During the breeding season, N. ostrina egg capsules were collected from seven mid- to high-intertidal sites between the months of May and July 2006 along the west coast of the United States across ~13° of latitude from Washington State to Central California. The collection locations are as follows: Cattle Point, San Juan Island, WA (CP); Boiler Bay, OR (BB); Strawberry Hill, OR (SH); Coos Bay, OR (CB); Humboldt, CA (HUM); Bodega Cove, CA (BML); Cambria, CA (RM) (see Table 1). At one collection site (Rancho Marino Reserve in Cambria, CA), N. ostrina overlaps in distribution with N. emarginata near Point Conception (Marko 1998). However, for this study, since genotyping was not conducted on the veligers, egg capsules collected from this site were called N. ostrina resembling the ‘northern form’ described by Palmer (1990) and later classified as N. ostrina by Marko (1998). At all sites, 15–30 egg capsules were haphazardly collected from 10–25 clutches by scraping egg capsules off the substrate and placing them in a mesh holding container. After collection, all egg capsules were immediately transported back to a marine facility near the site. Upon arrival, the egg capsules were placed in a flow-through seawater tank containing seawater adjusted to the temperature of the location where the capsules were collected (10–19°C). On average, experiments were run about 1–3 h after collection, and no collection of egg capsules was held for more than 12 h before experimentation began. Table 1

Collecting sites for Nucella ostrina

Using a protocol modified after Sewell and Young (1999), the egg capsules were exposed to a range of temperatures (from 13 to 34.5°C, see Table 2), including some of the higher temperatures that could be experienced in the intertidal zone during emersion (Zippay 2009). While some of the temperatures were outside the range of ecological relevancy, it was important for this study to identify the physiological limits imposed by temperature on survival for developing veligers. The number of temperature treatments varied slightly from site to site due to the number of egg capsules available for collection. For the actual assay, five to eight egg capsules were placed in an unsealed 20 mL glass scintillation vial containing filtered seawater (0.35 μm), in triplicate at each of the different temperatures (Table 2). To eliminate the confounding effect of desiccation stress as a variable, egg capsules were exposed to the temperature treatments while immersed in seawater. After the 1-h incubation, egg capsules (n = 5–8) were cut open, and larvae were examined under a dissecting microscope and were scored for survivorship. Larvae that appeared to have ciliary movement around their velum were considered alive and those with no movement were scored as dead. Each capsule contained approximately 5–22 larvae. Table 2

Temperature incubation regimes for Nucella ostrina egg capsules for a particular site

Methodologically, temperature exposures of 1–2 h in duration are commonly used to analyze the physiological performance of larvae (Boon-Niermeijer and Van de Scheur 1984; Sewell and Young 1999; Brown et al. 2004). Although it is likely that Nucella egg capsules might experience more prolonged exposures to thermal stress in the field, we took a conservative approach in this study and chose a 1 h period for the assessment of survivorship. It is important to note that only advanced veliger-stage larvae were used for this experiment (as described by Gallardo 1979). Namely, if the larvae did not have distinct velar lobes, appearance of shell, eyespots, formation of foot, and operculum, then the whole egg capsule was eliminated from the experiment. These thermal tolerance trials were designed to measure veliger mortality across different temperatures, and these data were used to calculate the LT₅₀ for that particular site. LT₅₀ refers to the temperature at which 50% of the total experimental population was killed by the heat treatment (Stillman and Somero 2000; Hamdoun et al. 2003).

To assess if the observed LT₅₀ was a true indication of larval mortality as opposed to a temporary state of torpor (i.e., a temporary state of decreased physiological activity) induced by the acute heat stress, a second experiment was conducted on N. ostrina larvae from Strawberry Hill, OR. Thirty-five egg capsules were collected from the field and subjected to a lethal (34.5°C) and an acute (31.7°C) heat shock for 1 h and allowed to recover for up to 1 h at the ambient seawater temperature for this population (11°C). N. ostrina veligers were exposed to 31.7 and 34.5°C because these temperatures represent the LT₅₀ and lethal temperature, respectively, for the Strawberry Hill, OR site.

Using the same thermal tolerance protocol as above, egg capsules (n = 35) were exposed to a 1 h heat shock, in triplicate, at the two temperatures (LT₅₀ = 31.7°C and lethal = 34.5°C). Immediately following the exposure, seven egg capsules were dissected and larvae survival rates were determined. The remaining 28 egg capsules for each temperature were placed in the seawater table to recover. Every 15 min up to 1 h, seven egg capsules were removed from the recovery seawater table and survival rates were determined. To obtain a baseline measurement of larval condition in the egg capsules with no heat stress (“control”), survival rates of larvae from 21 N. ostrina egg capsules were quantified prior to any high temperature exposures.

For the thermal tolerance assays, analyses were performed using JMP 7.0 statistical software (SAS Institute Inc.). The LT₅₀ of Nucella for each site was calculated using a 95% confidence interval where upper and lower limits of the 50% mortality were given. Analyses for the recovery experiment were performed using SigmaStat 3.0 statistical software. The average mortality was calculated for each recovery time point for veligers that were heat shocked at the LT₅₀ temperature of 31.7°C and those that were heat shocked at 34.5°C. All values were Box-Cox transformed to meet the assumptions of normality. The mean mortality values for each heat shock temperature were compared using one-way analyses of variance.

Veliger larvae from all sites survived temperatures from 13 to 27.5°C (Fig. 1). As temperatures further increased, survivorship rapidly decreased within a narrow window of temperatures. Overall, regardless of the biogeographic collection site of the Nucella ostrina veligers tested in the thermal tolerance trials, the temperature at which death was first observed was separated by only ~3–5°C from the temperature that caused nearly 100% mortality. For example, veligers from Cattle Point first experienced mortality at 29.5°C with 47.5% dead; this population experienced 100% mortality at 33.5°C, a range of only 4°C (Fig. 1). A similar pattern was seen for other sites, such as Strawberry Hill and Humboldt, where only a few degrees (5 and 3°C, respectively) separate the initial onset of mortality and 100% mortality (Fig. 1).Fig. 1

The thermal tolerance profile for Nucella ostrina veligers from four out of the seven sites during May–July 2006. The x-axis corresponds to the exposure temperatures of egg capsules after 1-h incubation. The y-axis is an index of survivorship based upon veligers with ciliary movement around the velum after exposure. Each line represents the number alive compared to the overall total number of larvae used (n = 5–8 egg capsules, n = 5–22 embryos per egg capsule)

In the thermal tolerance trials, there was a relationship between larval thermal tolerance and the latitude at which the Nucellaostrina egg capsules were collected. Specifically, the northern populations of veligers were more sensitive to higher temperatures when compared to the veligers collected in the southern part of the range (Fig. 3). N. ostrina veligers collected from the most northerly site of Cattle Point had an LT₅₀ of 30.1, 3.8°C lower than the LT₅₀ for N. ostrina veligers from the most southern site of Rancho Marino (LT₅₀ = 33.9°C, Fig. 3 and Table 2). When considered as a group, the veliger LT₅₀’s showed a pattern of least to most tolerance of high temperatures in the following order: CP < BB > SH ≈ CB < HUM < BML ≈ RM (Table 2). This latitudinal trend was confirmed in a linear regression model (Fig. 3, P = 0. 0011, R² = 0.8995).Fig. 3

Latitudinal comparison of veliger thermal tolerance for Nucella ostrina, distribution indicated by solid black line, during May–July 2006. Black dots, on the graph, represent the calculated LT₅₀ for each site with error bars representing ±95% confidence intervals. Y-axis represents latitude and correlates with the collection sites depicted by black dots on the map. The x-axis is the temperature at which 50% of the veligers died after a 1-h exposure (P = 0.0011, R ² = 0.8995)

In the thermal trials, there was a rapid, linear increase in mortality occurring over a few degrees of temperature. Specifically, there was only ~3–5°C difference between the temperature at which we first began to see mortality and the temperature where nearly 100% mortality was achieved across all the sites. Given the large-scale distribution of Nucellaostrina, a broader range of LT₅₀ temperatures might be predicted. However, our data revealed a narrow window of thermal tolerance. For example, we observed a very small gap between survival and mortality, where the LT₅₀ temperature for N. ostrina veligers collected from Strawberry Hill was 31.7°C when compared to the recorded lethal temperature of 34.5°C. These slight differences in thermal tolerance indicate that there is an abrupt threshold for mortality that could have profound affects on survival where “ecological surprises” of sudden increases in temperature could precipitate mass mortality events (Harley et al. 2006).

While our study is the first to assess thermal tolerance, our observation of the narrow window between tolerance and mortality is consistent with other abiotic stressor studies of Nucella that also found broad range of tolerance followed by a sudden decline in survivorship of encapsulated veligers. Although the nature of the egg capsule can mitigate some of the abiotic stresses inherent to life in the intertidal (desiccation stress, osmotic shock, and UV radiation) (Pechenik 1982, 1983; Hawkins and Hutchinson 1988; Rawlings 1990; Russell and Phillips 2009a, b), there are thresholds where the protective capacity of the capsule fails. For example, in a study on desiccation stress, late stage veligers (in the formally N. emarginata, now referred to as N. ostrina, Marko 1998) could withstand up to 80% water loss from the capsule chamber before suffering substantial mortality (Rawlings 1995). In other studies, it has been shown that capsule wall thickness provides some protection against UV radiation. In a comparative study, the thicker-walled capsules of N. emarginata (now referred to as N. ostrina) were more resistant to UV stress and protected from UV-A radiation when compared to the thinner-walled lower intertidal species, N. lamellosa (Rawlings 1996, Pechenik 1999).

Our finding of an acute increase in N. ostrina mortality at elevated temperatures is not isolated within larval biology; similar observations of sudden shifts in survivorship in response to thermal stress have been made in early life history stages of other marine invertebrates. Boon-Niermeijer and Van De Scheur (1984) found that a 1°C increase (39.5–40.5°C) in heat exposure for 3-day-old Lymnaea stagnalis, a gastropod mollusk, shifted mortality from 40 to 100%. Similarly, in a non-marine invertebrate, larvae of two species of Drosophila, D. melanogaster and D. simulans, had percent survivorship that shifted from 100% to near zero when exposure temperatures were increased from 37 to 39.5°C (Krebs 1999). Responses of developing embryos to temperature, and the particular temperatures encountered during the normal breeding season, have been most extensively investigated in echinoderms (Andronikov 1975; Strathmann 1987; Fujisawa and Shigei 1990). Purple sea urchin (Strongylocentrotus purpuratus) embryos displayed abrupt changes in survival over just a 2°C span for both gastrulae and 4-arm larvae (L. M. Hammond, unpublished results). In an early study on five tropical species of sea urchins, Rupp (1973) found that the upper thermal limits of cleavage were near 34°C, whereas egg fertilization was more thermally resistant and not affected until 36°C, suggesting that larval development would cease if larvae were trapped in flat reef pools or lagoons where temperatures exceed 36°C. Similarly, Sewell and Young (1999) reported that the ability for blastulae and later staged tropical sea urchin larvae (Echinometra lucunter) to survive short periods (2 h) of elevated temperatures (4–5°C higher than the 34°C upper limit for normal development) might be important if the larvae are carried to shallow waters of the Caribbean reef flats where seawater temperatures can reach 40°C. In another experiment, significant increases in mortality were observed in the larvae of a brooding tropical coral, Porites astreoides, when exposed to slight increases in temperature from 28 to 33°C (Edmunds et al. 2001). Finally, in a polar species, successful larval development was achieved in a 3°C window, from +6 to +9°C, for the sub-Antarctic crab, Paralomis granulosa (Anger et al. 2003).

While all the above examples help to explain the importance of understanding how temperature limitations influence early developmental stages of larvae, they also highlight the scarcity of data on the temperature impacts on non-planktonic larvae. From an ecological perspective, this study contributes to our understanding of how conditions in the intertidal might influence larval survival. In a more global context, these data also suggest that increased air and seawater temperatures due to global climate change may also increase mortality during extreme heat events (Hoegh-Guldberg 1999; Pandolfi et al. 2003; Denny et al. 2009).

Our data suggest a strong latitudinal relationship between calculated LT₅₀ temperature and site. Larvae from more southerly distributed populations were more thermally tolerant than northern populations. For N. ostrina veligers collected from Cattle Point, the LT₅₀ was 30.1°C when compared to an LT₅₀ of 33.9°C of veligers collected at the more southerly location of Rancho Marino Reserve (Table 2). Few other studies have addressed the relationship of biogeography and the physiological tolerances of larval forms. One such study found a modest correlation between thermal tolerance and biogeography for echinopluteus larvae of Strongylocentrotus purpuratus. Larvae spawned from Oregon adults (LT₅₀ = 29.7°C) were more thermally sensitive than those larvae collected from Baja California, Mexico urchins (LT₅₀ = 31°C) (L. M. Hammond, unpublished results).

In contrast to the limited data on larval forms, the correlation of thermal tolerance with latitude and environment has been documented in adult forms of many marine invertebrates (see review Vernberg 1962; Stillman and Somero 2000; Somero 2005; Sorte and Hofmann 2005; Jansen et al. 2007) and in other ectotherms (e.g., Drosophila melanogaster—Hoffmann et al. 2002, mosquitoes—Zani et al. 2005, Fundulus heteroclitus—Fangue et al. 2006). Similar to our results, Urban (1994) found a correlation between LT₅₀ temperature and geographical distribution for adults of 10 bivalve species from Peru and Chile over 22° of latitude.

In summary, the current study showed that Nucellaostrina veligers displayed different thermal tolerances as a function of the latitude at which they were collected. The data presented here are a conservative measurement of the upper limits of physiological thresholds of encapsulated larvae. Overall, these data suggest that larval stages of this intertidal invertebrate may not be a hyper-vulnerable life stage (see also Hamdoun and Epel 2007). This observation is further supported by other studies, for example, on an intertidal bivalve which found that early life stages and adults had similar thermal tolerances. The oyster, Ostreola conchaphila, brood their young and adults and veliger larvae survive the same lethal temperatures of 38.5–39°C (Brown et al. 2004). However, changing thermal patterns due to climate change may significantly increase mortality during larval development and result in population fragmentation based on the latitudinal timing of low tides and tendency for local adaptation (Sanford et al. 2006).