grouped into northern, central, and southern zones. Events were grouped into northern (e.g., Atacama and Coquimbo), central (e.g., Valparaíso), and southern (e.g., Biobío to Magallanes) zones. Higher numbers of events were observed in the northern and central zones. Spatial autocorrelation analysis indicated a significant clustering of MOS events (Global Moran’s I, p = .003). Space–time permutation analysis identified three statistically significant clusters (Table 2; Figure 1), located in northern (Caldera and Punta de Choros) and central Chile (Maitencillo). These clusters collectively involved up to seven stranded individuals across all events. Chronologically, clusters were detected in July 2016 (Cluster 1), January 2022 (Cluster 3), and July 2023 (Cluster 2), suggesting a recurring pattern in both time and space.
Figure 1
Table 2
| N | Zone | Latitude | Longitude | Radius (km) | Time frame | Obs. MOS | Exp. MOS | Obs./Exp. | p-value |
|---|---|---|---|---|---|---|---|---|---|
| 1 | North | -29.18 S | - 71.49 W | 0 | July 2016 to September 2016 | 7 | 0.4 | 17.4 | < 0.05 |
| 2 | North | -27.13 S | -70.90 W | 6.1 | July 2023 to September 2023 | 3 | 0.12 | 25.57 | < 0.05 |
| 3 | Central | -32.58 S | -71.45 W | 3.46 | January 2022 to March 2022 | 4 | 0.31 | 13.02 | < 0.05 |
Statistically significant spatio-temporal clusters (p < 0.05) detected using a space–time permutation model based on marine otter (Lontra felina) stranding data from February 2003 to December 2024 in north-central Chile.
Obs., observed; Exp., expected.
3.4 Time series analysis
Marine otter mortality events (MOS) occurred irregularly throughout the study period, with higher numbers in recent years (Figure 2a). The trend component (Figure 2b) shows an overall increase in MOS overtime. Seasonal decomposition revealed two recurring peaks per year, suggesting a biannual pattern (Figure 2c). These peaks were typically observed during late summer (January–February) and early spring (September), while winter (July) and early summer (December) showed consistently low values. The most pronounced seasonal factors were recorded for January (0.4), February (0.4), and September (0.3); the lowest were for July (-0.3) and December (-0.4). No significant autocorrelation or periodicity was detected (Figure 3). The autocorrelation (ACF) and partial autocorrelation (PACF) plots confirmed that all lag coefficients remained within the 95% confidence intervals (p > 0.05), indicating the events were not driven by cyclic temporal processes but occurred sporadically.
Figure 2
Figure 3
3.5 Causes of mortality and macroscopic pathological findings
In 62% (n = 111) of cases, the cause of death was undetermined (no evidence, no necropsy or non-diagnostic macroscopic lesions). In the remaining 38.5% (n = 69), the cause or probable cause was assigned based on macroscopic pathological findings, on-site reports by SERNAPESCA officers, and photographic/audiovisual evidence obtained from social media. Among cases with an assigned cause of death (38.5% of cases; n = 69), dog attacks were the most frequent identified cause (22; 31.9%), followed by bycatch (15; 21.7%) and non-specific trauma (13; 18.8%).
A total of 62 necropsies (34.6%) were performed, either partially or completely, depending on the condition of the carcass. In specimens classified as condition 5 (mummified), only external macroscopic samples were collected and analyzed. Based on the necropsies and macroscopic pathological findings, 14 mortality events were associated with dog attacks, four with bycatch (Figures 4A, B), and 13 with non-specific trauma. Additionally, three cases were linked to drowning in thermoelectric turbines, three to fossil fuel poisoning (Figures 4C, D), and two were attributed to hunting and potential infectious diseases.
Figure 4
Of the 22 deaths attributed to dog attacks, 15 (68.2%) occurred in the Valparaíso Region. Of the 15 events associated with bycatch, 9 (60%) occurred in the Coquimbo Region. Additionally, 75% (3 of 4 cases) of the deaths related to intoxication and 100% (3 of 3 cases) of the deaths due to drowning in thermoelectric plants were recorded in the Valparaíso Region. Two mortality events recorded in 2023 outside the core study area (northern Chile) were associated with avian influenza, with one case reported in the Arica y Parinacota Region and another in the Tarapacá Region.
3.6 Monthly etiological composition
The etiological composition varied among months (Figure 5). Summer (Jan–Feb), and, to a lesser extent, September, showed a higher relative contribution of anthropogenic categories (e.g., dog attack, bycatch, non-specific trauma, drowning in thermoelectric plants), whereas Unknown (cause not determined) remained relatively constant year-round and Hunting/Poisoning/Potential infectious disease were sporadic. Because each column shows within-month proportions (sum = 1), the heatmap reflects composition rather than variation in monthly effort. The global month×cause association was not significant (χ² [121] = 136.99, Monte-Carlo p = .533; Cramér’s V = 0.262). Cell-wise standardized residuals with FDR correction detected one robust deviation: bycatch in September was over-represented (q = 0.0037); apparent increases of anthropogenic categories in Jan–Feb did not remain significant after FDR.
Figure 5
4 Discussion
Our results support the hypothesis that anthropogenic pressures, particularly free-ranging dogs and fisheries interactions, are key drivers of mortality in the endangered marine otter (Lontra felina), and that these events exhibit non-random spatial clustering. Mortality events can have substantial demographic effects on otter populations (; ), particularly in key intertidal species such as the marine otter, which already faces multiple conservation pressures (). In Chile, we documented 173 mortality events involving 181 individuals between 2003 and 2024. These patterns highlight the role of human-dominated coastal environments and land–sea interactions in shaping mortality risk. These patterns may reflect increased reporting effort and improved detection, particularly in recent years, rather than a true increase in mortality.
The annual number of recorded mortality events is lower than those reported for other otters: 37.3–118 individuals per year for Enhydra lutris in California (; ) and 20–30 individuals per year for Lutra lutra in several European countries (; ; ). South American estimates are closer to ours (e.g., 8.6 individuals per year for Lontra longicaudis in southern Brazil; 5 individuals per year for L. felina in southern Peru; ; ). However, cross-study comparisons must be made cautiously because effort, coastline accessibility, carcass persistence, and sampling scale differ markedly. In Chile, coordinated necropsy/retrieval protocols were only consolidated after 2016–2017, and increased reporting effort likely over-represents recent years relative to earlier periods.
Among individuals with determined sex, adult males predominated (48 vs. 30 females), although sex could not be determined in a large proportion of cases (n = 101). This bias may reflect larger home ranges and more frequent long-distance movements typically associated with males, which in turn increase their exposure to anthropogenic hazards such as roads, fisheries, and pollution (; ; ; ). For marine otter, documented significantly larger and more variable home ranges in males along the Chilean coast, often encompassing risk-prone habitats near human settlements or fisheries. These findings suggest that movement ecology plays a central role in shaping sex-biased mortality patterns and may explain the disproportionate number of stranded males observed. This pattern may reflect increased exposure to anthropogenic risks associated with movement behavior rather than intrinsic differences in vulnerability.
Monthly counts peaked in January–February (austral summer), a period of intense coastal use associated with tourism, recreational beach activities, artisanal fishing activity, greater human mobility along the shoreline, and increased presence of free-ranging or off-leash dogs associated with higher human presence during summer months, all of which may increase both mortality risk and carcass detection. A formal test found no significant differences among seasons, and the overall month×cause association was not significant, although bycatch in September was over-represented (q = 0.0037). Because the heatmap uses within-month proportions, these patterns should be interpreted cautiously.
Dog attacks were the most frequent identified cause of mortality within cases with assigned causes, although this result should be interpreted cautiously given the limited number of cases. This pattern is unusual compared with other species, where dog-related mortality rarely exceeds 5% and is typically associated with roadkill (; ; ). In Chile, free-ranging domestic dogs have been documented as a significant threat to coastal wildlife, particularly in peri-urban and developed areas where human–dog interactions are common . Our results provide quantitative evidence that this threat may represent a dominant and spatially structured mortality driver in peri-urban coastal systems. Similar patterns have been reported in other coastal wildlife in Chile, including seabirds (; ; ), suggesting that these mortality drivers may operate at a broader ecosystem level. We did not explicitly evaluate local dog density or urban structure, which may influence the spatial distribution of dog-related mortality. Given the limited number of cases with assigned causes, these interpretations should be considered exploratory. Future studies should evaluate local dog density, ownership practices, and urban structure to better understand spatial patterns of dog-related mortality.
The spatial clustering of mortality events in developed coastal zones indicates predictable risk hotspots associated with human activity. This is particularly relevant considering recent coastal urban expansion, recreational use of beaches, and the presence of free-ranging domestic dogs. These interactions not only affect marine otters directly through attacks but may also indicate broader ecological risks, as similar dynamics have been associated with declines in co-occurring fauna such as seabirds (; ). Given the frequency and severity of these events, this threat should be treated as a key conservation concern for marine otters, requiring targeted mitigation strategies and further investigation. Despite these limitations, the analysis remains useful for identifying areas with higher reporting frequency, which can help guide monitoring and management efforts when interpreted cautiously.
Bycatch in gillnet fisheries remains a key threat. Our largest mortality event, involving seven otters in Punta de Choros, occurred during a period of intense coastal fishing activity targeting conger eel (Genypterus spp.). In this area, nearshore set nets overlap with rocky-shore and shallow coastal habitats frequently used by marine otters for foraging and movement, which may increase the risk of entanglement and drowning ().
Non-specific trauma likely reflects a pooled category of carcasses presenting multiple external injuries that may be associated with human-related interactions (e.g., boat strikes, entanglement) or natural interactions (e.g., interspecific aggression), but without pathognomonic lesions. This ambiguity underscores the need for complete necropsies in the central coast of Chile. The seasonal concentration in summer and the frequent detection in peri-urban areas further suggest a possible anthropogenic contribution (; ; ).
Hydrocarbon exposure/ingestion was inferred from the presence of tar-like material in the gastrointestinal tract and external contamination of the fur in individuals recovered near industrialized coastal zones. In one case, an individual admitted alive to a rehabilitation center later died, and necropsy confirmed the presence of hydrocarbon residues in the gastrointestinal tract, supporting this diagnosis.
Human-related causes dominated assigned mortalities (dog attacks, bycatch, industrial hazards) (; ; ), and sensitivity analyses strengthened the summer pattern without changing the non-significant month×cause test (). The high proportion of undetermined cases in 2023 should be interpreted with caution due to H5N1 emergency protocols prioritizing biosafety over complete necropsies (FAO; WHO; ).
Three spatio-temporal clusters were detected along the coast (including Caldera–Punta de Choros and Papudo/Maitencillo). Although cluster sizes were small, their recurrence in peri-urban areas suggests spatially structured risk, but results should be interpreted cautiously. This result may reflect, at least in part, uneven detection probability, as carcasses occurring in accessible and frequently visited coastal areas are more likely to be observed and reported, particularly where citizen reports and social media networks contribute to stranding detection. This may bias the apparent spatial distribution of mortality events. This potential bias has been also described for other mustelid species, as for example in Lutra lutra mortality series dominated by roadkills along accessible corridors (; ; ). Because of this limitation, caution when interpreting trends based on carcass recovery dynamics has been suggested for these studies (). This reinforces the need to account for potential reporting bias when interpreting spatial mortality patterns. Therefore, we favor conservative interpretation and recommend standardized carcass recovery/necropsy protocols and effort-corrected models that include proxies of search/reporting (tourism pressure, accessibility/mobility, municipal waste, fishery effort) to separate true incidence from detection.
These findings justify prioritizing preventive actions during summer, a period when multiple anthropogenic threats converge. Notably, recent data from 2025 and early 2026 indicate an increase in dog-related mortality, reinforcing the urgency to address this threat more proactively. Specifically, efforts should focus on dog control in peri-urban coastal areas, alongside measures to mitigate bycatch, improve waste management, and investigate non-specific trauma. This seasonal focus is consistent with broader otter literature, where mortality is often linked to human-related causes such as traffic collisions and fisheries interactions in freshwater otters (e.g., Enhydra lutris, Lutra lutra), and similar patterns have been described for marine otter in Chile and Peru (; ; ; ; ). Reinforcing these efforts during summer, when mortality peaks, may significantly reduce threats and enhance early response capacity for this endangered species.
Management actions should prioritize control of free-ranging dogs in coastal settlements, targeted bycatch mitigation during high-risk periods, and the implementation of standardized carcass recovery and necropsy protocols, as well as the integration of terrestrial and marine management frameworks, which will be essential to effectively reduce mortality risk in this species. In addition, effective mitigation will require stronger collaboration with coastal communities, including environmental education, responsible dog ownership, and local participation in monitoring and stranding response efforts. Given recent infectious-disease events, coordinated and biosecurity health surveillance is also warranted. Strengthening standardized monitoring will be essential to reduce the proportion of undetermined cases and improve causal inference in future assessments.
Statements
Data availability statement
The raw data supporting the conclusions of this article will be made available by the authors, without undue reservation.
Ethics statement
This study was based on opportunistic data from stranded marine otters and necropsies performed under permits issued by the Chilean National Fisheries and Aquaculture Service (SERNAPESCA). No experimental procedures or handling of live animals were conducted, and therefore ethical approval was not required.
Author contributions
FT: Conceptualization, Data curation, Formal analysis, Funding acquisition, Investigation, Methodology, Project administration, Resources, Software, Supervision, Validation, Visualization, Writing – original draft, Writing – review & editing. GM: Investigation, Methodology, Writing – original draft, Writing – review & editing. MA-R: Conceptualization, Data curation, Formal analysis, Investigation, Methodology, Software, Validation, Visualization, Writing – original draft, Writing – review & editing. BT-B: Conceptualization, Investigation, Methodology, Writing – original draft. JAl: Data curation, Formal analysis, Methodology, Project administration, Software, Validation, Writing – original draft, Writing – review & editing. JT: Methodology, Writing – review & editing, Investigation, Resources, Validation, Writing – original draft. CC: Data curation, Methodology, Supervision, Writing – review & editing. FC-J: Conceptualization, Investigation, Methodology, Writing – review & editing, Resources. PA: Investigation, Methodology, Validation, Writing – review & editing, Conceptualization. JL: Conceptualization, Investigation, Methodology, Resources, Visualization, Writing – review & editing. JAr: Conceptualization, Investigation, Methodology, Validation, Visualization, Writing – review & editing. SF: Investigation, Methodology, Resources, Writing – review & editing. PS: Investigation, Methodology, Resources, Writing – review & editing. MSep: Conceptualization, Data curation, Formal analysis, Methodology, Supervision, Validation, Writing – review & editing. MSeg: Conceptualization, Data curation, Formal analysis, Investigation, Methodology, Supervision, Validation, Writing – review & editing. GM-V: Conceptualization, Data curation, Methodology, Supervision, Validation, Writing – review & editing. TH-S: Data curation, Investigation, Methodology, Validation, Writing – review & editing. GP: Conceptualization, Data curation, Formal analysis, Methodology, Supervision, Validation, Visualization, Writing – review & editing.
Funding
The author(s) declared that financial support was not received for this work and/or its publication.
Acknowledgments
We thank the staff of the National Fisheries and Aquaculture Service (SERNAPESCA) from the regions of Arica y Parinacota, Antofagasta, Atacama, Coquimbo, Valparaíso, and Libertador General Bernardo O’Higgins for their continued collaboration and field support. We also thank the School of Veterinary Medicine of Universidad Santo Tomás, Viña del Mar campus, for its institutional support. We are grateful to Jhoann Canto from the National Museum of Natural History; to Guillermo “Willy” Barrera from Turismo Punta de Choros; and to Katherine Ramírez and Pablo Valladares from Universidad de Tarapacá. We further acknowledge the students and colleagues of Universidad Santo Tomás and Universidad de Valparaíso, as well as the Veterinary Medicine students of Universidad Santo Tomás (Viña del Mar campus) and the Marine Biology students of Pontificia Universidad Católica de Chile and Universidad de Valparaíso for their commitment and support throughout the different stages of this work.
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.
Generative AI statement
The author(s) declared that generative AI was used in the creation of this manuscript. Generative AI tools were used to assist with language editing and text refinement. All scientific content, analysis, and interpretations were developed 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.
References
1
Alvarado-RybakM.ToroF.AbarcaP.ParedesE.Español-JiménezS.SeguelM. (2020a). Pathological Findings in Cetaceans Sporadically Stranded Along the Chilean Coast. Front. Mar. Sci.7. doi: 10.3389/fmars.2020.00684
2
Alvarado-RybakM.ToroF.Escobar DoderoJ.KinsleyA. C.SepúlvedaM. A.CapellaJ.et al. (2020b). 50 Years of Cetacean Strandings Reveal a Concerning Rise in Chilean Patagonia. 10. doi: 10.1038/s41598-020-66484-x
3
ApazaM.FigariA. (1999). Mortandad de aves marinas durante “El Niño 1997–1998” en el litoral sur de San Juan de Marcona, Ica-Perú. Rev. Peruana Biología6, 110–117.
4
AvilaI. C.KaschnerK.DormannC. F. (2018). Current global risks to marine mammals: Taking stock of the threats. Biological Conservation.221, 44–58. doi: 10.1016/j.biocon.2018.02.021
5
BenjaminiY.HochbergY. (1995). Controlling the false discovery rate: A practical and powerful approach to multiple testing. J. R. Stat. Soc. Ser. B.57, 289–300. doi: 10.1111/j.2517-6161.1995.tb02031.x. PMID:
6
Bravo-NaranjoV.JiménezR. R.ZuletaC.RauJ. R.ValladaresP.PiñonesC. (2019). Selección de presas por perros callejeros en el humedal Estero Culebrón, Coquimbo, Chile. Gayana83, 102–113. doi: 10.4067/S0717-65382019000200102
7
Burek-HuntingtonK.GillV.BronsonN.TuomiP. (2005). ). A pictorial guide to sea otter anatomy and necropsy findings.
8
Burek-HuntingtonK. A.GillV. A.BerrianA. M.GoldsteinT.TuomiP.ByrneB. A.et al. (2021). Causes of mortality of northern sea otters (Enhydra lutris kenyoni) in Alaska from 2002 to 2012. Front. Mar. Sci.8, 630582. doi: 10.3389/fmars.2021.630582
9
CastillaJ. C.PaineR. T. (1987). Predation and community organization on Eastern Pacific temperate rocky shores. Rev. Chil. Hist. Nat.60, 131–151.
10
ChanT.BaryshevaA.GanleyY.WhiteP.van GeelN.WitcombT.et al. (2017). Integrating spatio-temporal patterns of marine mammal strandings into monitoring programmes. Methods Ecol. Evol.8, 1604–1613. doi: 10.1111/2041-210X.12814. PMID:
11
CórdovaO.RauJ. R.SuazoC. G.ArriagadaA. (2009). Estudio comparativo de la ecologíaalimentaria del depredador de alto nivel trófico Lontra felina (Molina) en Chile. Rev. Biología Marina y Oceanografía44, 429–438.
12
CórdovaL.RauJ. R. (2016). Dieta de Lontra felina en costas rocosas del norte de Chile. Latin Am. J. Aquat. Res.44, 874–883. doi: 10.3856/vol44-issue4-fulltext-13
13
ErlingeS. (1967). Home range of the otter Lutra lutra L. in southern Sweden. Oikos18, 186–209. doi: 10.2307/3565098
14
EstesJ. A.PalmisanoJ. F. (1974). Sea otters: Their role in structuring nearshore communities. Science185, 1058–1060. doi: 10.1126/science.185.4156.1058. PMID:
15
EstesJ. E.SmithN. S.PalmisanoJ. F. (1978). Sea otter predation and community organization in the western Aleutian Islands, Alaska. Ecology59, 822–833.
16
GarcésA.PiresE. (2024). Biological and conservation aspects of otter mortality: A review. Conservation2, 498–518. doi: 10.3390/conservation2030034. PMID:
17
GarnierS.RossN.RudisB.CamargoA. P.SciainiM.SchererC. (2021). Colorblind-friendly color maps for R.viridisLite Package Version 0.4.2. Available online at: https://CRAN.R-project.org/package=viridisLite.
18
GerberL. R.TinkerM. T.DoakD. F.EstesJ. A.JessupD. A. (2004). Mortality sensitivity in life-stage simulation analysis: a case study of southern sea otters. Ecol. Appl.14, 1554–1565. doi: 10.1890/03-5155
19
GoetheF. (1964). “ Das verhalten der musteliden,” in Handbuch der zoologie, vol. 8. ( Walter de Gruyter, Berlin), 1–80.
20
HauerS.AnsorgeH.ZinkeO. (2002). Mortality patterns of otters (Lutra lutra) from eastern Germany. J. Zool.256, , 361–368.
21
HeggbergetT. M.MyrbergetS. (1991). “ Sex and age distribution in Eurasian otters (Lutra lutra) killed by human activity,” In: ReutherC.RöchertR. (eds.), Proceedings of the V International Otter Colloquium. (Habitat) 6, 123–125.
22
KendallM. G.StuartA. (1983). The Advanced Theory of Statistics. (London: Griffin) 3, 410–414.
23
KulldorffM. (1997). A spatial scan statistic. Commun. Stat.- Theory Methods26, 1481–1496. doi: 10.1080/03610929708831995. PMID:
24
KulldorffM. (2024). SaTScan v10.0.2: Software for the spatial, temporal and space-time scan statistics (Calverton, Maryland, USA: Information Management Services, Inc).
25
KulldorffM.HeffernanR.HartmanJ.AssunçãoR.MostashariF. (2005). A space-time permutation scan statistic for disease outbreak detection. PloS Med.2, e59. doi: 10.1371/journal.pmed.0020059. PMID:
26
KulldorffM.NagarwallaN. (1995). Spatial disease clusters: Detection and inference. Stat. Med.14, 799–810. doi: 10.1002/sim.4780140809. PMID:
27
LanszkiJ.OroszE.NagyD. (2018). Mortality causes and body dimensions of otters (Lutra lutra) determined by post-mortem analysis in Hungary. Acta Veterinaria Hungarica66, 1–10.
28
MangelJ. C.WhittyT. S.Medina-VogelG.Alfaro-ShiguetoJ.CáceresC.GodleyB. J. (2011). Latitudinal variation in diet and patterns of human interaction in the marine otter. Mar. Mammal Sci.27, E14–E25. doi: 10.1111/j.1748-7692.2010.00414.x. PMID:
29
Medina-VogelG.BoherF.FloresG.SantibáñezA.Soto-AzatC. (2007). Spacing behaviour of marine otters (Lontra felina) in relation to land refuges and fishery waste in central Chile. J. Mammal.88, 487–494. doi: 10.1644/06-MAMM-A081R1.1. PMID:
30
Medina-VogelG.Delgado RodríguezC.AlvarezP. R. E.BartheldV. J. L. (2004). Feeding ecology of the marine otter (Lutra felina) in a rocky seashore of the south of Chile. Mar. Mammal Sci.20, 134–144. doi: 10.1111/j.1748-7692.2004.tb01144.x
31
Medina-VogelG.MerinoL. O.Monsalve AlarcónR.ViannaJ. D. A. (2008). Coastal–marine discontinuities, critical patch size and isolation: implications for marine otter conservation. Anim. Conserv.11, 57–64. doi: 10.1111/j.1469-1795.2007.00155.x
32
MeyerD.ZeileisA.HornikK. (2006). The strucplot framework: Visualizing multivariate data. J. Stat. Software17, 1–48.
33
MillerM. A.TinkerM. T.GillV. A.MoriartyM. E.DoroffA.BurdinA. M.et al. (2020). A review of mortality in southern sea otters (Enhydra lutris nereis): Patterns, drivers, and knowledge gaps. Front. Mar. Sci.7, 582. doi: 10.3389/fmars.2020.00582
34
MoranP. A. P. (1950). Notes on continuous stochastic phenomena. Biometrika37, 17–23. doi: 10.1093/biomet/37.1-2.17
35
MoxleyJ. H.NicholsonT. E.Van HoutanK. S.JorgensenS. J. (2019). Non-trophic impacts from white sharks complicate population recovery for sea otters. Ecol. Evol.9, 6378–6388. doi: 10.1002/ece3.5218
36
PfeifferD. U.RobinsonT. P.StevensonM.StevensK. B.RogersD. J.ClementsA. C. A. (2008). Spatial analysis in epidemiology (Oxford: Oxford University Press).
37
PhilcoxC. K.GroganA.MacdonaldD. W. (1999). Patterns of otter Lutra lutra road mortality in Britain. J. Appl. Ecol.36, 748–762. doi: 10.1046/j.1365-2664.1999.00441.x
38
PizarroT. (2008). Mortality of the marine otter (Lontra felina) in southern Peru. IUCN Otter Specialist Group Bull.25, 94–99.
39
PugliaresK. R.BogomolniA.TouheyA.TouheyK. M.HerzigS. M.HarryC. T.et al. (2007). Marine mammal necropsy: An introductory guide for stranding responders and field biologists (Woods Hole, Massachusetts: Woods Hole Oceanographic Institution Technical Report WHOI-2007-06), 1–131. doi: 10.1575/1912/1823
40
QuintelaF. M.Da SilvaF. A.AssisC. D.AntunesV. C. (2012). Data on Lontra longicaudis mortality in southeast and southern Brazil. IUCN Otter Specialist Group Bull.29, 5–8.
41
QGIS Development Team (2019). QGIS geographic information system, version 3.8.2 (Beaverton, Oregon, USA: Open-Source Geospatial Foundation Project).
42
R Core Team (2024). R: A language and environment for statistical computing (Vienna: R Foundation for Statistical Computing).
43
SáezK.SimeoneA.SchlatterR. P.CortésM.LobosV. (2016). Historia natural del piquero Peruano (Sula variegata) en los acantilados de La Quirilluca, Valparaíso, Chile. El Hornero31, 1–6.
44
SandersonC. E.AlexanderK. A. (2020). Unchartered waters: Climate change likely to intensify infectious disease outbreaks causing mass mortality events in marine mammals. Global Change Biology.26, 4284–4301. doi: 10.1111/gcb.15163
45
Servicio Nacional de Pesca y Acuicultura (2023). Influenza aviar: SERNAPESCA información oficial sobre detecciones de H5N1 en fauna marina y medidas operativas de vigilancia (Valparaíso, Chile: SERNAPESCA website).
46
SielfeldW. (1983). Mamíferos marinos de Chile ( Ediciones de la Universidad de Chile).
47
SielfeldW.CastillaJ. C. (1999). Estado de conocimiento y conservación de Lontra felina en Chile. Estudios Oceanológicos18, 69–79.
48
SimeoneA.Luna-JorqueraG.BarcelóG.LunaN.RiquelmeV.RuizJ.et al. (2021). Spatial and temporal patterns of beached seabirds along the Chilean coast. Biol. Conserv.256, 109069. doi: 10.1016/j.biocon.2021.109069. PMID:
49
SinghG. G.MarkelR. W.MartoneR. G.SalomonA. K.HarleyC. D.ChanK. M. (2013). Sea otters homogenize mussel beds and reduce habitat provisioning in a rocky intertidal ecosystem. PloS One8, e65435. doi: 10.1371/journal.pone.0065435
50
SotoR. (1996). Estructura gremial de un ensamble de depredadores de la zona intermareal rocosa en Chile central. Investigaciones Marinas24, 97–105.
51
TinkerM. T.HatfieldB. B.HarrisM. D.AmesJ. A. (2016). Dramatic increase in sea otter mortality from white sharks in California. Mar. Mammal Sci.32, 309–326. doi: 10.1111/mms.12276
52
TraplettiA.HornikK. (2019). tseries: time series analysis and computational finance. (Vienna, Austria: R Foundation for Statistical Computing).
53
UlrichJ. (2020). TTR: technical trading rules.(Vienna, Austria: R Foundation for Statistical Computing).
54
ValquiJ. (2012). The marine otter Lontra felina (Molina 1782): A review of its present status and prospects. Mamm. Biol.77, 75–83. doi: 10.1016/j.mambio.2011.08.004. PMID:
55
ValquiJ.RheingantzM. L. (2015). Lontra felina. IUCN Red List Threatened Species, e.T12303A21937779. doi: 10.2305/IUCN.UK.2015-2.RLTS.T12303A21937779.en
56
ViannaJ. A.AyerdiP.Medina-VogelG.MangelJ.ZeballosH.ApazaM.et al. (2010). Phylogeography of the marine otter (Lontra felina): Historical and contemporary factors determining its distribution. J. Heredity101, 676–689. doi: 10.1093/jhered/esq081. PMID:
57
VillatoroF. J.Naughton-TrevesL.SepúlvedaM. A.StowhasP.MardonesF. O.Silva-RodríguezE. A. (2019). When free-ranging dogs threaten wildlife: public attitudes toward management strategies in southern Chile. J. Environ. Manage.229, 67–75. doi: 10.1016/j.jenvman.2018.09.035
58
WickhamH. (2016). ggplot2: Elegant graphics for data analysis (New York: Springer).
59
WickhamH.AverickM.BryanJ.ChangW.McGowanL. D.FrançoisR.et al. (2019). Welcome to the tidyverse. J. Open Source Softw4, 1686. doi: 10.21105/joss.01686
60
WickhamH.SeidelD. (2022). scales: Scale functions for visualization. (Vienna, Austria: R Foundation for Statistical Computing).
61
WilliamsonM. J.StockinK. A.OramsM. B.RoeW. D. (2021). Reviewing marine mammal necropsy diagnostics: understanding causes of mortality in a changing world. J. Comp. Pathol.184, 1–17. doi: 10.1016/j.jcpa.2021.10.004. PMID:
62
World Health Organization (2023). Human infection caused by avian influenza A (H5N1) – Chile. (Geneva, Switzerland: World Health Organization).
Summary
Keywords
anthropogenic mortality, free-ranging dogs, Lontra felina, marine otter, mortality risk, peri-urban environments, spatio-temporal clustering
Citation
Toro F, Mallea G, Alvarado-Rybak M, Toro-Barros B, Alarcón J, Trivelli J, Calvo C, Cruz-Jofre F, Abarca-Mora P, Leiva J, Arriagada J, Frenkel S, Salah P, Sepúlveda M, Seguel M, Medina-Vogel G, Huerta-Santander T and Pavez G (2026) Using spatio-temporal clustering to identify mortality patterns and anthropogenic drivers in the marine otter (Lontra felina) in north-central Chile. Front. Mar. Sci. 13:1835525. doi: 10.3389/fmars.2026.1835525
Received
20 March 2026
Revised
24 April 2026
Accepted
06 May 2026
Published
03 June 2026
Volume
13 - 2026
Edited by
Peijun Zhang, Chinese Academy of Sciences (CAS), China
Reviewed by
Minmin Chen, Anqing Normal University, China
Casandra Gálvez, National Polytechnic Institute (IPN), Mexico
Updates
Copyright
© 2026 Toro, Mallea, Alvarado-Rybak, Toro-Barros, Alarcón, Trivelli, Calvo, Cruz-Jofre, Abarca-Mora, Leiva, Arriagada, Frenkel, Salah, Sepúlveda, Seguel, Medina-Vogel, Huerta-Santander and Pavez.
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: Frederick Toro, frederick.toro.c@gmail.com
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.