The risky nightlife of undersized sea urchins

Time-lapse photography

Previously, 15 time-lapse videos filmed in eastern Tasmania during summer (November 2009–February 2010) were analysed for urchin movement as described in Flukes et al. (2012). Sites were specifically chosen to be spatially independent and represent one of three distinct habitat types: widespread urchin barrens (grazed areas >10⁴ m²) composed of flat rock; widespread barrens composed of boulders; and incipient barrens (grazed patches 1–10 m²). Nikon D200 digital SLR and Pentax Optio W80 digital compact cameras equipped with red lighting were installed by SCUBA divers on adjustable tripods. Camera time-lapse photos were saved every 5 min and recording ran for 12 h each night, commencing prior to sunset (19:30 hours) and continued to after sunrise (07:30 hours). The measure tool in imageJ (ver. 1.52, W. S. Rasband, US National Institutes of Health, Bethesda, MD, USA, see https://imagej.nih.gov/ij/; Schneider et al. 2012) was used to measure the diameter of the SC of each urchin, recorded in pixels. For each individual urchin, SC was measured four times and the results averaged across replicate measurements to account for different positions of the urchins during the video. For distance travelled and displacement data, only the subset of measured urchins that remained within the field of view (FOV) for the entire night were included (n = 182). Urchins which emerged from and retreated to a crevice within the FOV during any point in the video were also included for distance and displacement measurements. For each of these urchins, visibility of the urchin at the beginning and the end of the video was manually sampled. The FOV photographed by the cameras varied from ~5 to 30 m² depending on both the camera system used and the adjusted height of the tripod (0.8–3 m from the substratum, depending on topography). The position of each urchin in each frame was determined by x–y coordinates in pixels. Calibration of the time-lapse frame (mm per pixel) was obtained prior to this study, by measuring known distances between features in the FOV (Flukes et al. 2012).

Spine canopy – test diameter ratio

The test diameter (TD) and SC of 211 urchins collected from Fortescue Bay, Tasmania (incipient barren) in May–June 2023 were measured using digital callipers. A log-linear model was applied to determine the relationship between SC and TD and convert SC measurements of urchins in the overnight videos to TD.

Data analysis

The predation limit relationship between lobster size (CL) and sea urchin TD (TD = 0.6 × 5.12e^0.023CL; Ling et al. 2009) was used to classify sea urchins into three size classes based on (a) the reported size of lobsters capable of urchin predation (140 mm CL) and (b) the lobster fishery size limit (110 mm CL), as used by Smith et al. (2022). Sea urchin size classes were ‘small’ (<39 mm TD, considered non-emergent and accessible to small lobsters <110 mm CL), ‘medium’ (39–77 mm TD, considered emerging and accessible to medium lobsters 110–140 mm CL), and ‘large’ (>77 mm TD, considered emergent and accessible only to large lobsters >140 mm CL). Movement rate was determined by the proportion of frames within which an urchin was scored as moving (i.e. the urchin was in a different x–y position from the previous frame, with a minimum step length of 10 mm as in Flukes et al. 2012). Movement rate was then averaged across the urchin size class for daylight period (daylight, dusk, night, dawn) as determined by times at sunrise and sunset at each deployment site. A linear mixed effects model (lmer) was used to determine the importance of size class, daylight period, and their interaction on urchin movement during the recording period. Individual urchin was included as a random factor to account for the individual urchins measured across multiple daylight periods. Habitat type, which is known to influence urchin movement (Flukes et al. 2012), was pooled for analysis following initial analysis which showed it did not interact with size to influence movement. For distance travelled, a linear model was used to determine the importance of size class in the total distance moved by each urchin. Net displacement of each urchin was calculated as the straight-line distance between its x–y position in the first frame and in the final frame. The movement model was fitted using the package lme4 (ver. 1.1-31, see https://cran.r-project.org/package=lme4; Bates et al. 2015) as lmer(movement ~ daylight-period × size-class, (1|Urchin-ID)), the distance travelled and displacement models were fit as lm(distance ~ size-class) and lm(displacement ~ size-class). ANOVA was used to determine each model’s predictiveness and post hoc multiple comparisons were used to compare between factors (Bonferroni corrected). Interactions were explored using t-tests and visualised using the package sjPlot (ver. 2.8.12, see https://cran.r-project.org/package=sjPlot) if significant. Data analysis was conducted with R (ver. 4.1.1, R Foundation for Statistical Computing, Vienna, Austria, see https://www.r-project.org/) in RStudio (Posit Software, PBC, Boston, MA, USA, see https://posit.co/products/open-source/rstudio/) using the packages tidyverse (ver. 2.0.0, see https://cran.r-project.org/package=tidyverse; Wickham et al. 2019), lubridate (ver. 1.9.2, see https://cran.r-project.org/package=lubridate; Grolemund and Wickham 2011) and multcomp (ver. 1.4-20, see https://cran.r-project.org/package=multcomp; Hothorn et al. 2008).

Test diameter and spine canopy ratio

Centrostephanus rodgersii SC was positively correlated with TD and followed a log-linear relationship:

where y is SC and x is TD (R² = 0.87, Fig. 1). The proportionally longer spines on smaller urchins is likely a predator avoidance adaptation, as long spines are known to confer greater resistance against predators (Ling and Johnson 2009). A gap in the data relating to urchins between 41 and 60 mm TD may be attributed to a low recruitment event at this site. Despite this gap in the measured TD-SC data, the high R² value confirms the model fits well and is suitable to use.

Fig. 1.

Measured relationship between test diameter and spine canopy of Centrostephanus rodgersii (n = 211). Log-linear relationship is modelled by y = −287.924 + 112.615 × log(x).

Movement of urchins overnight

The time-lapse footage identified a total of 292 urchins ranging in size from 75 to 253 mm SC, which converted to 20–122 mm TD, this conversion facilitates association to the size at which urchins may be subject to predation by lobster (Ling et al. 2009). Using the upper limit of lobster–urchin predation size classes from Ling et al. (2009), urchins of all three size classes were identified in the time-lapse videos: ‘small’ (n = 32), ‘medium’ (n = 224) and ‘large’ (n = 36). Both medium and small urchin size classes (88% of urchins recorded) include those considered to be cryptic and inaccessible to predation. Previously, nocturnal activity levels of small urchins has not been reported, and thus their cryptic, crevice-dwelling daylight behaviour has only been considered (Andrew and Underwood 1989; Ling et al. 2009; Ling and Johnson 2012; Byrne and Andrew 2020). Here, overnight underwater filming (by Flukes et al. 2012) and identification of individual urchin sizes has allowed size-specific monitoring of nocturnal urchin movement and proved that small urchins leave their crevices during nocturnal hours (Fig. 2). Of those urchins whose track remained within the FOV all night, only 50% of small urchins were visible at the beginning and end of the filming period (daytime), which was lower than medium (65–70%) and large (67–72%) urchins, indicating that small urchins have heightened cryptic behaviour.

Fig. 2.

Photographs taken from the timelapse footage at one site showing Centrostephanus rodgersii during daylight (a) and night (b) to demonstrate the crevice-dwelling behaviour seen during the day and the increase in movement during the night. Small black circles are C. rodgersii. Also visible in the field of view is kelp cover (left), boulders, and camera or light tripod legs.

Size class, daylight period and their interaction all had a significant effect on the movement rate of sea urchins (ANOVA: size F_(2,388) = 8.04, P < 0.001; daylight period F_(3,2540) = 366.0, P < 0.001; interaction F_(6,2539) = 3.15, P = 0.004). Overall, urchin movement was highest during night-time (z_{(night–dawn)} = 26.7, P < 0.001; z_{(night–day)} = 41.7, P < 0.001; z_{(night–dusk)} = 18.6, P < 0.001) and showed the lowest movement during the daylight period (z_(day–dawn) = −10.1, P < 0.001; z_(dusk–day) = 13.3, P < 0.001; z_{(night–day)} = 41.7, P < 0.001, Fig. 3a, Table 1), corroborating the results found by Flukes et al. (2012). The interaction between size and daylight period was explained by small urchins moving significantly less during the dusk period (Supplementary Fig. S1, t_{(small–medium)} = −4.14, P < 0.001; t_{(small–large)} = −4.20, P < 0.001), when the medium and large urchins had increased movement rates, indicating a delayed onset of nocturnal activity in smaller urchins (Fig. 3a). At night-time, small urchins also moved marginally less than the larger size classes (F_(2,270) = 3.36, P = 0.036). Post hoc tests indicated there was limited difference in movement between small and medium size urchins (z_(small-med) = −2.39, P = 0.05), but no significant difference between the other groups (z_{(medium–large)} = −0.68, P = 1; z_{(small–large)} = −2.26, P = 0.07). Increased sample size may further elucidate the pattern seen here, but this result suggests that small urchins are just as active overnight as larger size classes. Habitat was confirmed not to interact with size to influence movement (F_{(4, 368)} = 1.31, P = 0.27). High and consistent levels of movement across all size classes of urchin overnight demonstrates that all sizes of urchins, even those that are cryptic in the day, are nocturnally active. The reduction in movement of all urchins towards dawn aligns with the previously described behaviour of C. rodgersii returning to their crevices in the morning and becoming diurnally cryptic (Byrne and Andrew 2020).

Fig. 3.

Differences in urchin movement rate and distance between size classes. All error bars represent ±s.e. (a) Mean movement rate between each daylight period (daylight, dusk, night, dawn). (b) Mean distance travelled (mm) by each urchin size class across the 12-h recording period. (c) Mean distance travelled scaled to test diameter (mm) by each urchin size class across the 12-h recording period.

The interaction between urchin size and timing of movement may be influenced by variations in ambient light within the habitat, as suggested in Flukes et al. (2012). Our research revealed that smaller urchins exhibit a delayed increase in movement at dusk compared to larger size classes (Fig. 3a), meaning that their period of heightened movement is relatively shorter, perhaps due to lesser demand for food (Hart and Chia 1990). During the daytime, smaller urchins tend to seek refuge in crevices, whereas larger urchins can often be found on open rock areas. Consequently, if changes in ambient light serve as the trigger for urchin activity, it is possible that smaller urchins residing in dark crevices during daylight hours may not perceive these changes until later in the night, resulting in delayed onset of activity.

Distance travelled by the urchins overnight was lowest for small urchins, compared to medium and large size classes (Fig. 3b, Table 1, ANOVA: F_(2,179) = 7.05, P = 0.001). Small urchins that remained in the FOV all night (n = 24) moved on average 2.5 ± 0.17 m per night (Fig. 3b) and had the lowest average displacement (0.3 m ± 0.05) from their starting point, but this was not significantly different to the displacement of other size classes (ANOVA: F_(157,2) = 2.60, P = 0.08). In comparison, medium urchins (n = 140) moved on average 3.7 m ± 0.17 with a displacement value of 0.4 m ± 0.04 and large urchins (n = 18) 4.6 m ± 0.6 with displacement of 0.6 m ± 0.1. When scaled to urchin size, size did appear to slightly influence distance travelled (Fig 3c, Table 1, F_{(2, 179)} = 3.05, P = 0.05). Post hoc analysis showed small urchins travelled further on average than large urchins (t_{(small–large)} = 2.47, P = 0.04) but there was no difference between small and medium t_{(small–medium)} = 1.42, P = 0.47), or medium and large (t_{(medium–large)} = 1.82, P = 0.21). The distance travelled by urchins overnight has similar implications to their movement rates regarding their predation susceptibility. The more time away from, and further distance travelled from their crevice, the more opportunity there is for nocturnal predation. Since only the urchins that remained within FOV all night were used (50% large urchins, 63% medium and 75% small), there is potential for underestimation of movement data, particularly in the larger size class, which had a greater tendency to leave the FOV. Despite this, it is clear that all sizes of urchins, including smaller ones, leave their cryptic refugia and are vulnerable to predation overnight during periods of greater movement.

Temporal overlap of predators and prey

Examining the nocturnal movement of urchins provides insight into their susceptibility to predation during night-time hours, particularly of smaller, diurnally cryptic urchins. In a similar urchin species, Centrostephanus coronatus, nocturnal foraging behaviour is considered a predator-avoidance strategy which evolved in response to daytime predators (Nelson and Vance 1979). Similarly, the tendency of small C. rodgersii urchins to delay onset of movement later than other size classes is likely to be an evolutionary behavioural adaptation to avoid predation during daylight hours. By delaying their emergence, the smaller urchins potentially reduce the risk of being targeted by predators as the larger urchins are fed on first. Alternatively, the small urchins may benefit from the increased number of conspecifics emerging first, reducing their likelihood of predation by a ‘group protection’ effect.

The observed nocturnal increase in movement among small urchins could put them at higher risk of predation in Tasmania. The nocturnal emergence of smaller urchins directly overlaps with the heightened activity levels seen in lobsters overnight, particularly the activity of small and medium lobsters (Frusher et al. 2009). Laboratory studies have shown small lobsters do predate on small urchins (Ling et al. 2009; Smith et al. 2022), whereas field studies show high prevalence of urchin in small wild lobster diet (Smith et al. 2023). Based on this evidence and the results obtained about urchin temporal behaviour in this study, we conclude that small urchins are temporarily susceptible to predation by small lobsters (as well as other size classes) when highly active on open reef spaces throughout the hours of darkness. This has important implications for management of C. rodgersii with predators, and may switch the focus from more large predators to more predators of all sizes.