Using biomimetic loggers to measure interspecific and microhabitat variation in body temperatures of rocky intertidal invertebrates
Post on 08-Apr-2017
Using biomimetic loggers to measure interspecificand microhabitat variation in body temperaturesof rocky intertidal invertebrates
Justin A. LathleanA,C, David J. AyreA, Ross A. ColemanB
and Todd E. MinchintonA
AInstitute for Conservation Biology and Environmental Management & School of Biological
Sciences, University of Wollongong, NSW 2522, Australia.BCentre for Research on Ecological Impacts of Coastal Cities, School of Biological Sciences,
The University of Sydney, NSW 2006, Australia.CCorresponding author. Email: firstname.lastname@example.org
Abstract. Until recently, marine scientists have relied heavily on satellite sea surface temperatures and terrestrialweather stations as indicators of theway inwhich the thermal environment, and hence the body temperatures of organisms,vary over spatial and temporal scales. We designed biomimetic temperature loggers for three species of rocky intertidalinvertebrates to determine whether mimic body temperatures differ from the external environment and among species and
microhabitats. For all three species, microhabitat temperatures were considerably higher than the body temperatures, withdifferences as great as 11.18C on horizontal rocky substrata. Across microhabitats, daily maximal temperatures of thelimpet Cellana tramoserica were on average 2.1 and 3.18C higher than body temperatures of the whelk Dicathais orbitaand the barnacle Tesseropora rosea respectively. Among-microhabitat variation in each species temperature was equallyas variable as differences among species within microhabitats. Daily maximal body temperatures of barnacles placed onsoutherly facing vertical rock surfaces were on average 2.48C cooler than those on horizontal rock. Likewise, dailymaximal body temperatures of whelks were on average 3.18C cooler within shallow rock pools than on horizontal rock.Our results provide new evidence that unique thermal properties and microhabitat preferences may be importantdeterminants of species capacity to cope with climate change.
Additional keywords: Australia, barnacle, Cellana tramoserica, climate change, Dicathais orbita, habitat temperature,limpet, Tesseropora rosea, whelk.
Received 1 November 2013, accepted 13 March 2014, published online 26 November 2014
In a warming world, interspecific differences in body tem-peratures and thermal tolerances will produce a variety ofresponses among key species within a community. Rockyintertidal shores and their associated biological communities
have emerged as excellent study systems to investigate howclimate change will affect marine organisms (Helmuth et al.2006b). As for many other biological communities, the persis-
tence of rocky intertidal organisms will be dependent on theirability to adapt to the changing conditions, either physiologi-cally or behaviourally (by seeking out more benign micro-
habitats) (Helmuth et al. 2006a; Tomanek 2008; Somero 2010).Because many intertidal organisms live at or close to theirthermal limits (Denny and Harley 2006; Somero 2010), suchphysiological responses may be limited. Rocky intertidal shores
are characterised by high degrees of topographic complexityproducing a variety of microhabitats, including those that mayfunction as thermal refugia for organisms sheltering from
extreme heat stress (Denny et al. 2011; Lathlean et al. 2012).
Our ability to assess how these microhabitats mitigate the neg-
ative effects of global warming for rocky intertidal communitiesis primarily dependent on characterising interspecific variationin body temperatures within and outside thermal refugia.
Due to the dynamic nature of the marine environment,
measuring body temperatures of marine organisms has provensomewhat difficult, and ambient temperature measurementstaken by simple waterproof sensors may not reflect the actual
body temperatures of an organism (Fitzhenry et al. 2004).Technological advances within the past decade, however, haveallowed marine ecologists to use biomimetic loggers to record
the temperatures experienced by target organisms over extendedperiods of time (Pincebourde et al. 2008; Broitman et al. 2009;Lima and Wethey 2009; Szathmary et al. 2009). These biomi-metic loggers are capable of estimating interspecific variation in
body temperatures because they incorporate the unique mor-phological characteristics (i.e. size, colour, shape and composi-tion) of each species. Biomimetic loggers have been used
successfully tomeasure broad-scale body temperatures of single
Marine and Freshwater Research, 2015, 66, 8694
Journal compilation CSIRO 2015 www.publish.csiro.au/journals/mfr
species (Helmuth 2002; Helmuth et al. 2006a; Seabra et al.2011) but we know of only one study that has made interspecific
comparisons using biomimetic technology (Broitman et al.2009) and one other that uses infrared thermography (Cox andSmith 2011). These studies revealed that body temperatures
of different species occupying the samemicrohabitat could varyat times by 5 to 108C, demonstrating the need for additionalinterspecific comparisons in order to predict how spatial het-
erogeneity of the thermal environment will impact on intertidalcommunities. Therefore, by developing biomimetic loggers forseveral co-occurring species, ecologists can investigate the roletemperature plays in regulating interspecific interactions.
On rocky intertidal shores of south-eastAustralia, interactionsbetween several dominant benthic invertebrates have previouslybeen shown to strongly influence community structure (Denley
and Underwood 1979; Creese 1982; Jernakoff 1983; Underwoodet al. 1983; Fairweather 1988a, 1988b). Key species include thebarnacle Tesseropora rosea, the limpet Cellana tramoserica and
the predatory whelkDicathais orbita. Here, adult T. rosea have anegative effect on C. tramoserica by reducing grazing activitiesand growth rates (Underwood et al. 1983). In turn, grazingactivities ofC. tramoserica reduce the survival of recently settled
T. rosea larvae. Processes affecting the relative abundance ofT. rosea are particularly important since: (1) T. rosea is thepreferred prey of several predatory whelks, including D. orbita
(Fairweather 1988a, 1988b); (2) adult T. rosea provide importantmicrohabitats for several small gastropods (Creese 1982;Chapman 1994); and (3) the presence of T. rosea promotes the
settlement and growth ofmacroalgae and conspecifics (Jernakoff1983; Lathlean et al. 2012, 2013). Understanding how bodytemperatures vary among each of these interacting specieswithin
various microhabitats will improve our ability to predict futurespecies interactions and community-level responses to climatechange (Kordas et al. 2011).
The aim of this study was to compare the body temperatures
of three interacting species of intertidal invertebrates acrossseveral common microhabitats on a single rocky shore. We didthis by designing and deploying biomimetic temperature loggers
that mimic the habitat-forming barnacle T. rosea, the predatorywhelk D. orbita, and the intertidal limpet C. tramoserica. Ourfirst objective was to test whether the unique morphological
characteristics of each species influenced the body temperaturesof individuals exposed to the same environmental conditions.Our second objective was to investigate whether the bodytemperatures of invertebrates are lower when they occupy
certain microhabitats thus allowing those habitats to functionas thermal refugia.
Study location and species
The study was undertaken on an exposed rocky shore at GarieBeach (34810038.100S, 151803057.800E) near Sydney in south-eastern Australia. The rocky platform at Garie Beach is pri-
marily composed of siltstone and is grey in colour. The platformhas an east to south-easterly aspect and an overall slight tomoderate (0 to 208) inclination. The topographic landscape ofthe rocky platform is moderately complex, producing a variety
ofmicrohabitats including,most notably, vertical and horizontal
surfaces, rock pools and crevices. T. rosea and C. tramosericaare typically found on horizontal to vertical emergent rocky
substrata exposed to full sunlight within the mid shore region onexposed rocky shores (Denley and Underwood 1979; Under-wood et al. 1983; Hidas et al. 2013).D. orbita is, however, more
abundant within crevices and shallow rock pools within both themid- and low-shore regions and is typically not found on verticalsurfaces (Phillips and Campbell 1974; Fairweather 1988a,
1988b). Rocky shores within this region experiencemixed semi-diurnal tides with a daily tidal range of 1.5 to 2m.
Biomimetic logger design and deployment
Biomimetic loggers for each of the three species were designedfollowing Lima and Wethey (2009). For each logger, thisinvolved dissecting a DS1922L iButton and removing the
internal circuit board and lithium battery. Two exposed con-stantan wires penetrating the shells of the limpet and whelk, andthe test of the barnacle, were used as contacts for logger pro-
gramming and subsequent data retrieval (Fig. 1). These wereconnected to the logger circuit board by soldering two pieces ofwirewrap wire to either the negative or IO terminal. Once con-nected each logger was coated in a waterproof resin (3M
Scotchcast 2130 Flame Retardant Compound) and placed insidean empty test or shell where extra resin was used to fill anyinterstitial space. Temperatures recorded by these biomimetic
loggers will therefore be strongly influenced by the thermalproperties of this resin. Lima and Wethey (2009) demonstratedthat 3M Scotchcast 2130 effectively mimicks the thermal
properties of the limpet Tectura persona. Therefore, the resinlikely acts as an effective surrogate for the internal tissue ofmany marine invertebrates.
Since intraspecific variation in body size could lead todifferent heating rates and body temperatures, we selectedgastropod shells and barnacle tests of roughly equivalent volumeand filled these with approximately the same amount of resin.
Particular care was takenwhen constructing the barnacle loggersto ensure that the operculum remained intact and in the correctposition. Once resin had set, loggers were calibrated in the
laboratory and deployed into the field usingA-788Z-Spar SplashCompound (Koppers Co., USA) (Fig. 1). Calibration involvedplacing all loggers under identical thermal conditions in a
temperature controlled room. Differences among all loggersunder these conditions were typical between 0.1 and 0.28C.
Biomimetic loggers for each species were deployed acrosscommon and abundant microhabitats within the midshore
region at Garie Beach between 5 March and 11 April 2013and body temperatures were recorded every 10min at a resolu-tion of 0.06258C. Four biomimetic loggers for C. tramosericaand four T. rosea loggers were attached to either horizontal orsouthward facing vertical emergent rocky substrata (n 2loggers per microhabitat), while four D. orbita loggers were
either submerged in rock pools ,0.1m deep or attached toemergent horizontal rocky substrata (n 2 loggers per micro-habitat). These microhabitats encompass the most thermally
distinct regions within the mid-intertidal region at Garie Beachfor each species (hereafter referred to either benign or stressfulmicrohabitats). The two microhabitats selected for D. orbitawere also intended to reflect the body temperatures of aerially
exposed animals foraging among barnacles and non-foraging
Interspecific and microhabitat variation in body temperature Marine and Freshwater Research 87
animals submerged in shallow rock pools. Two unmodifiedTidbiT v2 Temp data loggers (Onset Stowaway logger,
model UTBI-001, accuracy 0.28C) with a sampling freq-uency of 10min were also attached to horizontal and southerly
facing vertical rocky substrata (n 2 loggers per microhabitat),0.5m to 1m away from biomimetic loggers to comparetemperatures of the microhabitat with that of the three studyspecies. Later comparisons between these unmodified Tidbitsand iButton loggers embedded in nearby (.1m) rocky substratarevealed a strong relationship between the two temperaturemeasurements (R2 0.883, P, 0.001, n 1315). Therefore,temperatures measured by the Tidbit loggers represent an
appropriate estimate of habitat temperature.The accuracy of biomimetic loggers was assessed by
comparing biomimetic body temperatures with that of three orfour similarly sized live specimens on horizontal rock for three
hours during three non-consecutive low tides. Thiswas achievedby inserting a small thermocouple (HANNA HI 935005K-Thermocouple Thermometer, resolution 0.18C) through a2-mm hole drilled into the shell or test of the live specimens.Similarly, to biomimetic loggers, the thermocouple was cali-brated in the laboratory before use and produced accuracies
between 0.1 and 0.28C. Live specimens were located between5 and 10 cm from biomimetic loggers and were all orientated inthe same direction to ensure small-scale thermal heterogeneitydid not affect our comparisons. We deliberately undertook our
comparisons during daytime low tides as these conditionsproduce the greatest absolute temperature differences and ther-mal stress as opposed to night time or during high tide, when
body temperatures equilibrate with sea temperatures. Bodytemperatures of live limpets were highly correlated with biomi-metic temperatures (Fig. 2a), with differences being on average
0.18C (0.1 s.e.) and no greater than 1.28C. Likewise, bodytemperatures of live barnacleswere significantly correlatedwithbiomimetic temperatures (Fig. 2b), with differences being on
average 0.88C (0.1 s.e.) and no more than 1.58C. Whenmeasuring body temperatures of live whelks in the field,individuals would retreat into their shells after being disturbedand would not reattach to the substrata for several hours. This
behaviour precluded appropriate comparisons between livewhelks and biomimetic loggers. Nevertheless, body tempera-tures of live whelks were highly correlated with biomimetic
temperatures (Fig. 2c) with differences being on average 2.08C(0.3 s.e.) and a maximal difference of 4.48C. Such largediscrepancies between body temperatures of live whelks and
biomimetic loggers were most likely the result of the observedretraction into shell. Consequently, comparisons of biomimeticdata with in situ body temperatures of live whelks should beviewed with caution.
Temperatures recorded by paired biomimetic or TidbiT loggers
within each microhabitat were averaged to produce a singlevalue for each 10min interval. Two-factor analysis of variance(ANOVA) was then used to test for statistically significant
differences in body temperatures (mean daily maxima, mean,and mean daily minima) among species (C. tramoserica,T. rosea and D. orbita) between the different microhabitats
(benign and stressful). One-factor ANOVA was used to test forstatistically significant differences in microhabitat temperatures(mean daily maxima, mean, and mean daily minima) recordedby TidbiT loggers deploy...