Nutrient Requirements in Diets: Fundamental Issues in Sustainable Aquaculture Development

Abstract

The dietary requirements for essential nutrients serve as fundamental guidelines for formulating animal feeds. Insufficient intake of these nutrients can lead to biological issues, while excessive intake can result in economic inefficiencies and environmental harm. Despite their importance, these dietary requirements have not been consistently determined or reported, with data often varying across different sources, even within the same species. This paper critically examines the mechanisms behind these inconsistencies and proposes several strategies to address them. Although the primary focus is on dietary phosphorus in fish nutrition, the concepts discussed may also be relevant to other nutrients and animal species.

1. Introduction

In a book released by a prestigious publisher, Hepher [1] compiled data on the dietary phosphorus (P) requirements necessary for optimal growth in various aquacultured fish species (p. 242). His table shows a discrepancy in the requirements for channel catfish Ictalurus punctatus: one study indicates a need for 0.8% P in the diet, while another study reports a requirement of only 0.4% for the same species. Similarly, the table indicates a P requirement of 0.68% for red seabream Pagrus major, compared to just 0.11% for black seabream Acanthopagrus schlegelii. Furthermore, the dietary P requirement for rainbow trout Oncorhynchus mykiss varies widely, ranging from 0.65% to 1.09%. The National Research Council [2] has compiled the dietary requirements for various aquaculture species. One table reports P requirements for channel catfish ranging from 0.33% to 0.8% per diet, for rainbow trout from 0.54% to 0.8%, and for Japanese flounder Paralichthys olivaceus from 0.6% to 1.5%. The dietary requirements for other minerals and vitamins also vary considerably, even within the same or closely related species, depending on the reference source.

How can we trust these values? Which values are accurate, if any? Regrettably, the authors did not provide explanations for these discrepancies. Obviously, relying on such inconsistent values is difficult and risky, as insufficient intake can lead to poor growth or even mortality in cultured fish, while excessive intake may be toxic to fish and cause economic losses and environmental harm. Therefore, it is crucial to identify the underlying factors contributing to these discrepancies and to find solutions to address them.

The term “dietary requirement”, though so commonly used, may be misleading. In reality, the requirement is only for specific physiological needs such as growth, reproduction, and disease resistance. The diet is merely a vehicle for each nutrient. Hence, when the requirement values are expressed based on dry diet weight, they can be highly inaccurate and misleading. Since Adolph’s experiment in 1947 [3], it has been shown in several monogastric animals that when diets are diluted with inert materials, such as cellulose and kaolin, to produce diets with varying energy densities, the animals are able to adjust their food intake so that the amount of calories (energy) eaten remains constant [4,5,6,7].

In other words, when a nutritionally complete, energy-dense diet is mixed with an indigestible ingredient, for example, in a 1/1 ratio, the mixed diet should still be nutritionally complete. The only difference is that the animal must consume twice as much of the diluted diet to compensate for the reduced energy density in order to achieve comparable growth. Using goldfish Carassius auratus, Rozin and Mayer [8] demonstrated this relationship. Similar findings have been reported in rainbow trout [9,10,11,12].

In the diluted diet, the concentrations of all nutrients, including vitamins, minerals, amino acids, and protein, among others, are reduced by half. This demonstrates that expressing dietary nutrient requirements based on diet weight can result in serious inaccuracies. (When cellulose is used as a diluent, its bulkiness can limit feed intake due to the gastrointestinal tract’s capacity. Therefore, denser fillers are better suited for studying fish’s compensatory feeding ability.)

Traditionally, commercial aquaculture feeds have been primarily composed of fish meal. However, it has become increasingly common to incorporate significant amounts of plant ingredients into modern or sustainable aquaculture feeds [13,14]. These plant ingredients vary widely in energy content, protein, lignocellulose, anti-nutritional factors, and digestibility. Additionally, the availability of feed ingredients differs greatly by region. In industrialized countries, high-quality ingredients are often used to produce energy-dense, highly digestible feeds for profitable aquaculture, whereas in developing countries, only fibrous plant materials or waste products may be available as feed ingredients. As a result, the nutrient and energy density of fish feeds can vary greatly [15]. Therefore, it is important not to express or report nutrient requirements based solely on diet weight. Alternative methods for expressing nutrient requirements are discussed below.

2. How to Express Nutrient Requirements

2.1. Per Digestible Energy

The nutrient requirements are much more accurately expressed on a digestible energy (DE) or metabolic energy (ME) basis than per weight of feed. This is because fish, like other animals, consume feed to meet their energy needs and stop eating once their caloric demands are fulfilled [7,16,17]. There is a clear inverse relationship between the expected total feed intake (feeding rate) and the DE of the feed [18,19]. Notably, earlier research has reported that the dietary protein requirement of brook trout was approximately 14% of their total caloric intake [20].

However, expressing dietary requirements based on the energy content of a diet presents several challenges. When the dietary protein-to-DE ratio falls below the protein requirement, fish growth becomes limited by insufficient protein intake—similar to kwashiorkor in humans. At this point, other nutrients can no longer be accurately expressed in terms of DE. Interestingly, chickens appear to eat primarily to meet their protein requirements rather than their energy needs, making DE-based measurements similarly inaccurate [21].

Moreover, the source of energy—whether fat, protein, or carbohydrate—has different physiological effects even when the caloric values are equivalent. Carbohydrate utilization, in particular, varies greatly depending on factors like digestibility, dietary levels, and fish species. Many carnivorous fish species have a limited ability to metabolize dietary carbohydrates. The balance of amino acids in protein, their bioavailability, and their dietary levels can shift protein utilization from growth (deposition) to consumption (ATP synthesis). Dietary fats, on the other hand, are either stored in the body or used as an energy source, depending on the overall nutrient balance, level of intake, and the fish’s physiological state. Additionally, exercise, stress, and cultural conditions can divert energy flow away from growth and toward consumption [1,2,18].

Physiologically, retained energy (growth) and consumed energy are fundamentally different. Dietary protein used for somatic growth requires a corresponding amount of P to maintain the proper P/N stoichiometric ratio in the body [22]. However, when dietary protein is consumed as an energy source, no additional P is needed. This balance is highly variable and difficult to predict. As noted by Grisdale-Helland et al. [23], the voluntary feed intake of Atlantic salmon Salmo salar and rainbow trout at satiation was significantly higher with a high-energy diet compared to a low-energy diet, resulting in much higher energy intake in fish fed the high-energy diet. This clearly contradicts the principle of feed intake, where an inverse relationship between feed intake and DE content is typically expected. Additionally, the feed efficiency (FE), measured as weight gain per unit of feed consumed, was similar or slightly lower with the high-energy diet, which also contradicts the usual response of higher FE with high-energy feeds.

Boersma and Elser [24] and Benstead et al. [25] explored the P/C (phosphorus/carbon) ratio to determine the optimal dietary P level for maximizing fish growth. However, this approach essentially expresses P requirements in terms of total energy and, as such, may be subject to the same concerns mentioned above.

2.2. Per Day

For humans, the Recommended Dietary Allowance (RDA) represents the recommended daily intake of nutrients, established based on the standard growth rates for different age groups. Expressing nutrient requirements per unit of body weight per day is generally acceptable for homeotherms, as their normal growth rates, varying by age, are well understood [26]. However, for poikilotherms, including fish, daily feed intake—and consequently, growth rate—depends heavily on the rearing temperature, which directly affects the metabolic rate. This makes the “per day” expression of nutrient requirements inappropriate for these species.

2.3. Per N Retention (Stoichiometric Ratio)

Since fish retain nutrients in proportions that align with the stoichiometric composition of their bodies, nutrient requirements—especially those contributing to body composition—can be accurately expressed based on growth increments [22]. The stoichiometric formulation of fish feeds is based on the nutrient composition of the body and the assimilation (retention) rate of these nutrients from the diet. Nutrient assimilation rates can vary significantly depending on numerous factors, which must also be considered when formulating balanced diets.

Early studies indicate that P is required for growth (N retention, protein accretion), while N is essential for P retention. An appropriate balance between P and N intake is necessary. For example, Gevaerts [27] found that P excretion in the urine of rats on a P-free diet was much lower than during starvation. Additionally, diets of sucrose with protein resulted in much less urinary P compared to a sucrose-only diet. Gregersen [28] found that in rats, even with an abundant intake of P in an assimilable form, no P was retained from a protein-free diet. Wolf and Oesterberg [29] noted that feeding starving dogs with a small amount of protein reduced urinary P excretion to a very low level, while feeding starch and fat had little or no effect on P excretion. Kleiber et al. [30] calculated the partial relative protein catabolism (ΔN in urine/ΔN digested), finding it to be 0.68 in P-deficient beef heifers compared to only 0.10 in control animals, indicating that P is necessary for N retention and, consequently, for growth.

In undernourished humans, Rudman et al. [31] found that the retention of P, potassium (K), sodium (Na), and chloride (Cl) virtually ceased when N (amino acids) was removed from an otherwise complete hyper-alimentation fluid. At all levels of N intake, these five elements—including N—were retained by the body in a fixed ratio. Similarly, the withdrawal of P halted the retention of the other elements.

The dietary requirements for nutrients, particularly those that constitute body tissues, may be most accurately expressed per unit of weight gain (more precisely, N retention or lean gain [16]). The protein-to-ash ratio in the bodies of animals of the same species remains relatively constant across different life stages and nutritional states, except in cases of starvation or malnutrition [32]. The allometric effect of body size on the N-to-P ratio (N/P) is also negligible for fish beyond the larval or early juvenile stages [33,34], as well as for growing pigs [35]. Due to this stoichiometric homeostasis between N and P, expressing the dietary available P requirement based on retained protein (N) should be accurate and stable.

In rainbow trout, Lanari et al. [36] noted that the fish retained more dietary P and excreted less P by increasing the dietary protein content. Also in trout, Sugiura [37] reported the available P requirement per unit of N gain. The data show that fish weighing 200 g require 0.277 g of P per gram of N gain calculated based on non-fecal excretion and 0.293 g based on 95% body saturation. In growing pigs, the NRC [35] listed dietary P and Ca requirements, which varied depending on the protein deposition rate (g/day).

Theoretically, during growth, there should be a parallel relationship between P and N retention in the body, with the P/N stoichiometric ratio reflecting that of the whole body. However, during starvation, the parallel stoichiometric ratio is reflected in the excreta, where the P/N ratio follows that of the waning muscle tissues rather than the whole body, as bone resorption is negligible [38]. Up to the maintenance N intake level, there is no net requirement for P, and all dietary P should be excreted on a net basis [38].

Schindler and Eby [39] demonstrated the relationship among fish growth rate, dietary P content, and P excretion. At zero growth (i.e., zero N balance, maintenance level), P excretion in fish is directly proportional to dietary P content, as expected. However, as fish growth rates increase, those fed a low-P diet gradually become P-deficient. This suggests that P deficiency is unlikely to occur during periods of low growth—caused by factors such as low feed intake, low water temperature, poor feed quality, poor rearing conditions, or stress—even if the dietary P content is much lower than the requirement values reported in the literature. In other words, under these conditions, providing standard levels of P in the diet can lead to excessive urinary P excretion, contributing to the pollution of surrounding water bodies.

Jahan et al. [40] reported that P excretion by fish increased as body weight exceeded approximately 600 g, attributed to slower growth in larger carp Cyprinus carpio. During this period, with water temperatures below 15 °C, growth was inhibited, resulting in maintenance feed intake where most dietary P was excreted. This is similar to adult humans consuming P-containing food daily but not growing, leading to the excretion of all dietary P on a net basis.

Expressing P requirements relative to N retention (following P/N stoichiometry) is quite reasonable. However, it poses a practical challenge, as N retention (i.e., lean gain or protein accretion) cannot be accurately predicted based on data available in the literature. Nonetheless, when evaluating the accuracy of experimental data, expressing the requirement per unit of N gain—which must be determined through chemical analysis—provides a rational basis for comparison with data from other experiments conducted using different diets, fish sizes, and culture conditions [22].

2.4. Per Feed Efficiency (FE: Fish Weight Gain/Feed Fed)

This metric is a slightly compromised alternative to the N retention method discussed above. In this case, fish weight gain is used instead of N gain to normalize or standardize the requirement value. The compromise lies in using weight (instead of N), which does not account for differences in fish body composition, such as protein, fat, or moisture levels. Shearer [41] and Asgard and Shearer [42] indicated significant differences in reported dietary requirement values even within the same species and noted that the differences are strongly related to FE.

Conventional FE is reasonably accurate and highly practical, as it only requires weighing fish on site. It is considered “practically accurate” because body protein (N) content is known to be relatively stable across different life stages and dietary conditions in both fish [43] and mammals [5], except under marasmic conditions. This is because the total water and fat content in the whole body—both of which are non-N components—remains relatively constant even though water and fat content can vary individually [32,44,45,46]. For example, in starved mice, whole-body fat content decreased, and water content increased, while ash and protein content remained unchanged [5].

In practice, dietary requirements, typically reported per unit weight of feed, can be normalized by dividing by the FE of that particular diet (or multiplying by the feed conversion ratio, FCR: feed fed/fish weight gain). This adjustment provides the requirement per unit of fish weight gain, which represents the dietary requirement when FE is 100% or 1.0 (Figure 1). This rectification was proposed by Sugiura et al. [47] as the standard requirement or Requirement Coefficient. The standard requirement is universal and can be multiplied by the FE of any diet used in practical feeding; namely, the dietary requirement of a practical diet (g/g feed) = standard requirement (constant) × FE of that diet. This standardization ensures that the requirement value is both accurate and practical. The standard requirement for each nutrient remains constant, regardless of feed composition, fish size, and most culture conditions.

In Figure 1 (see the horizontal dashed line), when fish require 1% available P in a diet with an FE of 100% (vertical dashed line), the dietary requirement decreases to 0.5% in a diet with an FE of 50% (vertical dashed line; e.g., in low-protein feeds, for large adult fish, during winter feeding, under stressful conditions). Conversely, the requirement increases to 1.5% in a diet with an FE of 150% (vertical dashed line; e.g., in high-performance feeds, for larval–early-juvenile fish). In either case, the standard requirement remains 1% (horizontal dashed line). Dietary requirements should change proportionally with FE. If this relationship is ignored, fish fed high-performance diets or fast-growing juveniles may become P-deficient, while fish fed economical diets or slow-growing adults may excrete excessive soluble P, leading to environmental problems.

Practically, in a diet with an FE of 100%, the dietary available P requirement to maintain normal body P levels should closely match the fish’s wet-body P content (discussed below). This is because the definition of (apparent) available P is the net amount of P absorbed into the body [48], and renal endogenous obligatory P loss is negligible at or below the requirement intake level [37,47]. Following this rationale, the available P content in each diet should be standardized (normalized) based on the FE of the respective diet, assuming dietary P is not directly limiting the FE. By doing so, the adequacy of the dietary P level can be accurately assessed by comparing it to the standard P requirement or standard body P content, as illustrated below.

(1)

Standardized P requirement = (available P/feed)/FE.

Since FE = body weight gain/feed, (1) can be rewritten as follows:

(2)

Standardized P requirement = available P/body weight gain.

When (2) is lower than the standard body P content of fish (wet weight basis), the diet is likely P-deficient. Conversely, if (2) exceeds the standard body P content, the diet is likely P-excessive.

As an example, Miñoso et al. [49] fed milkfish Chanos chanos (initial body weight 2.6 g, final 36 g) for 22 weeks with a semi-purified diet containing a normal P level. A P-deficient diet was also created by omitting the inorganic P supplement from the normal-P diet. The P-deficient diet contained 0.296% of total P (mostly available P). However, the reported feed conversion ratio (FCR) of the normal-P diet was 1.81 (equivalent to an FE of 55.2%), meaning the standardized P content of the P-deficient diet was 0.536% (= 0.296 × 1.81). Since the whole-body P content of milkfish is 0.6–0.7% [50], it is evident that the P-deficient diet was only slightly deficient in P, despite its low dietary P content.

In another study with juvenile milkfish, Borlongan and Satoh [51] determined the dietary P requirement using several response criteria with semi-purified diets containing incremental P concentrations. They reported that the dietary P level required for optimal growth and mineralization was 0.85% of the dry diet. However, the FE of their P-sufficient diet was only 65%. Therefore, the standardized requirement would be 1.31% (=0.85/0.65), which is about twice the whole-body P content. Obviously, this discrepancy warrants further discussion. The following table (

Table 1. Examples of highly variable feed efficiency (FE) observed in practical feeding.

Fish 1 (Body wt: Initial–Final)	Diet Type	FE (%) 2	Ref. 3
Mirror carp (28–106 g)	Economical feed (locally made)	65	1
Mirror carp (28–202 g)	Commercial feed (for carp)	105	1
Mirror carp (28–248 g)	Commercial feed (high protein)	130	1
Atlantic salmon	Commercial feeds of the 1970s	42	2
Atlantic salmon	Commercial feeds of the 2000s	118	2
Rainbow trout	Commercial feeds of the 1970s	51	2
Rainbow trout	Commercial feeds of the 2000s	91	2
Yellowtail (40–1244 g)	Commercial feed	56–67	3
Yellowtail (900–4600 g)	ibid.	45–48	3
Amberjack (100–210 g)	Commercial feed (CP 54%, fat 15%)	105	4
Amberjack (1500–2100 g)	ibid.	54	4
Atlantic salmon (486–962 g)	Commercial feed (CP 53%, fat 28%)	103	5
Rainbow trout (415–859 g)	ibid.	90	5
Atlantic cod (439–682 g)	ibid.	143	5
Red seabream (10–15 g)	Commercial feed (CP 52%, fat 15%)	77–91	6
Red seabream (80–110 g)	ibid.	67	6

¹ Mirror carp Cyprinus carpio; Atlantic salmon Salmo salar; Rainbow trout Oncorynchus mykiss; Yellowtail Seriola quinqueradiata; Amberjack Seriola dumerili; Atlantic cod Gadus morhua; and Red seabream Pagrus major. ² FE: feed efficiency (%) = fish wet weight gain × 100/dry feed weight. ³ Sources: 1 Sugiura [52]; 2 Hardy and Gatlin [53]; 3 Miyashita [54]; 4 Watanabe [55]; 5 Grisdale-Helland et al. [23]; and 6 Takii et al. [56].

) provides examples of highly variable FEs, highlighting the importance of standardization.

3. Additional Issues

Feed efficiency (FE) varies not only with feed composition but also with factors such as growth rate and physiological state (

Table 2. Dietary requirements * can change depending on the following factors.

1. Factors that can be normalized by feed efficiency (FE)

► Feed composition, including protein and energy density, nutrient balance, deficiencies in essential nutrients, palatability, anti-nutritional factors, digestibility of macronutrients, indigestible matter content, immunomodulators, feed processing-storage conditions, etc.

► Growth rate that is influenced by fish size or age, species or strain, feeding rate and frequency, water temperature, water quality, various stressors, and other culture conditions.

► Physiological state that is influenced by maturation, reproduction, smoltification, and disease infection, among others.

2. Factors independent of feed efficiency (FE): technical factors

► Feeding duration, i.e., growth magnification to offset initial body reserves; thus, also diet history (described in the text below).

► Nutrient supply from non-dietary sources, e.g., natural organisms, bioflocs, soil, and water (branchial and surface absorption of various ions and drinking water containing minerals). When these factors are significant, dietary requirements will be water-dependent.

► Leaching loss of soluble nutrients (significant in microparticulate diets and loose pellets) and uneaten feed due to excess feeding, poor feeding practices, and bad pellets (inadequate sizes, fines, and loose pellets).

► Response criteria used, e.g., growth, mortality, bone P content, gene expression, immune function, enzyme activity, blood levels, and other indicators (described in the text below).

► Methodology including dose–response, factorial estimates, balance method, statistical methods (e.g., logistic regression, polynomial, and broken-line), and the level of plateau chosen (e.g., 90%, 95%, and 100%) (described in the text below).

* Dietary requirements should always be expressed based on the available (absorbable) nutrients rather than the total amounts. Total amounts include both absorbable and non-absorbable fractions, making them an unreliable metric for expressing the dietary requirements for essential nutrients. By definition, non-absorbable forms of “nutrients” are not classified as nutrients. Digestion and absorption are factors independent of nutrient requirements or systemic physiology and should be considered separately.

(Part 1)). These factors influence FE and, therefore, can be standardized by FE. However, numerous other factors, independent of FE, can also impact the measurement of dietary requirements (Table 2 (Part 2)). These technical factors must be addressed with careful consideration.

3.1. Growth Magnification

Suppose the feeding duration is limited to only one week; the dietary requirement estimates (of vitamins and minerals) based on growth are likely to be zero. Conversely, estimates based on nutrient retention would be infinite (i.e., no saturation plateau). A feeding duration long enough to allow substantial growth from the initial body size (i.e., growth magnification) is essential for accurately assessing dietary requirements. In other words, growth is essential for inducing nutrient deficiencies and studying dietary requirements.

Roloff [57] fed dogs a diet low in Ca and observed the development of rickets, noting that the rapidity of growth and the degree of Ca deficiency influenced the onset of the disease. E. Voit [58] fed puppies a mixture of meat and lard with or without Ca supplementation and found that those lacking Ca developed rickets, with the severity of the disease correlating directly with the growth of the animals. In fish, Embody and Gordon [59] noted that rapidly growing young trout in hatcheries require more Ca and P for skeletal development. Murakami [60,61] reported that fast-growing carp exhibited more severe bone deformities (due to P deficiency) than slow-growing carp cultured in the same pond.

A latent period often precedes the appearance of clinical deficiencies. Kleiber et al. [30] found that beef heifers fed low-P diets grew normally for six months but then stopped growing and maintained their weight for a year, eventually losing weight, while heifers on normal-P diets continued to grow throughout. Similar findings have been reported in pigs [62], rats [63], and chicks [64].

Nose and Arai [65] reported that Japanese eel Anguilla japonica required 0.29% P in the diet for optimal growth. In their study, the highest weight gain was only 45% of the initial weight. When growth magnification is at such a low level, the dietary requirements for most nutrients can be underestimated if fish growth is used as the response criterion because fish utilize not only dietary nutrients but also those stored in their body reserves.

Hardy et al. [66] fed juvenile rainbow trout a P-deficient diet, a P-adequate diet, or a mixture of the two diets in various ratios over eight weeks. Fish fed the P-deficient diet exhibited clinical signs of P deficiency, including anorexia, transient lethargy, reduced growth, and dark coloration, within five weeks. Fish fed a 9/1 ratio of the P-deficient and P-adequate diets showed these signs after seven weeks. Subclinical P deficiency did not affect fish growth until body P stores were reduced below a certain threshold level. Uyan et al. [67] also observed that juvenile flounder (1 g body weight) fed a P-deficient diet grew well for the first 20 days without showing any signs of P deficiency, compared to fish fed P-sufficient diets.

Similar relationships exist between body nutrient stores and growth magnification for other essential nutrients. Dupree [68] fed channel catfish a vitamin B12-free diet, and the fish grew normally for many months, similar to those fed a VB12-supplemented diet. However, after this latent period, the vitamin-deficient fish began to exhibit reduced growth rates, indicating that growth only declines once body stores are depleted below a certain threshold. In the same study, similar results were reported with several other vitamins. The initial body store, or pool size, can vary depending on the dietary history of the fish or animals and the specific nutrient in question [21].

The feeding duration must be long enough to minimize the effect of the body’s reserves or diet history. This is particularly relevant when working with large fish or conducting experiments under suboptimal conditions. However, errors arising from these factors cannot be entirely eliminated, making it essential to verify the accuracy of the estimated requirements. One approach is to determine the requirement at various time intervals (e.g., every other week) until the values stabilize [21]. In trout, Sugiura et al. [69] noted that serum P responded quickly to dietary P levels, but it took several weeks to reach a stable response. Bone P responded more slowly to dietary P, requiring an extended feeding period to achieve a stable response. Molecular markers (mRNA abundance) did not show stable responses; many genes responded clearly during the acute phase but less so during the chronic phase.

As nutrition research on large fish is particularly difficult, it may require different methods than those used for smaller fish. Using sensitive indicators, like urinary P excretion, is more practical for assessing dietary P adequacy. Studies by Rodehutscord [70] and Sugiura [37] show that the P requirement of large trout, based on growth, is lower than that estimated from urinary, plasma, or bone P levels. This difference may arise from indicator sensitivity and varying definitions of “requirement” (i.e., minimum necessary vs. tissue saturation). Since urinary P is a major pollutant in aquaculture, minimizing it through diet is environmentally important [71].

3.2. Rational Criteria

The assessment of nutritional adequacy and the determination of dietary requirements for essential nutrients have traditionally relied on observable deficiency signs. Among these indicators, the mortality rate is the most definitive and can also be used to establish dietary requirements, particularly during early development. Various deficiency signs have been documented for each essential nutrient across different animal species, such as growth rate, feed efficiency, disease resistance, and environmental impact, among others.

However, the justification for using factors like bone density maximization, tissue accumulation (saturation), blood levels, enzyme levels, and gene expression as bases for determining dietary requirements is questionable. Animals may simply be physiologically adapting to changes in dietary intake, compensating for either increases or decreases without necessarily reflecting a clinical deficiency. Maximizing tissue saturation of a particular mineral or vitamin does not imply that the animal requires such high intake levels to maintain optimal health. In fact, at the point of body saturation, fish might excrete significant portions of the absorbed nutrients through urine or gills. Thus, response criteria for determining dietary requirements should not be selected solely based on sensitivity or responsiveness. Instead, the physiological and practical importance of those criteria should guide the choice.

In pigs and chickens, it is well established that the dietary P requirements for maximum bone strength and bone ash content are higher than those for maximum weight gain [72,73]. Similarly, Ogino and Takeda [74] found that the dietary requirement of available P for maximum growth in juvenile carp (initial body weight ~4 g, final 7–12 g) was 0.6–0.7%, while the requirement for maximum bone mineralization was higher, around 1.5%. Bureau and Cho [75] reported that increasing dietary P intake had no significant effect on the growth and feed efficiency of rainbow trout but did significantly increase the P content in the whole carcass, vertebrae, plasma, and urine. Rodehutscord [70] also observed in rainbow trout that the P requirement for maximum weight gain (3.7 g/kg diet) was lower than that for maximum P deposition or bone calcification (5.6 g/kg diet).

For some essential nutrients, the oral dose required to maximize certain responses (e.g., disease resistance and enzyme activity) is often a pharmacological level far exceeding the minimum requirement necessary for normal performance in animals and humans. While using these response criteria is valid, such cases should be categorized separately—not as essential nutrients (dietary requirements) but as pharmacological agents or immunostimulants.

In medicine, early detection and early treatment are of paramount importance, leading to the development of various diagnostic tools (e.g., blood or urine analysis, CT, MRI, and biochemical and molecular markers). In fish, however, invasive methods can be employed (with the exception of endangered species and valuable fish like Koi). Therefore, diagnostic methods for fish can be more direct, such as analyzing bone P content, rather than relying on expensive medical instruments that are often less accurate or reliable. Using advanced instruments does not mean the research is advanced.

3.3. Factorial Approach for Comparison

The factorial approach has been explored by several researchers to estimate the dietary P requirements of fish.

• Pfeffer and Pieper [76] used the following formula to derive the requirement value:

Gross requirement = (Amount retained + Endogenous loss) × 100/Availability.

• Shearer [77] refined the dietary requirement calculation as follows:

Dietary requirement = Gross requirement/Feed consumed, where Gross requirement = (Requirement for new tissue synthesis + Endogenous loss − Uptake from water)/Bioavailability.

• Nakashima and Leggett [78] provided the following formula:

P_surplus (i.e., urinary P excretion) = P_ingestion (food) − P_growth − P_egestion (feces) − P_maintenance. From this, the dietary P requirement (P_ingestion when P_surplus is zero) can be expressed as dietary P requirement = P_growth + P_maintenance + P_egestion.

The P requirement for maintenance (endogenous obligatory loss) may not be accurately determined in starved fish, as starvation is a catabolic state, while growth is an anabolic state. Growing fish require minerals, while starving fish do not. During starvation, fish excrete P in proportion to the obligatory N loss from wasting muscle, based on the N/P ratio. Therefore, fish fed a P-free diet lose much less, or negligible, P compared to starving fish [37]. This phenomenon is also well known in higher animals and humans, as noted above.

Regarding fecal obligatory loss, this is also negligible in fish, as the apparent digestibility of sodium or potassium phosphates is typically ~98% [38]. This indicates that only up to ~2% of dietary P (at normal intake levels) might be attributed to obligatory endogenous loss, assuming true digestibility is 100%. Notably, if apparent availability is used in the aforementioned formulas to calculate the requirement, fecal obligatory loss should be disregarded, and only non-fecal loss becomes relevant [48], which, as mentioned earlier, is nearly zero. Thus, for P, the factorial approach can provide a reasonably accurate estimate of dietary requirements. However, for other minerals, factorial estimates may be less reliable due to factors such as the absorption of waterborne minerals, endogenous excretions, and the lower bioavailability of dietary sources.

In factorial calculations, Ogino [79] observed that the amino acid composition of fish bodies closely corresponds to the dietary amino acid requirements established through dose–response feeding experiments. However, Schwarz [80] noted that while the factorial deduction method can provide approximate values for mineral requirements, the dose–response technique is the most reliable approach for estimating these needs. This may hold true for small, fast-growing fish, but for adult or post-juvenile fish, the dose–response technique can be ineffective or even misleading, as discussed earlier.

4. Conclusions

Reporting inaccurate data is a futile effort, leading only to confusion. This review examined the root causes of errors in nutrient requirement studies, with the primary issue being the lack of standardization (normalization) of raw data based on FE. The reported values of dietary requirements can be considerably more robust and reliable when (1) rational response criteria are chosen, (2) dose–response data are validated through repeated measures, (3) raw data are standardized based on FE, and (4) the results are compared by factorial calculations.