Do risk-prone behaviours compromise reproduction and increase vulnerability of fish aggregations exposed to fishing?

Human disturbances can prompt natural antipredator behaviours in animals, affecting how energy is traded-off between immediate survival and reproduction. In our study of male squaretail groupers (Plectropomus areolatus) in India's Lakshadweep archipelago, we investigated the impact of fishing pressure on anti-predatory responses and reproductive behaviours by comparing a fished and unfished spawning aggregation site and tracking responses over time at the fished site. Using observational sampling and predator exposure experiments, we analysed fear responses (flight initiation distance, return time), as well as time spent in vigilance, courtship, and territorial defence. Unpaired males at fished sites were twice as likely to flee from simulated predators and took longer to return to mating territories. Contrastingly, males in the presence of females at both sites took greater risks during courtship, fleeing later than unpaired males, but paired males at the unfished site returned earlier. Our findings suggest that high fishing pressure reduces reproductive opportunities by increasing vigilance and compromising territorial defence, potentially affecting mate selection cues. Altered behavioural trade-offs may mitigate short-term capture risk but endanger long-term population survival through altered reproductive investment. Human extractive practices targeting reproductive aggregations can have disruptive effects beyond direct removal, influencing behaviours crucial for population survival.

Methods
Refer to paper and Supplementary Materials (ESM)

Methods
2.1.Field sites
Data were collected from a ‘fished’ squaretail grouper spawning aggregation site (Bitra) during the new moon phase in January and from an ‘unfished’ site (Site 2) in February 2023 and 2024, in the Lakshadweep archipelago (details in electronic supplementary material 1, site characteristics). Historical data wasere also used from the fished site from January 2013 and 2014 (i.e., when it was in an ‘’unfished’ state, [20]). Sampling, site and aggregation characteristics are summariszed in table 1. Two prominent modes of active in-water reef fishing were observed at the fished site: hook-and-line and spearfishing, both involveing fishers free-diving in the water (RK, personal observation, details in ESM, electronic supplementary material 1). Anti-predator behavioural responses have been demonstrated in response to both fishing scenarios [16,17,21].

2.2.Squaretail grouper spawning aggregations
Squaretail groupers aggregate at both study sites for 5–6 days around the new moon from December to March [20]. Male groupers arrive 2–3 days earlier, establishing temporary territories (5–15 m2) that are vigorously defended from neighbours. Females typically join a day before the new moon and stay for 2–3 days after. The mating system is described as a ‘'lek’' [20]. Given the males’' stationary behaviour and distinct body markings facilitating easy identification, our study focussed on analysing male reproductive behaviours.

2.3.Population surveys
The total aggregation area, including the ‘'core’' lekking zone, was marked during snorkel surveys (refer to [20] and electronic supplementary material 1, for methods). Groupers were surveyed within this core area by two observers (RK, IMK) using underwater visual sampling on SCUBA along permanent 50 mx× 10 m belt transects. The number of transects varied based on the core area size (n = 5 fished site, n = 7 unfished site). Abundance estimates per transect were used to calculate average site-level population density (grouper.500 m−2). Observers also recorded fish lengths (5 -cm bins) and sex in situin situ [20]. Habitat characteristics such as structural complexity and percentage of live coral cover were assessed along three 50 -metre transects at the sites (details in electronic supplementary material 2).

2.4.Behavioural observations
Males were ‘paired’ with a female, if females were observed within male territories during focal follows and if displaying ‘'mate-guarding’' behaviours, such as maintaining close contact with females or hovering nearby (electronic supplementary material 2: table S1).

(i) Reactive response:
Reactive responses were measured at the fished and unfished sites during surveys done in 2023–2024. To determine the reactive fear response in territorial male groupers, we used ‘Flight Initiation Distance (FID)’, i.e., the distance at which the focal animal moves to avoid an approaching human observer [22]. FID is commonly used as an indicator to test for fish wariness in response to fishing [16]. We simulated predation risk by horizontally approaching, opportunistically selected territorial males (n = 72) on SCUBA at a steady pace from a standardiszed distance of five5 metres [23]. The approaching observer dropped a marker when the male began to move away and the linear distance between the grouper position (prior to movement), and the marker was measured by a second observer with a tape measure. Time away from territory is known to positively influence territory intrusion in spawning territorial damselfish [24]. Therefore, we retreated five5 metres from the male territory, immediately after the simulated disturbance to record how quickly an individual returns to its territory. In a pilot assessment of 20 males, over 70% of males returned within a minute, suggesting that immediate responses were critical for territory holders. We therefore used 120 secondsec as a cut-off time in our extended dataset.

(ii) Proactive response:
Focal video sampling was used to estimate timed-activity budgets, comparing behaviour among the fished and unfished sites in 2023–2024, and within the fished site in 2013–2014 &and 2023–2024. Observations were conducted on 108 territorial males over three days at each site, encompassing a day before, during, and after the new moon. Male behaviours were categoriszed into three states: ‘'courtship’' (male-–female interactions), ‘'territorial aggression’' (male-–male interactions), and ‘'vigilance’' (electronic supplementary material 2: table S1). The proportion of time spent in each activity was calculated by dividing the total time in different states (seconds) by the total video length (seconds).

To determine whether fishing rather than site-level population and environmental factors drove time activity budgets, we collected information on focal male body size (cm) and the number of males and females in an area of 25 m2 (i.e., competitor and mate density) from the centre of focal male territories. Habitat condition (live coral cover) and habitat structure (structural complexity) were ranked on a scale of 0–5 for each territory (electronic supplementary material 2,: table S2), owing to their influence on anti-predator behaviours like shelter-seeking [25,26] and predator avoidance [27]. Context data were not available for the historical dataset (2013-–2014), which was excluded from the analysis of drivers.

Statistical analysis
(i) Reactive response:
We used bootstrap resampling (R = 1000) to generate 95% confidence intervals to compare mean Flight Initiation Distance (FID) and territory return times between paired and unpaired males at fished and unfished sites. Non-overlapping confidence intervals indicated significant differences [28]. Proactive response: We also employed bootstrap resampling to compare activity budgets between paired and unpaired males at fished and unfished sites and within the fished site (2013–2014 pre-fishing and 2023–2024 fishing scenario). Ternary plots illustrated trade-offs between courtship, aggression, and vigilance.

(ii) Drivers of activity budgets:
Bayesian Zero-and-One Inflated Dirichlet (zoid) regressions [29] were used to model the influence of fishing status, habitat condition, male size, and social context on time activity budgets. The analysis involved 70 individuals surveyed in 2023–2024. All analysis was conducted in R (electronic supplementary material 3,: Table 4table S4).