Plasticity and evolution shape the scaling of metabolism and excretion along a geothermal temperature gradient

Read the free Plain Language Summary for this article on Journal blog. Ecological responses to environmental warming depend organisms’ physiological characteristics and extent which these can change in response increasing temperature (Chown et al., 2010). Beneficial be achieved via two interacting mechanisms: adaptive phenotypic plasticity evolutionary adaptation (Ghalambor 2007; Stamp & Hadfield, 2020). Plasticity allow organisms cope with (Crozier Hutchings, 2014; Merilä Hendry, 2014), but if rises beyond range of existing plastic responses, population persistence may potential rapid (Visser, 2008). Moreover, even within thermal window persistence, ongoing tune physiology a manner that alters ecology species interactions (Hendry, 2016; Nadeau Urban, 2019). In way, ecological outcomes rest combination evolved changes key traits (Levis Pfennig, 2020; Price 2003). Physiological processes are heavily influenced by body size temperature. Of shape organism performance function, metabolism is one strongest predictors dynamics, community composition ecosystem as climates warm (Bruno 2015). Metabolic rate dictates energy demand, influencing food intake top-down (Carr Bruno, 2013; Jochum 2012). Thus, dependence associated fundamental parameters many models attempting predict outcome 2015; Schramski Sentis 2017). Most based assumptions metabolic theory (MTE), predicts B scales exponentially hypo-allometrically size, according = b 0 M α e - E / k T , where b0 normalisation constant, mass, allometric scaling exponent (i.e. dependence), activation sensitivity or Boltzmann's constant (Brown 2004; Gillooly 2001). The MTE assumes independently temperature, problematic, it disregards interactive influence scaling, well evolution shaping (Lindmark 2018; Moffett 2018). Counter predictions, effects not independent (Glazier, 2005; Killen 2010; Lindmark For example, positive relationships between have been observed mosquitofish (Moffett 2018), black porgy (Li 1999) some intertidal gastropod (Huebner, 1973; Newell, 1973), among others 2005). cases, warmer temperatures closer one, meaning rises, per unit mass becomes more similar across opposite pattern has also found several aquatic invertebrate fish Ohlberger studies rates examined at interspecific level. slope (α) evolve species, commonly decreasing different were assayed Rubalcaba White Although intraspecific variability difficult conclusively determine whether results from and/or level (Fossen 2019; Glazier, Nespolo, 2003; Réveillon Populations environments display below expectations, suggesting reduce (‘countergradient variation’, ‘metabolic cold hypothesis’; Pilakouta 2012; although see Alton Studies recently demonstrated reduction arise over few generations (<100 years), assessed relative contribution versus patterns Along metabolism, excretion nutrients waste process consumers processes. particular, availability base webs, ecosystems bottom-up nutrient subsidy (Vanni, 2002). demands drive acquisition processing; thus, reasonable expect scale similarly However, tends variable than factors such stoichiometry, starvation period rate's (Hall Hirst 2017; Vanni McIntyre, 2016). Given biomass abundance balance recycling (O’Gorman 2012), question how relationship affected either exacerbate mitigate effect warming. We disentangled model organism, western Gambusia affinis. six established (~100 years) populations geothermal springs span broad gradient (19–33°C). Prior field surveys shown sizes decline slopes rise low high Furthermore, common-rearing experiments revealed somatic growth countergradient fashion reduced adult (Fryxell influences tested, interdependent (Kerkhoff, so reflect metabolism. On other hand, show growth, slower growing tend excrete faster because they retain less nitrogen protein synthesis (Bucking, common reared each four rearing (23, 26, 30 32°C) test rates. If wild result primarily alone, we expected all populations. involved, degree differing populations, consistent previously patterning described above affinis small (<6 cm), livebearing native southern United States northern Mexico. It was introduced diverse worldwide mosquito biocontrol throughout 20th century, creating natural experiment genotypic 2021). Mosquitofish generalists, capable reproduction water ranging 16 35°C, tolerant short-term exposures 45°C (Pyke, 2005, 2008; Wood Here, studied invaded California. 1922, taken 1–2 localities Texas, stocked into single pond California, systematically spread state through series transplantations (Dill Cordone, 1997). Our focal occur thermally stable Inyo Mono Counties, 19 33°C. These experience relatively around year (Figure 1) human-made barriers dispersal habitats addition, chemistry characteristics, geographically close (maximum distance sites ~70 km, Figure do contain piscivorous fishes, making them an ideal system study without typical confounding along gradients Table S1). collected wild, transported controlled environment facilities, then performed generations. Fish collections approved our institutional animal ethics committee (UCSC protocols PALKE-1311 PALKE-1801) local wildlife agency (CADFW permit SC-12752). Collection detail Fryxell al. (2020). Briefly, using hand nets February 2018. F0 (wild-caught generation) environment-controlled greenhouse UC Santa Cruz, separately 568 L tanks heated intermediate 26°C separate tanks). After 1 month, began collecting offspring F1 generation fish). Groups newborn same day birth together mesh baskets hung 57 room 26°C. When had least 90 population, euthanised, after drained, cleaned refilled, again tanks. grown maturity, their (F2 months. fed flake pellets 2× daily rearing. transferred F2 rooms (R.W. Smith Co) treatment 32°C), spanned most experienced assigned unique ID number individually inside cylinders petri dish bottom open-top (250µm mesh, 7cm diameter, 20cm height) no contact during Cylinders sunk upright 100 tubs (91 × 61 20 cm) filled off-gassed city water. Tubs allocated rooms. room, set air 23°C 23, 32°C. 19°C randomly temperatures, five replicate tubs. given sequentially born equalise sample population. minimise tub differences density time. densities various dates. excess Frystartr (Skretting, Stavanger, Norway) three times daily. Metabolism assays being 12 118 days, needed measure coefficients. A depending (Table S2). measured routine (RMR) oxygen consumption (MO2) individual (Sinclair 2006), experimental protocol used Routine comparable ecologically relevant basal under normal activity, including spontaneous movements (Clarke, Hudson 2013). closed-system respirometers consisted 40 ml rectangular acrylic chambers fitted optical sensor dissolved concentration (DO) Sensors monitored FireSting four-channel logger (PyroScience, Aachen, Germany), allowing us run simultaneously. Nitrogen estimated ammonia (NH4+–N) closed assays, (2018). 5 weeks September 11 October Individual reared. Each week, conducted day. order randomised week. Before measuring day, selected 20–30 attempted origin arose due reproductive Food digestion increase MO2 (Chabot time vary 2005); starved hr before assays. To so, held 475 cups placed (69 47 22 ensure guts empty Measurements 95 ‘assay’ bin dechlorinated appropriate aerated overnight aquarium bubbler. Assay changed start assay bin, added respirometer, sealed keep constant. allowed settle minutes monitoring DO 10–15 min, duration adjusted magnitude large enough mix fin movements, linear declines (r2 > 0.9; Clark case juveniles, necessary end run, difference initial final divided chamber instead (59% measurements). ran blanks beginning averaged subtracted value MO2. calculated μmol O2/min. measurement, took filtered 0.7 μm glass microfiber filter (GF/F) stored ice 15 tube. euthanised obtain blotted wet weight measurements week NH4+ fluorimetry (Holmes Trilogy Laboratory Fluorimeter (Turner Designs). calculate rates, total spent chamber. concentrations increased linearly runs blank regression first last highly fluctuating traces excluded analyses; 415 analysed RMR, 452 rate. First, RMR individuals, accounting (evolved differences) (phenotypic plasticity), calculating simple ln-transformed data. Second, fitting OLS models. started full specification: ln(rate) ~ ln(mass) + RearingTemp OriginTemp OriginTemp. removing terms comparing resulting log-Likelihood ratios tests (χ2). significant interaction term would indicate broadly shared common-reared populations). An suggest α. Finally, support themselves mediated heritable (evolution plasticity; 1). work 2018) (as energy, E) mosquitofish. explored estimating Arrhenius (expressed 1/kT; Brown 2004) (minimum, mean maximum values datasets). predicted coefficients allometry temperatures. pooled affects Slope All statistical analyses R v3.6.2 (R Core Team, R’s lm() function. selection process, (χ2) lrtest() function lmtest v0.9-37 package (Zeileis Hothorn, Final verified plotting residuals covariate model, whereas variance inflation factor (VIF) assess collinearity covariates vif() car v3.0-8 (Fox Weisberg, VIF ln(body mass), lower cut-off recommended collinear variables (Gareth summary() extract models, predict() log-transformed ±95% confidence intervals masses, specific extracted emtrends() emmeans v1.4.7 (Lenth, Across combined, (α 0.76 ± 0.06) predictions (R2 0.58, p < 0.001). showed its 2). Instead, best included interaction. (p 0.049), lack rate, indicated variation (Tables S3 S4). estimate (E) RMR. rose 0.60 S5; 3a). Consequently, differed rising 0.17 0.68 eV smallest largest dataset S6; 3b). steeper 0.82 0.06; R2 0.60, 0.001) 0.06). contrast three-way indicating (log-likelihood −323.80; χ2 6.58; 0.010). Allometry S7), reflecting 4). Notably, reversed S8; 4a). coldest decreased warmest exhibiting declined S9; 4b). Excretion scaled (slope 0.74 0.08, 0.47, 0.001; 5). despite S10; disentangle Previously, juvenile greater investment contributing At time, plasticity, genetic evolution. contrast, evidence Despite both individuals treatments, followed expectations 0.75 when account, what words, increased, mass. This reflects interdependently, contrary negative (or standard) multiple activity movement; non-sessile organisms, better energetic nature (Hudson species’ ontogeny (Gaitán-Espitia 2006; Norin Gamperl, High metabolism's slow shielding genes (Price Phylogenetic comparative correlated (White Therefore, correlation traits, likely constrain capacity independently, constraining time-scales (Futuyma, did directly affect operate indirectly connection size. Per (2020), there good component early Hence, phenotype shaped indirect (plastic) 6). Evolution towards smaller help constraints imposed higher reducing whole-organism demands. Additionally, probably Riemer could facilitate further, equalises size-specific potentially eliminating advantage efficiency larger note age plays tended younger. possible young allometry. Gamperl (2018) decreases age. Ultimately, inextricably linked disentangle. Contrary fish, 0.06), great 0.58–0.98). Field-tested exhibited springs, spring driven diet 2018, 2021; nonlinear excretion's decrease increases. warming, exploring normally acclimated short periods, ignoring (Morgan 2001; Wright Nutrient allocate resources structural vital activities (Hirst portion fishes’ fuelled proteins (Wright As tolerance limit, proportion supported (Alsop 1999; Consistently, colder except enhanced fish. Large (Wood protein-fuelled abruptly tolerances involves mechanisms besides exhibit relationship, slow-growing consumed amino acids (Bonaldo 2011; Bucking, sizes, might Evolutionary alter mosquitofish's consumers’ coupled fashion, able maintain current turnover primary producers communities proportionally 0.08), field-tested 0.85; similarity suggests (Carter Brafield, 1992; Uliano processes, like metabolising glycogen lipids reserves subjected periods (Ferreira 2001); methodological limitations, form urea, account 10%–30% teleost nitrogenous (Uliano Plastic will climate change. Results previous populations’ face 6; highlight importance respond differently demand Overall, findings caused contemporary traits. Combining parse improving understanding thank Taylor Vance, Jim Velzy, Daniel Alvarez Finn Arnesen assistance research affiliated institutions: US National Science Foundation (GRF D.C.F., DEB-1457333 E.P.P., DEB-1457112, OIA-1849227, OIA-1826777 M.T.K.), Royal Society New Zealand Marsden Fund (16-UOA-023 K.S.S., E.P.P. M.T.K). E.P.P received NOAA Cooperative Institute Marine, Earth, Atmospheric Systems M.T.K Maine Agricultural Forest Experiment Station. Open access publishing facilitated Monash University, part Wiley University agreement Council Australian Librarians. [Correction 11-May-22, online publication: CAUL funding statement added.] declare conflict interest publication. authors contributed design; D.C.F. led (with J.N.B.); J.N.B. K.S.S.), analyses, writing manuscript. critically drafts gave approval Data available Dryad Digital Repository https://doi.org/10.5061/dryad.2280gb5ts (Benavente 2022). Please note: publisher responsible content functionality any supporting information supplied authors. Any queries (other missing content) should directed corresponding author article.