Elemental C and N balances evidence stoichiometric adjustments to dietary protein content in growth phenotypes of the Manila clam (Ruditapes philippinarum)

Aquaculture 555 (2022) 738147 Available online 24 March 2022 0044-8486/© 2022 The Authors. Published by Elsevier B.V. This is an open access article under the CC BY-NC-ND license (http://creativecommons.org/licenses/bync-nd/4.0/). Elemental C and N balances evidence stoichiometric adjustments to dietary protein content in growth phenotypes of the Manila clam (Ruditapes philippinarum) Kristina Arranz * , I˜ naki Urrutxurtu, Enrique Navarro Departamento de Gen´ etica, Antropología Física y Fisiología Animal, Facultad de Ciencia y Tecnología, Universidad del País Vasco/Euskal Herriko Unibertsitatea, UPV/ EHU, Apartado 644, 48080 Bilbao, Spain ARTICLE INFO Keywords: Nutrient balances Stoichiometry Growth phenotypes Ruditapes philippinarum C:N ratio Homeostasis ABSTRACT Juveniles from two separately bred families of the Manila clam Ruditapes philippinarum were used to constitute fast (F)- and slow (S)-growing groups by size segregation. The growth phenotypes that resulted from these 4 family*growth category combinations were then used to measure elemental C and N balances in the acute and chronic responses to two different microalgal diets that were isocaloric but differed (x3) in their protein:energy ratios. Both diets were based on the same phytoplankton species (Rhodomonas lens) that were harvested in the exponential (low C:N ratio) or stationary (high C:N ratio) phases of an indoor culture. Aims were to assess the differential acquisition and processing of elemental nutrients according to differences in their dietary availability and the variable requirements set by the growth trends exhibited by different phenotypes. Clams fed the high N diet achieved the highest C and N balances through a combination of higher feeding rates and absorption efficiencies of overall organics. However, the main differential effects were associated with increased absorption efficiencies for N relative to C, which were particularly observed in clams subjected to chronic N deficit in the diet. This occurred in all growth phenotypes and resulted in the partial homeostatic regulation of nutrient imbalance operating at the preabsorptive level. Further adjustments occurred at the postabsorptive level with increased N disposal in the form of ammonia excretion and the resulting decrease of metabolic C:N indices recorded in clams fed the high N diets. The main phenotypic differences were observed in the intrafamily (F vs. S growth groups) rather than in the interfamily comparisons, with F clams exceeding the absorption rate of nutrients (both C and N) of S clams by 50%. Physiological responses to the acute dietary change were found to be faster in F clams, indicative of a higher plasticity of this phenotype. Whereas, stoichiometric adjustments by S clams resulted in higher N release through excretion, suggesting less efficient protein turnover. 1. Introduction Growth is an important attribute of animal individuals and populations of marine bivalves and a factor to be accounted for in the management and sustainability of coastal ecosystems, as well as an economic target in the aquaculture industry. From an energetic point of view, growth can be viewed as “the outcome of acquisition and utilization of nutrients and energy” (Bayne, 2017), and this perspective has dominated the field since the early growth models by von Bertalanffy (1938) that considered growth to result from a balance between metabolic processes of assimilation and catabolism. Bioenergetic models currently in use, either the classical scope for growth (SFG) approach or the dynamic energy budget (DEB) theory (Kooijman, 2010), derive from the same principles, although relying on a more detailed physiological account of energy flows. These models have provided useful tools for predicting overall growth that are of general interest in aquaculture production, while the experimental testing of these models has enlarged our knowledge of the mechanisms of growth with regard to the specific influence of both endogenous and exogenous (food, temperature, etc.) factors on the various components of the energy budget. However, the use of energy as the sole currency in these budgets has clear limitations stemming from the fact that large differences are often encountered between consumed food and consumer body tissues with regard to elemental composition. This has special relevance in * Corresponding author. E-mail address: [email protected] (K. Arranz). Contents lists available at ScienceDirect Aquaculture journal homepage: www.elsevier.com/locate/aquaculture https://doi.org/10.1016/j.aquaculture.2022.738147 Received 17 September 2021; Received in revised form 27 January 2022; Accepted 14 March 2022 Aquaculture 555 (2022) 738147 2 herbivorous or detritivores relying on vegetal resources, such as most bivalve molluscs, where for instance,the rather constant C:N ratios reported for body composition (between 4.1 and 6.2) contrasts with the highly variable values of C:N ratios in their diets, ranging between 4.8 and 22.5 on a seasonal/experimental basis (Bayne, 2009; Bayne and Svensson, 2006; Fielding and Davis, 1989; Grant and Cranford, 1991; Smaal and Vonck, 1997; summarized in Bayne, 2017, Table 5.9 on p. 299). In the stoichiometric approach (Sterner and Elser, 2002), these mismatches are addressed in terms of nutrient homeostasis that involves the resource in a suit of physiological mechanisms able to compensate, whenever necessary, for the imbalance between the diet and consumer tissues. Nutrient limitation implies that C or energy is in excess, and “to maintain overall homeostasis, consumers should release elements on food in excess of requirements while retaining most of the limiting element…. On the other hand, consumers may also be limited by carbon (energy), in which case it is the ingested nutrients that are subject to efficient recycling” (Anderson et al., 2005). According to this description, for stoichiometry to be realized in growing tissues, any component of the diet in excess of the appropriate ratio should be disposed of, whence important constraints on growth can be expected that would hardly be identified from only the account of energy flow measurements. This approach thus replaces reliance on energy budgeting by assessing nutrient (elemental) balances based on physiologically determined components of growth. Several early studies emphasized the need for an evaluation of nutrients, together with energy, in the analysis of growth (Cranford, 1995; Grant and Cranford, 1991; Bayne et al., 1993; Hawkins and Bayne, 1985), and modelling efforts have tended to incorporate parameters that take into account the elemental composition of diet to fit a better approach of actual growth (Scholten and Smaal, 1999; Scholten and Smaal, 1998; Smaal and Scholten, 1997; Bourl` es et al., 2009; Brigolin et al., 2009; Emmery et al., 2011; Granger´ e et al., 2009). Mechanisms for homeostatic nutrient regulation in bivalves can be identified at several different levels in the chain of physiological processes connecting feeding to growth. These include preingestive as well as preand postabsorptive levels. The preingestive selection of N-rich particles has been broadly reported in feeding experiments using natural seston or mixtures of phytoplankton with suspended sediments containing organic detritus (Hawkins et al., 1996; Iglesias et al., 1996; Soletchnik et al., 1996; Urrutia et al., 1996), a selective process that would be mainly based on the ability of gills and palps to discriminate between microalgae (low C:N ratio) and detrital particles (high C:N ratio). In fact, higher selection efficiencies (SE) observed for chlorophyll relative to overall organics were found to correlate with higher SE for N (SE N ) compared with SE for C (SE C ) in some of these studies (e.g., Iglesias et al., 1996; Urrutia et al., 1996). Differential selection of species of phytoplankton according to their food value (Beninger et al., 2008; Cognie et al., 2001; Pales-Espinosa et al., 2008; Pales-Espinosa et al., 2007; Shumway et al., 1985) might further contribute to ingested N enrichment. As evidence of the potential of these selective mechanisms to contribute to nutrient homeostatic regulation, Bayne (2009) reported that the ratio of SE N to SE C is a positive function of the nutrient imbalance (N limitation) represented by the difference between the C:N ratio in the seston and the C:N ratio in oyster tissues. Similar selective mechanisms at the postingestive level can also be invoked to account for the N enrichment of the absorbed ration from complex diets (i.e., seston) since digestive selection (Navarro et al., 2016) has been shown to result in the preferential utilization of microalgae (low C:N ratio) vs. sedimentary organic particles, including phytodetritus (high C:N ratio). The expected increment in the absorption efficiency for N compared to C in such scenarios has been certainly reported (Bayne, 2009; Cranford and Grant, 1990; Grant and Cranford, 1991; Prins and Smaal, 1989; Iglesias et al., 1996; Urrutia et al., 1996). Even in the absence of any scope for digestive selection (e.g., more homogenous diet formulations as those used in the present study), the net N enrichment may result from the preferential absorption of dietary proteins relative to carbohydrates and lipids reported in cockles (Cerastoderma edule) fed phytoplankton (Ibarrola et al., 2000b). The almost complete absorption of labile protein from detrital particles (above the AE for overall POM) found in oysters (Adams et al., 2019) might also account for the preferential absorption of N relative to C. Postabsorptive mechanisms concern the metabolic fate of different elements contributing to final adjustments of the assimilated ration to meet the specific requirements for tissue biosynthesis. The release of excess C (or other nonlimiting nutrients) at this postabsorptive level may represent a significant fraction of the metabolic budget in animals feeding on relatively unbalanced foods that are common in both terrestrial and aquatic systems (Anderson et al., 2005). This has been broadly documented in bivalve molluscs (namely, mussels) with regard to several features of protein turnover, where the efficiency of protein deposition was reported to increase with rates of dietary protein absorption (Hawkins, 1985). The increased ratio of protein breakdown to protein synthesis reported under conditions of reduced N input (Hawkins, 1985) probably reflects the metabolic role of dietary protein in fuelling the N pool through transamination reactions for protein synthesis, with the consequent waste of most of the protein C. This would explain the disproportionately higher efficiency of tissue deposition for the amino-N fraction compared with the amino-C fraction reported in mussels (Kreeger et al., 1996; Kreeger et al., 1995) as the result of a N sparing mechanism at the expense of energy (C) even under conditions of negative energy balance (Hawkins and Bayne, 1991). Conversely, N release when N input is in excess of requirements occurs in the form of high rates of ammonia excretion, reflecting the metabolic fate of proteins in supplying energy through deamination reactions. Resort to experimental diets providing a range of variation in the relative proportion of different nutrients, particularly in the protein/energy ratio, have thus proven useful in studying the contribution of these postabsorptive processes to homeostatic nutrient regulation in bivalves. In a previous work (Arranz et al., 2020), we analysed the effects of two different diets that were isocaloric but differed broadly in the C:N ratios on the physiological performance of juvenile clams (Ruditapes philippinarum) measured in terms of the energy budgets. These diets consisted of the same microalga species (Rhodomonas lens) cultured at different growth stages (i.e., exponential and stationary phases). Different groups of clams were food conditioned (15 d.) with these two diets and physiological components of growth and SFG measurements were then performed in response to both the acute and chronic change in dietary composition. Additionally, the study was performed on different growth phenotypes that were obtained by combining selective breeding of two families and size-group segregation inside each family in an attempt to assess how physiological responses to diet composition were modulated according to the variable growth demands of endogenous origin set by the occurrence of these different phenotypes. In the present study, the same experimental design developed to compute energy balances (Arranz et al., 2020) was extended to include the elemental balance of nutrients (C and N), aiming to test the following hypothesis concerning the occurrence of compensatory responses for differential stoichiometric imbalances between high and low N diets: H1. Increased feeding rates acted to compensate for decreased N availability in the diet (overfeeding response). H2. Chronic N deficit of dietary origin is compensated through the preferential absorption of N relative to C (energy): The pre-absorptive level. H3. Metabolic activities further contributed to the stoichiometric adjustments, with metabolic C:N indexes reflecting the differential release of elements in excess: The post-absoptive level. H4. Higher growth demands of endogenous origin are better accomplished under conditions of high N availability which resulted in the maxima balances of both N and C (energy) being achieved in fast growing phenotypes fed the N rich diet. K. Arranz et al. Aquaculture 555 (2022) 738147 3 H5. Less efficient protein metabolism of slow growers is revealed in higher rates of ammonia excretion and reduced values of N net growth efficiency, relative to fast growers. H6. Time course of physiological adjustments to diet quality differ between food conditioned groups and reveals differential phenotypic plasticity. 2. Material and methods 2.1. Animal maintenance, diet characteristics and experimental design Features of diets, clam families, growth categories and the maintenance of them, including experimental design, have been previously described in Arranz et al. (2020). Briefly, specimens belonging to any of two full-sib families (1 and 8) of the Manila clam Ruditapes philippinarum were size-segregated to constitute F (fast growth) and S (slow growth) categories by choosing the largest and smallest clams inside each family. The mean shell lengths for the F and S categories were 23.50 (1.67) mm and 13.10 (1.65) mm, respectively. The different growth category*family combinations resulted in four growth phenotypes (F1, F8, S1 and S8) that were used in subsequent experiments. Maintenance was carried out under constant ambient conditions in a recirculating seawater (34 PPT) system regulated at 17 ◦C. Two groups of clams from each of the above phenotypes were each fed the same ration (~ 0.5 mg POM L −1 ) of the microalgal species R. lens reared in either the exponential (C:N ratio =4.94) or the stationary (C:N ratio = 14.5) phases of the culture (Arranz et al., 2020). The chosen notation was N+for diets based on microalgae in the exponential phase and Nfor those in the stationary phase. These diets differed by ca. 3-fold in N content and were used to assess elemental balances of clams in the context of the acute and chronic physiological response to dietary change. Clam groups were conditioned for 15 days to diets N+or N-. Following that “acclimation” period, physiological measurements were carried out while subgroups of clams from each conditioning group were exposed to either the conditioning diet (chronic response) or the alternative diet (acute response). This design resulted in the 16 experimental groups reported in Table 1. 2.2. Elemental balances Samples for diet characterization were performed twice a week in quadruplicate during the acclimation period and 5–6 times in quadruplicate during the exposure experiments (Arranz et al., 2020). Elemental analysis (CNH) of diets was conducted on samples collected over preweighed glass fibre filters (GF/C) by filtering a known volume of water from the feeding tanks and washing with 50 ml of filtered seawater. Individual samples of faeces produced over8–10 h were similarly collected over preweighed GF/C filters and washed with 50 ml filtered seawater. Both types of samples were immediately frozen at −20 ◦C, lyophilized, and maintained at −20 ◦C until being analysed in a Euro EA Elemental Analyser (CHNS) from Euro Vector, using acetanilide as a standard. A subset of samples was calcined for 6 h at 450 ◦C and subsequently measured in an elemental analyser to be used as a control. Physiological components of the scope for growth were determined as previously described (Arranz et al., 2020), where in short, five replicates per condition were used to determine the clearance rate under the flow-through chamber method (Crisp, 1971; Filgueira et al., 2006), absorption efficiency (Conover, 1966), ammonia excretion rate (Sol- ´ orzano, 1969) and metabolic rate (see mean values in Table 1). Individual values of this dataset were combined with elemental analysis of diets and faeces to compute N and C balances as follows: 2.2.1. Ingestion rates Particulate organic N and C (PON and POC, respectively) were calculated as the product between POM and the proportion of organic N and C present in the diet in the following way: PON (mg L−1)=POM ×N% 100 POC (mg L−1)=POM ×C% 100 The ingestion rates of N and C (IR N and IR C : mg h −1 ) were estimated by multiplying the clearance rate by the particulate organic matter of each element, in these terms: IRN=PON ×CR IRC=POC ×CR 2.2.2. Absorption rates and absorption efficiencies Absorption rate (AR N and AR C , mg h −1 ) constitutes the difference between ingestion (IR N and IR C : mg h −1 ) and egestion rate (ER N and ER C : mg h −1 ), which is calculated through the use of organic ingestion rate (OIR), absorption efficiency (AE) and the proportion of organic N and C present in the faeces, as follows: ERN=OIR × (1–AE) × N%(feces) 100 ERC=OIR × (1–AE) × C%(feces) 100 Absorption efficiency (AE N and AE C , decimal units) is then given by Table 1 Mean (SD) values of clearance rate (CR, L h −1 ), absorption efficiency (AE, decimal units), oxygen consumption rate (VO 2 , ml h −1 ) and ammonia excretion rate (VNH 4 - N, μ g h −1 ), from Arranz et al. (2020), used to calculate the present elemental balances. Food conditioning Acute exposure Growth category Family CR AE VO2 VNH 4 -N N+ N+ F 1 1.22 (0.16) 0.84 (0.01) 0.06 (0.01) 4.26 (1.42) 8 1.1 (0.22) 0.85 (0.01) 0.07 (0.02) 6.31 (7.1) S 1 1.1 (0.13) 0.85 (0.01) 0.06 (0.02) 12.58 (1.74) 8 0.72 (0.2) 0.85 (0.01) 0.08 (0.02) 15.22 (11.56) NF 1 0.9 (0.13) 0.58 (0.04) 0.07 (0.04) 2.83 (1.35) 8 0.74 (0.13) 0.62 (0.02) 0.08 (0.03) 3.62 (1.48) S 1 0.41 (0.11) 0.58 (0.02) 0.06 (0.03) 4.86 (1.21) 8 0.47 (0.11) 0.61 (0.02) 0.06 (0.03) 3.97 (1.06) N− N+ F 1 0.95 (0.09) 0.81 (0.03) 0.1 (0.04) 9.16 (2.64) 8 0.78 (0.41) 0.75 (0.06) 0.07 (0.02) 9.38 (3.25) S 1 0.66 (0.16) 0.74 (0.03) 0.08 (0.04) 19.79 (7.36) 8 0.56 (0.22) 0.78 (0.03) 0.07 (0.02) 11.74 (2.18) NF 1 1.1 (0.22) 0.44 (0.03) 0.05 (0.02) 1.29 (0.35) 8 0.83 (0.44) 0.48 (0.04) 0.07 (0.02) 0.95 (0.03) S 1 0.9 (0.05) 0.53 (0.06) 0.04 (0.03) 0.82 (0.06) 8 0.63 (0.23) 0.65 (0.03) 0.04 (0.02) 0.63 (0.24) K. Arranz et al. Aquaculture 555 (2022) 738147 4 the quotient between AR and IR of each element: AEN=ARN IRN AEC=ARC IRC 2.2.3. Carbon and nitrogen losses Carbon loss due to respiration (R C , mg C h −1 ) was derived from respiration rates (ml O 2 h −1 ), assuming a respiratory quotient RQ =0.9 mol CO 2 mol O 2−1 . Respiration rates were obtained with the aid of oximeters by monitoring the decline in oxygen concentration in sealed chambers filled with seawater where clams had been introduced. Nitrogen loss due to excretion (E N , mg N h −1 ) was assessed through the determination of increments in ammonia concentration by the phenolhypochlorite method (Sol´ orzano, 1969), employing flasks filled with 30 ml of filtered seawater (0.2 μ m Millipore membranes) for 2–3 h. For both determinations, chambers without animals were employed as controls. 2.2.4. Elemental balances and growth efficiencies of carbon and nitrogen Elemental balances of carbon (SFG C , mg C h −1 ) were calculated as the difference between the amount of C absorbed and respired: SFGC=ARC–RC, whereas the elemental balance of nitrogen (SFG N , mg N h −1 ) was calculated as the difference between the amount of N absorbed and excreted: SFGN=ARN–EN The growth (deposition) efficiencies of both elements were computed as follows: NGEN=SFGN ARN NGEC=SFGC ARC After concluding the experiments, clams were dissected to compute the dry weight of tissues, and consequently, all rates were standardized to the common dry weight of soft tissues (85.95 mg) using mass exponents of 0.6091 (CR), 0.6967 (R) and 1.00 (VNH 4 -N). 2.3. Data analysis Relationships between two or more quantitative variables were explored through regression analysis, while comparisons for qualitative factors, family, growth category, acclimation diet and exposure diet were tested through 4-way ANOVA using R (R Core Team, 2018). These analyses were performed after testing for the normality (Shapiro–Wilk) and homoscedasticity (Levene) of the data. 3. Results Results of the present experiments have been disposed in order to fit the sequence of hypothesis set in the introductory section. 3.1. Diet effects H1. Increased feeding rates acted to compensate for decreased N availability in the diet (overfeeding response) Concerning elemental (C and N) composition, acquisition parameters (i.e., rates of food processing and digestive balances: IR, ER and AR) were mostly influenced by the composition of the diet supplied to the animals during physiological determinations (exposure diet) than for previous conditioning, although some effects concerning the latter variable will be considered. In this way, the ingestion rates of N were significantly higher when clams were exposed to the N rich diet, doubling the values recorded in those fed the N poor diet, while egestion rates of N were increased under N restricted conditions (Tables 2 and 3 and Fig. 1). As a consequence, absorption rates were 3 times higher when clams were fed the N+diet compared to the Ndiet. Previous conditioning did not affect physiological performance, although ingestion and absorption rates tended to increase in those individuals previously acclimated to the N+diet. Ingestion rates of C reached the highest values in those groups fed the same diet used in conditioning (e.g., N +N+and N-N-; Fig. 1). However, trends were similar for C and N as regards the net acquisition (higher absorption rates of C in clams fed the N+diet). Thus, maximization of energy obtained from high quality foods appear to be the rule in these experiments, rather than compensatory overfeeding for N income regulation on low quality diets. An interpretation of these features based on the occurrence of digestive constraints will be discussed in the corresponding section. H2. Chronic N deficit of dietary origin is compensated through the preferential absorption of N relative to C (or energy): The pre-absorptive level. Absorption efficiencies for both N and C were clearly superior (p < 0.001) when exposed to the N+diet; acclimation to this high quality diet also had a positive effect but only on AE C . Differential performance for N and C in the different acclimation*exposition groups of clams were analysed by plotting elemental AEs against the overall AE values for organics (Conover) reported in Arranz et al. (Arranz et al., 2020; Fig. 2). As expected, the AE C closely fitted the Y =X line (AE C =1.04 AE – 0.003, R 2 =0.997, p < 0.001), while AE N clearly departed from this relationship, especially in the lower range of AE values (AE N =0.88 AE + 0.13, R 2 =0.88, p < 0.001). As a general rule, AE N was above the AE for overall organics (equivalent to AE C ), but this discrepancy tended to be higher in clams submitted to chronic N deficit (i.e., conditioned to Ndiets). In fact, results of ANOVA testing the difference between AE N and AE (Conover),using acclimation and exposure diet as factors, resulted in significant effects for the acute response (F =9.6, p =0.005) and highly significant effects for both the chronic response (F =114.2, p < 0.001) and the interaction term: (F =43.7, p < 0.001). H3. Metabolic activities further contributed to the stoichiometric adjustments, with metabolic C:N indices reflecting the differential release of elements in excess: The postabsoptive level. N excretion was highly affected by diet composition, such that clams exposed to N+excreted ca. 5 times more ammonia-N than those fed the Ndiet (Fig. 3). This acute response varied depending on the conditioning diet, since clams acclimated to N+and transferred to the Ndiet reduced their rates of excretion but to a lesser extent than animals fully acclimated to N-. Conversely, when clams conditioned on the Ndiet were transferred to the N+diet, excretion rates increased up to 50% more than those of N +-acclimated clams. Carbon loss due to metabolic processes reflected fewer fluctuations between diets; clams exposed to N+reached approx. 20% increment of respired C. As a result of the previous differences, elemental balances for N and C showed significant differences (Table 3) for conditioning, exposure and their interaction. As before, acute exposure to diets with different N contents affected the balance of both C and N to a greater extent, and the best performance was accomplished in clams fed the N+diet (Fig. 3 and Table 2). Conditioning diet was also important, but its effects on elemental balances were only noticeable in animals fed the N+diet, since feeding the N-poor diet tended to reduce the C and N balances to the point of cancelling out the potential benefits of high-quality (N+) conditioning. Overall, metabolic C:N indices (Table 2) reflected the differential stoichiometric release of both elements according to the diet supplied: Both chronic and acute exposure to diets of high N content promoted a 2 to 4-fold reduction of this index (highly significant; Table 3), that was mainly based on increased rates of ammonia excretion. K. Arranz et al. Aquaculture 555 (2022) 738147 5 The above was clearly reflected in the net growth efficiencies for C and N. The efficiency for C (NGE C ) was fairly constant across diets, while NGE N increased significantly in clams fed the Ndiet (Tables 2 and 3) pointing to compensation for N deficit. Regarding both elemental growth efficiencies, interaction terms for acclimation and exposure diets (Table 3) were significant, accounting for the fact that the highest efficiencies were achieved in the experimental groups where conditioning and exposure diets were coincident (N +N+and N-N-). Physiological parameters for each combination of acclimation*exposition diets were determined with a large standard deviation because values were pooled means of different phenotypes represented by families and intrafamily segregated growth groups. Comparisons corresponding to these endogenous sources of variation will be considered next. 3.2. Growth genotype effects H4. Higher growth demands of endogenous origin are better accomplished under conditions of high N availability, which resulted in maxima balances of both N and C (energy) being achieved in fast growing phenotypes fed the N rich diet. Generally, significant effects on physiological parameters involved in elemental balances were mainly due to growth category (Table 3), although differences between families were also detected. Fast-growing juveniles registered higher values for both ingestion and egestion rates of both N and C (Fig. 4). The resulting digestive balances represented by absorption rates were also significantly higher. Indeed, increments of absorption in F compared to S clams ranged from 35 to 40% for N and C, respectively. Similar tendencies between families were not significant (except for N egestion rate), although the results pointed out to a better performance of Family 1 compared to Family 8, whose mean absorption increments amounted only to 4% in terms of N and 8% for C. Consequently, the combination of interand intrafamiliar effects led to a gradual decrease in processing rates from the F1 to S8 groups (Fig. 4). Fig. 5 shows the distribution of AE values (both C and N) across dietary conditions and phenotypes. The mains effect on these parameters consists in AEs increasing their values by ~70% in the change from Nto N+diets. This combines with the ranking distribution of food processing rates among phenotypes (Fig. 4) to result in fast growers fed the high quality diet achieving the highest values for both C and N balances (Tables 2; Fig. 7). In addition, both endogenous factors (family and growth category) exerted weak although significant effects on AE N and AE C (Table 3), and with few exceptions, these efficiencies were higher for F8 specimens than for F1 specimens (<10% difference between families). H5. Less efficient protein metabolism in slow growers is revealed in higher rates of ammonia excretion and reduced values of net growth efficiency for N. Phenotype effects on C respiratory losses were negligible, with only a slight trend for a higher rate in F clams (Table 3). Instead, N excretion of F juveniles (mean: 5.56 μ g h −1 ) was significantly exceeded by S juveniles (mean: 8.02 μ g h −1 ) (Table 3), which is mainly accounted (Fig. 6) for by strong F vs. S differences recorded with the N+diet (low C:N index). Phenotype effects on elemental balances for C and N were restricted to growth category, whereas family effects were not significant (Fig. 7; Table 3). F vs. S differences were highly significant (Tables 2 and 3) and resulted in elemental balances of N that were ca. 3 times higher for F than for S clams, compared with two fold differences for C balances (Table 2; Fig. 7). Net growth efficiencies were, in general; higher for F clams than for S clams (Table 2; Fig. 8), but only differences in NGE N attained the significance level (Table 3). 3.3. Acute vs. chronic dynamics of N and C balances and net growth efficiencies H6. Time course of physiological adjustments to diet quality differed between food conditioned groups and revealed differential phenotypic plasticity. Considering the 4 factors (conditioning, exposure, growth category and family), N and C balances are represented in Fig. 7 to account for the interand intrafamiliar differential performances to acute and chronic dietary changes. As both (N and C) balances were subjected to the same fluctuations, yet each differed in the intensity of these changes, the plots of Fig. 7 will be described together. When clams conditioned to N+diet (N +N+groups) were transferred to N- (N +Ngroups), a sharp decrease of the net balances for both N and C was observed in the four phenotypes resulting from growth group*family combinations (F1, F8, S1 and S8), involving the fact that only about one third of N and half of C was being retained following this change from high to low quality diet. Full conditioning to the Ndiet (N-Ngroups) barely affected the fast growers performance, while S clams nearly doubled their values. The acute response of Nconditioned clams to N enrichment (N-N+groups) also involved differential responses in fast and slow growers: while F clams partly recovered the initial retention of C and N, S clams maintained or even lowered their levels, as occurred in Family 1 S specimens. These differential effects were accounted for by the diet exposure*- growth phenotype interaction term (Table 3), which was only significant for SFG N . Finally, compared to low N conditioning, high N conditioning promoted a neat increase in the net C and N balances for all growth Table 2 Pooled mean values (SE) for conditioning, exposure, growth category and family for IR N and IR C (ingestion rates for N and C), ER N and ER C (egestion rates of N and C), AR N and AR C (absorption rates for N and C), AE N and AE C (absorption efficiencies for N and C), R C (C losses by respiration), E N (N excreted), M C:N (metabolic C:N indices),SFG N and SFG C (elemental balances for N and C) and NGE N and NGE C (net growth efficiencies for N and C): Units are μ g h −1 for rates and SFG and decimal for efficiencies. Conditioning Exposure Growth category Family N+NN+NF S 1 8 IR N 44.64 (5.3) 38.22 (3.82) 58.55 (4.14) 27.01 (1.79) 49.73 (4.29) 30.01 (3.47) 42.61 (4.5) 39.94 (4.7) IR C 329.48 (21.78) 313.78 (22.38) 316.88 (22.31) 324.86 (21.93) 377.29 (17.27) 246.53 (15.53) 337.13 (19.37) 305.38 (24.21) ER N 10.42 (0.87) 9.3 (0.8) 8.57 (0.58) 10.87 (0.91) 11.67 (0.77) 7.38 (0.49) 10.82 (0.82) 8.84 (0.8) ER C 99.96 (11.21) 124.71 (18.26) 61.07 (4.12) 155.77 (14.68) 132.67 (16.65) 86.6 (10.45) 123.88 (15.46) 101.97 (15.64) AR N 34.22 (5.33) 28.92 (3.64) 49.98 (3.76) 16.14 (0.98) 38.06 (4.52) 22.63 (3.38) 31.79 (4.55) 31.1 (4.5) AR C 229.51 (20.57) 189.07 (13.02) 255.81 (19.67) 169.1 (9.5) 244.62 (15.82) 159.94 (12.04) 213.25 (16.74) 203.41 (18.1) AE N 0.7 (0.03) 0.72 (0.03) 0.85 (0.01) 0.61 (0.01) 0.71 (0.03) 0.71 (0.03) 0.69 (0.03) 0.74 (0.03) AE C 0.68 (0.03) 0.63 (0.03) 0.8 (0.01) 0.53 (0.02) 0.65 (0.03) 0.65 (0.03) 0.64 (0.03) 0.67 (0.03) R C 35.81 (2.38) 37.17 (3.86) 41.7 (3.41) 32.24 (2.87) 39.93 (3.42) 31.98 (2.53) 35.93 (4.26) 37.12 (1.83) E N 5.97 (1.13) 7.2 (1.56) 11.57 (1.43) 2.53 (0.37) 5.56 (1.06) 8.02 (1.76) 6.4 (1.43) 6.83 (1.34) M C:N 9.94 (2.16) 17.01 (3.68) 5.04 (0.88) 20.75 (3.36) 14.79 (2.91) 12.12 (3.52) 12.56 (2.94) 14.72 (3.41) SFG N 28.25 (5.01) 21.72 (3.01) 38.41 (4.56) 13.61 (1.11) 32.5 (4.05) 14.6 (2.41) 25.39 (4.16) 24.27 (4.04) SFG C 193.7 (20.44) 151.9 (12.12) 214.11 (19.8) 136.86 (9.81) 204.69 (15.83) 127.96 (12.25) 177.32 (16.37) 166.29 (17.66) NGE N 0.77 (0.04) 0.78 (0.04) 0.71 (0.05) 0.82 (0.03) 0.85 (0.02) 0.67 (0.05) 0.78 (0.04) 0.76 (0.04) NGE C 0.82 (0.02) 0.79 (0.02) 0.81 (0.02) 0.80 (0.02) 0.82 (0.02) 0.78 (0.02) 0.82 (0.02) 0.79 (0.02) K. Arranz et al. Aquaculture 555 (2022) 738147 6 group*family phenotypes. Similar acute vs. chronic dynamics were also observed for net growth efficiencies (Fig. 8), with differential effects accounting for 3rdand 4thorder interaction terms (Table 3), which were only significant for nitrogen. 4. Discussion One important issue from the stoichiometric approach concerning heterotrophic organisms is that trophic relationships might be “misleading” if only energy balances are considered. In the physiological Table 3 Diet conditioning (C) and exposure (E), growth category (G) and family (F) effects on 4-way ANOVA table of the parameters involved in elemental balances of nutrients (IR N and IR C : ingestion rates for N and C, ER N and ER C : egestion rates of N and C, AR N and AE C : absorption rates for N and C, AE N and AE C : absorption efficiencies for N and C, R C : C losses by respiration, E N : N excreted, M C:N (metabolic C:N indices), and SFG N and SFG C : elemental balances for N and C). IR N IR C ER N ER C AR N AR C AE N AE C C F =3.986, p = 0.056 F =0.444, p = 0.511 F =1.975, p = 0.172 F ¼4.675, p ¼ 0.04 F =4.018, p = 0.056 F ¼7.449, p ¼ 0.011 F =2.819, p = 0.105 F ¼29.48, p < 0.001 E F ¼100.671, p < 0.001 F =0.074, p = 0.788 F ¼7.585, p ¼ 0.011 F ¼72.351, p < 0.001 F ¼169.552, p < 0.001 F ¼37.66, p < 0.001 F ¼422.92, p < 0.001 F ¼686.95, p < 0.001 G F ¼23.435, p < 0.001 F ¼30.676, p < 0.001 F ¼31.768, p < 0.001 F ¼26.447, p < 0.001 F ¼17.565, p < 0.001 F ¼23.4, p < 0.001 F ¼5.44, p ¼ 0.028 F ¼11.017, p ¼ 0.003 F F =1.663, p = 0.209 F =1.706, p = 0.203 F ¼5.492, p ¼ 0.027 F =2.193, p = 0.151 F =0.746, p = 0.396 F =0.871, p = 0.359 F ¼10.34, p ¼ 0.003 F ¼5.636, p ¼ 0.025 C*E F ¼4.341, p ¼ 0.047 F ¼6.87, p ¼ 0.014 F =1.959, p = 0.173 F =3.575, p = 0.07 F ¼8.716, p ¼ 0.007 F ¼7.337, p ¼ 0.012 F ¼14.84, p ¼ 0.001 F =0.203, p = 0.656 C*G F =0.306, p = 0.585 F =0.769, p = 0.388 F =0.984, p = 0.33 F =0.002, p = 0.968 F =0.14, p = 0.711 F =2.036, p = 0.166 F =2.425, p = 0.131 F ¼4.509, p ¼ 0.043 E*G F =0.974, p = 0.333 F =0.62, p = 0.438 F ¼5.879, p ¼ 0.023 F ¼9.113, p ¼ 0.006 F =3.715, p = 0.065 F =1.164, p = 0.29 F ¼9.569, p ¼ 0.005 F ¼14.286, p ¼ 0.001 C*F F =0.141, p = 0.711 F =0.189, p = 0.667 F =0.077, p = 0.784 F =0, p =0.99 F =0.291, p = 0.594 F =0.492, p = 0.489 F =0.005, p = 0.946 F =0.832, p = 0.37 E*F F =0.163, p = 0.689 F =0.004, p = 0.951 F =1.256, p = 0.273 F =1.022, p = 0.321 F =0.685, p = 0.415 F =0.464, p = 0.502 F ¼13.15, p ¼ 0.001 F ¼8.746, p ¼ 0.007 G*F F =0.349, p = 0.56 F =0.347, p = 0.561 F =0.297, p = 0.591 F =0.657, p = 0.425 F =0.308, p = 0.584 F =0.097, p = 0.758 F =0.927, p = 0.345 F ¼7.022, p ¼ 0.014 C*E*G F =0.191, p = 0.666 F =0.224, p = 0.64 F =0.085, p = 0.773 F =0.219, p = 0.643 F =0.197, p = 0.661 F =1.243, p = 0.275 F ¼4.76, p ¼ 0.038 F ¼7.834, p ¼ 0.01 C*E*F F =0.002, p = 0.967 F =0.023, p = 0.88 F =0.879, p = 0.357 F =0.308, p = 0.584 F =0.111, p = 0.742 F =0.034, p = 0.854 F ¼4.94, p ¼ 0.035 F =3.53, p = 0.072 C*G*F F =0, p =0.998 F =0.393, p = 0.536 F =3.804, p = 0.062 F =3.434, p = 0.075 F =0.339, p = 0.566 F =0.188, p = 0.668 F =3.87, p = 0.06 F ¼12.416, p ¼ 0.002 E*G*F F =0.34, p = 0.565 F =0.031, p = 0.862 F =1.152, p = 0.293 F =0.041, p = 0.842 F =0.149, p = 0.703 F =0.189, p = 0.667 F =0.946, p = 0.34 F =0.017, p = 0.898 C*E*G*F F =2.375, p = 0.135 F =2.813, p = 0.106 F =0.419, p = 0.523 F =1.689, p = 0.205 F =2.818, p = 0.105 F =2.768, p = 0.108 F =0.61, p = 0.441 F =0.066, p = 0.8 R C E N M C:N SFG N SFG C NGE N NGE C C F =0.114, p = 0.738 F =1.313, p =0.262 F ¼5.596, p ¼ 0.026 F ¼5.305, p ¼ 0.03 F ¼7.027, p ¼ 0.013 F =0.131, p =0.72 F =0.932, p = 0.343 E F ¼5.385, p ¼ 0.028 F ¼69.54, p < 0.001 F ¼30.271, p < 0.001 F ¼80.96, p < 0.001 F ¼26.77, p < 0.001 F ¼9.81, p ¼0.004 F =0.359, p = 0.554 G F =2.963, p = 0.097 F ¼10.294, p ¼ 0.004 F =2.79, p =0.107 F ¼26.25, p < 0.001 F ¼16.86, p < 0.001 F ¼32.399, p < 0.001 F =1.07, p = 0.311 F F =0.038, p = 0.846 F =0, p =0.984 F =1.001, p =0.326 F =0.637, p = 0.432 F =0.86, p =0.362 F =0.108, p =0.746 F =0.89, p = 0.354 C*E F ¼5.751, p ¼ 0.024 F ¼8.387, p ¼ 0.008 F ¼9.702, p ¼ 0.004 F ¼14.82, p ¼ 0.001 F ¼9.969, p ¼ 0.004 F ¼29.558, p < 0.001 F ¼6.68, p ¼ 0.016 C*G F =0.013, p =0.91 F =0.02, p =0.888 F =2.305, p =0.141 F =0.162, p =0.69 F =1.876, p = 0.182 F =3.07, p =0.092 F =0.63, p = 0.434 E*G F =0, p =0.989 F ¼7.509, p ¼ 0.011 F =0.083, p =0.776 F ¼8.031, p ¼ 0.009 F =1.035, p = 0.318 F =3.764, p =0.063 F =0.057, p = 0.814 C*F F =0.253, p = 0.619 F =3.653, p =0.067 F =1.856, p =0.185 F =0.048, p = 0.828 F =0.282, p =0.6 F =0.327, p =0.573 F =0.598, p = 0.446 E*F F =1.025, p = 0.321 F =0.005, p =0.942 F =1.346, p =0.257 F =0.64, p =0.431 F =0.146, p = 0.706 F =0.955, p =0.337 F =0.001, p = 0.981 G*F F =0.557, p = 0.462 F =1.599, p =0.217 F =1.098, p =0.304 F =0.002, p = 0.968 F =0.01, p =0.92 F =1.526, p =0.228 F =0.532, p = 0.472 C*E*G F =0.507, p = 0.483 F =0.207, p =0.653 F =1.09, p =0.306 F =0.343, p = 0.563 F =0.75, p =0.394 F ¼5.983, p ¼ 0.022 F =0.246, p = 0.624 C*E*F F ¼5.023, p ¼ 0.034 F =3.674, p =0.066 F =2.68, p =0.114 F =0.171, p = 0.682 F =0.158, p = 0.694 F =0.836, p =0.369 F =0.829, p = 0.371 C*G*F F =0.794, p = 0.381 F =0.661, p =0.424 F =1.751, p =0.197 F =0.723, p = 0.403 F =0.403, p = 0.531 F =0.444, p =0.511 F =1.175, p = 0.288 E*G*F F =1.988, p =0.17 F =0.632, p =0.434 F =1.113, p =0.301 F =0.004, p = 0.953 F =0.591, p = 0.449 F =0.125, p =0.727 F =1.14, p = 0.295 C*E*G*F F =2.998, p = 0.095 F =1.307, p =0.263 F =2.644, p =0.116 F =3.993, p = 0.056 F =1.258, p = 0.272 F ¼4.963, p ¼ 0.035 F =0.378, p = 0.544 Significant differences (p <0.05) are shown in bold. K. Arranz et al. Aquaculture 555 (2022) 738147 7 description of growth, consideration of specific nutrient balances becomes essential to account for features from feeding, digestion and metabolic behaviour that appear to be associated with the necessary coupling between the available food and the growing tissues (Bayne, 2017). The computation of C and N balances provides a first approach in the road to a more detailed nutritional characterization of these interactions of the organism with the food environment. In the present study, these balances were addressed in a double experimental context: 1) the analysis of the effects of acute and chronic changes between two different diets (N+and N-) that were isocaloric but maintained a ~ 3-fold range of variation in the C:N ratio and 2) the use of different phenotypes to provide for a broad range of growth variability of endogenous origin. 4.1. Diet effects In this section, the results of nutrient (N and C) balances reported with clams maintained with N+and Ndiets will be discussed with reference to potential mechanisms of nutrient homeostatic regulation, providing a means for matching the composition of the assimilated ration and those of the clam tissues. Since the C:N ratios of food in the Ndiet (14.54) largely depart from the corresponding values for body composition that closely resemble instead the composition of food in the N+diet (C:N ratio =4.94), a useful approach will be to analyse physiological responses in the shift from N+to Ndiets as potential mechanisms to offset dietary N limitations. As described in the Introduction section, the preferential ingestion of N based on sorting between different types of particles during pseudofaecal production has proven to be an extremely effective mechanism for the compensation for nutritional deficits (Bayne and Svensson, 2006), but no choice for such selection was available in the present experiments wherein food particles were homogeneous and feeding conditions prevented pseudofaeces. However, the possibility of preingestive selection playing an important role in homeostatic nutrient regulation in clams fed more heterogeneous diets (e.g., seston) cannot be discarded, as it has been reported in different species of bivalves, including oysters, mussels and cockles (Hawkins et al., 1996; Iglesias et al., 1996; Soletchnik et al., 1996; Urrutia et al., 1996). As proposed by Sterner and Elser (2002), another level in the stoichiometric coupling between the animal and the food environment might consist of adjustments of the assimilation patterns wherein the ingestion/digestion/assimilation rates for a given element in default might be upregulated. Bayne (2009) considered the possibility of an increase in ingestion rates to compensate for low nutrient availability as a stoichiometric mechanism to maintain elemental homeostasis without the need for selective feeding. In fact, the results of experiments, very similar to the present study, showed that mussels (M. edulis) fed phytoplankton cultures with a low N content increased feeding rates compared with mussels fed phytoplankton with a high N content (Bayne, 2017; see his Fig. 5.38). However, no such resort to an increment of the feeding rates was observed in the present work, which resulted in ingestion rates of N being reduced by half on average when clams were exposed to the N-poor diet (see Table 2). In a previous analysis, we reported digestive constraints in the processing of Ndiets, resulting in Fig. 1. Acute and acclimated response to dietary N changes for ingestion (IR), egestion (ER) and absorption (AR) rates of N (left) and C (right). Fig. 2. Relationship between AE N and AE C and Conover AE, wherein circles represent the AE C values, triangles represent AE N from clams conditioned on the N+diet, and squares represent AE N from clams conditioned on the Ndiet. For AE N data points, black symbols represent animals exposed to N+, while grey symbols are from those exposed to N-. K. Arranz et al. Aquaculture 555 (2022) 738147 8 poor AE (for overall organics), further hindered by a strong negative dependence between AE and the ingestion rate that does not occur with the N+diets (Arranz et al., 2020). Although we presently lack an explanation for the comparatively poor digestive performance of clams fed R. lens in the stationary phase (the organic component of Ndiets), it becomes clear that these constraints would tend to preclude the efficacy of any strategy based on “overeating” to compensate for reduced dietary N. Conversely, the higher digestibility of N+diets can be inferred from significantly increased values of the absorption efficiency for both C and N exhibited by clams exposed to these diets of high N content (Tables 2 and 3). However, the differential increase in AE N over AE and AE C when fed a N-poor diet (Fig. 2) was the relevant mechanism employed to compensate for the nutrient imbalance, which agrees with the previous literature described in the Introduction section (Cranford, 1995; Cranford and Grant, 1990; Grant and Cranford, 1991; Hawkins, 1985; Hawkins and Bayne, 1985; Kreeger et al., 1996; Smaal and Vonck, 1997). Similarly, Ibarrola et al. (2000a) reported that cockles (C. edule) fed cultures of Tetraselmis suecica in the stationary phase (low N content) improved the absorption efficiency of proteins relative to carbohydrates and lipids compared with the digestive behaviour in cockles fed these cultures in the exponential phase (high N content). While the perfect fit between AE C and overall AE for organics (Conover) suggests a lack of limitations concerning this nutrient across the different combinations of exposure*acclimation diets, nearly consistent departures of AE N from the general relationship strongly point to stoichiometric compensation. Fig. 3. Acute and acclimated response to dietary N changes for absorption rates (AR), N excretion (E; left) or C respiration (R; right) and elemental balance (SFG) rates of N (left) and C (right). Fig. 4. Growth phenotype comparison for elemental acquisition parameters (ingestion (IR), egestion (ER) and absorption (AR) rates) of N (left) and C (right). K. Arranz et al. Aquaculture 555 (2022) 738147 9 On the other hand, significant differential effects on AE N are associated with conditioning rather than exposure to low N diets. For instance, clams conditioned to Nshowed a clear discrepancy of AE N with respect to overall AE that is not seen in those conditioned to N+, even when exposed to the Ndiet. Therefore, regardless of the digestive mechanism involved in the preferential absorption of N, it becomes clear that time is required to elicit a functional response, very likely for a noticeable N deficit to be developed. Additionally, acclimation time might be necessary provided that digestive adaptation proceeds through digestive enzyme induction (Ibarrola et al., 1999; Ibarrola et al., 1998). Bayne (2009) reported AE N :AE C ratios in oysters that range between approx. 0.8 in July to 1.0 in November and 1.3 in March, suggesting that seasonally variable requirements for specific nutrients might be met by enzyme induction. When considering digestive balances (net absorption), one alternative/complementary explanation for the preferential absorption of N may be the C enrichment of faecal materials contributed by endogenous materials in the form of metabolic faecal losses (MFL; Hawkins and Bayne, 1985) that were found to consist mainly of lipids in cockles (e.g., up to 66% compared with 25%of proteins and 6% carbohydrates; Ibarrola et al., 2000a). Finally, the regulation of nutrient balances can be achieved in the postabsorptive phase by adjusting the proportion of C and N released in metabolic activities according to differential income in the diet. This is seemingly achieved through a shift in the composition of the pool of substrates fuelling metabolic energy production. High levels of N consumption would enhance protein degradation, and consequently, high amounts of N are excreted as a consequence of deamination reactions resulting from protein utilization for energy purposes, whereas low N availability may trigger biochemical mechanisms preserving this nutrient for protein synthesis that might, however, be energetically expensive (e.g., high rates of protein recycling have been related to a decrease in protein absorption in mussels; Hawkins, 1985). According to the present results (see Table 2), the N+diet promoted a general increase in metabolic activities, both ammonia excretion and respired C, but the important point is that the C:N ratio for metabolism (Table 2) changes from 5 (fully compatible with the metabolic breakdown of proteins) in clams feeding on the N+diets to 20 in those feeding the Ndiets, which suggests the majority use of other energy substrates. These differences could become even larger if a more precise RQ (respiratory quotient) value reflecting protein-based catabolism (i.e., RQ =0.8) was applied to N+diets instead of the common RQ =0.9. A further source of energy expenditure associated with N-preserving mechanisms, more prominent with the Ndiets, could have been disregarded in the present postabsorptive measurements, since C disposal corresponding to protein metabolism in the form of the amino-C fraction was found to occur in mussels mainly by the excretion of dissolved organic matter (DOM; not determined here) rather than in the form of respired CO 2 (i.e., 60 and 10%, respectively, of the amino-C assimilated) (Kreeger et al., 1996). Taking into account the observed impacts of diet quality on the parameters of the energy balance, a reasonable conclusion is that the acquisition processes contributed to a larger extent to differences in the elemental balances between N+and Ndiets. Elemental balances for both C and N were increased under the low C:N conditions, despite the slight decrease in C availability under the N+diet. This behaviour coincides with the elemental balances calculated from the data shown in Bayne (2009), where both SFG C and SFG N decreased proportionally as Fig. 5. AE N (left) and AE C (right) for every combination of factors: conditioning (N+: dark region, N-: light region), exposure (see x axis), growth category (see x axis) and family (Fam 1: dark boxes, Fam 8: white boxes). Fig. 6. Ammonia-N excretion of fast (solid circles) and slow (open circles) growing juveniles in relation to the diet dosed. K. Arranz et al.