The buccal mass comprising the beak, radula and associated muscles is the point at which food enters the body with the size of the bolus determined by the gape of the beak and the bite force. The food is mixed with secretions from the anterior and posterior salivary glands which contain digestive enzymes but also in the case of the posterior glands toxins used to subdue or kill prey (see below for details and Ponte and Modica, 2017).

From the buccal mass, pieces of food are propelled along the cuticle lined oesophagus which passes between the sub- and supra-oesophageal lobes of the brain linked by the connectives. The distensibility of the oesophagus and the encircling brain tissue will limit the bolus size together with the compressibility of the food by the oesophageal muscle which in its proximal part has functional characteristics of striated muscle (Andrews and Tansey, 1983a).

Following passage along the oesophagus, the bolus enters the crop with very limited evidence for a sphincter between the two structures (for details see Sykes et al., 2020). Diagrams of the digestive tract in O. vulgaris (see Figure 1) sometimes show a bridge of tissue linking each posterior salivary gland to the anterior crop and this can also be seen in published pictures (for O. vulgaris see fig. 1 in Baldascino et al., 2017; for O. americanus see fig. 3 in Avendaño et al., 2020). Although this structure appears like a duct transporting salivary secretions into the crop, this has not been demonstrated (G. Ponte and P.L.R. Andrews, unpublished observations). Studies of this structure should be undertaken in paralarvae to investigate patency as the structure in adults may be vestigial. Another possibility is that the connexion serves to hold the salivary glands in position as the crop fills.

When distended with food, the crop can be seen to have a more distensible anterior half (see Figure 2) which accommodates ∼75% of the content and a narrower posterior part connecting to the gastric vestibule (Andrews and Tansey, 1983a) with no strong evidence for a sphincter. The gastric vestibule is one of two thin layers of muscle holding the two thick muscle blocks closely apposed and responsible for the movement of one thick block against the other. The stomach has a relatively thick cuticle lining. The anatomical similarity of the stomach in octopus (and to a lesser extent cuttlefish) to that of the koilen-lined “gizzard” present in some species of bird was first noted by Aristotle (1910). The gastric vestibule connects with both the spiral caecum and the intestine with evidence for sphincters between these structures lacking. The major enzymes responsible for digestion (see below) originate in the digestive gland, are passed into the caecum (DG ducts) from where they can enter the gastric vestibule to contact the food. The capacity of the stomach in octopus is small in relation to the crop so only relatively small boluses of food can enter for trituration while enzymes continue to act and then the partially digested food progresses to the either caecum or intestine. Alternatively, the crop and stomach may act as a single functional unit with food interchanged between the “hopper” (crop) and “mill” (stomach) until it is physically and chemically degraded and is passed to the caecum or intestine. Currently, there is no definitive evidence to decide between the two, not incompatible, options. The common cavity between the crop, stomach, caecum and intestine means that contents can readily interchange depending upon the intraluminal pressure gradient generated by the muscles in each region.

In O. vulgaris, the oesophagus, crop and particularly the stomach are all lined with a cuticle. The cuticle protects the underlying epithelium against abrasion by ingested prey (e.g., crab carapace and limbs, fish bones) and in the stomach provides a surface for trituration. The presence of a continuous, variable thickness, layer of cuticle makes it highly unlikely that the oesophagus, crop or stomach has an absorptive function. However, it should be noted that limited evidence has been provided for uptake of radio-labelled glycine/leucine by the crop in Nautilus pompilius (Westermann and Schipp, 1998a) and O. vulgaris (Wells, 1978) but uptake into the epithelium does not provide evidence for transepithelial transport.

The caecum is a thin-walled muscular tube, coiled into about two and a half turns forming a cone and lined with lamellae which will increase the surface area. It is stated that the caecum filters food from the stomach “discarding gross particles that pass directly to the intestine” (Anadón, 2019, p. 4). Whilst this is a likely function, together with the addition of mucus to begin formation of faecal ropes, there are no direct studies in octopus demonstrating this but studies of the caecum in squid (Bidder, 1950) are supportive. Additionally, cilia have been to proposed to be responsible for the movement of “fine particles in suspension” from the caecum into the digestive gland (Fernández-Gago et al., 2019, p. 331) but no studies have investigated the relative roles of the cilia, caecal muscle constricting the cone or DG duct peristalsis in this process. The large surface area of the caecum is also consistent with an absorptive function, but Fernández-Gago et al. (2019, p. 331) concluded that in O. vulgaris that there was “no clear evidence for absorption in the caecum” although Boucaud-Camou et al. (1976) considered it was involved in absorption as is also the case in N. pompilius (Westermann and Schipp, 1998b). In the squid the caecum/caecal sac is proposed to have a major role in absorption (Boucher-Rodoni and Mangold, 1977; Boucaud-Camou and Boucher-Rodoni, 1983). Published studies used uptake of radiolabelled molecules by the epithelium or the presence of lipid containing vesicles in the epithelium but have not demonstrated transepithelial transport into the haemolymph from the caecum and this is therefore a major knowledge gap.

The control of the directionality of movement of contents between the common cavities of the crop, stomach and vestibule, caecum (and hence the DG duct) and proximal intestine has not been investigated. Evidence for the presence of sphincters is inconclusive or absent in octopus although there may be a flap-valve between the caecum and the digestive gland duct at the tip of the columella (Budelmann et al., 1997; see figure 64A, p. 197) and a thickening of circular muscle at the exit of the ink duct and anus is consistent with it acting as a sphincter (Young, 1967).

The tubular intestine shows little external variation in octopus throughout its course from the stomach vestibule/caecal junction to the anal opening, but turns anteriorly in the middle of its course. Internally, the mucosa has longitudinal folds, two of which on opposite sides are larger forming a “typhlosole” in the proximal intestine. It is assumed that the function of the typhlosole is to increase surface area and whilst this is a logical assumption the question of the reason why a surface area increase is required has not been answered; usually this is to facilitate absorption but it could also be related to secretory capacity (e.g., mucus). It is reported that the epithelium of the typhlosoles is ciliated but that of adjacent intestinal epithelial tissue has a luminal border with microvilli (Anadón, 2019; Fernández-Gago et al., 2019) suggesting a differentiation of function. Ciliary motion would be capable of moving mucus with trapped particles so investigation of the direction of ciliary beating in fresh specimens would be of particular interest. The function of the typhlosoles is not established, but constriction of the circular muscle would lead to the division of the lumen into two channels by apposition of the typhlosoles. A further issue regarding intestinal function in octopus is the extent to which it is involved in transepithelial transport of nutrients, ions and water which has not been investigated directly. In Loligo limited evidence was presented by Bidder (1950) for some absorption in the intestine while caecal absorption is in progress noting that the intestine may have a more major role when the gonads enlarge, limiting filling of the caecal sac. The capacity to absorb nutrients from contents bypassing the caecum/digestive gland but which have been exposed to digestive enzymes would be an advantage, as would involvement in water and ion homeostasis.

The anal opening in O. vulgaris has paired muscular anal flaps or wings. Their absence in species lacking an ink sac suggests involvement in inking rather than a digestive tract function but further investigation is required. The flaps are innervated by the atrio-rectal nerve branch of the visceral nerve and Young (1967, figure 18, p. 9) noted the presence of a plexus of fine fibres in the flap commenting “perhaps afferent.”

The faeces are formed as mucus coated ropes but their composition has not been studied in any detail and the mechanisms, presumed to be neural, controlling of defaecation are unknown (see Ponte et al., 2017 for discussion).

The toxicity of the exocrine secretions from the posterior salivary glands of O. vulgaris was first noted by Lo Bianco working at the Stazione Zoologica Naples, in the early C19^th but the toxic fraction was not identified until 1959 (Ghiretti, 1959). Two main constituents were identified with differing biological effects on crabs; amines inducing hyperexcitability and cephalotoxin, lethality. Cephalotoxin is a glycoprotein now known to occur in α and β forms (Cariello and Zanetti, 1977). The current view is that the saliva is a chemical cocktail including tachykinins (i.e. Eledoisin from Eledone, Erspamer and Anastasi, 1962), OctTK 1, phospholipase A2, other proteins, and enzymes (e.g., peptidase, hyaluronidase, chitinase) that also play key roles in toxicity of coleoid venom, supplemented by other small organic molecules, peptides, and non-enzymatic proteins (for review see: Ruder et al., 2013; Gonçalves and Costa, 2021).

Analysis of the phylogenetic history and molecular evolution of coleoid venom shows comparable diversity and complexity with those found in snakes (Ruder et al., 2013) with evidence of active evolution of some constituents (e.g., cysteine-rich pacifastin and kallikrein). Characterization of the molecular structure and composition of cephalotoxins is still incomplete (maybe except for Sepia esculenta cephalotoxin, Ueda et al., 2008). Apart from the glycosylation of the protein, characterization allowed the identification of conserved domains: EGF-like, Sushi, TSP type-1 and LDL-receptor class A (review in Gonçalves and Costa, 2021). Of particular interest is the epidermal growth factor (EGF) domain known to paralyse crustaceans via Na_v block; cephalotoxins, should be described as EGF neurotoxins. The increasing availability of cephalopod genomes has provided insights about tissue-specific genes expressed in the posterior salivary gland of three cephalopod species (non-tetrodotoxin-bearing octopods vs. tetrodotoxin in Hapalochlaena maculosa; Whitelaw et al., 2020).

Although cephalopod salivary gland research is focused on the nature of the toxins and their biological effects little is known about the physiological mechanisms involved in their secretion, how the ratio of the various constituents is determined, and the neural control of secretion together with the transport of saliva along the ducts. The discrete nature of the salivary glands makes them suitable for in vitro studies of intact glands and subsequently isolated cells comparable to those applied to mammalian salivary glands.

Knowledge of the physiological consequences of structural similarities and differences in the digestive tract between species (“functional morphology,” Mangold and Young, 1998) is essential to understanding how the animal adapts to its ecological niche. Mangold and Young (1998) and Voss (1988) utilised digestive tract morphology for systematics and revealing phylogenetic relationships. A recent review of the structure and function of the digestive system in molluscs makes more general comparisons (Lobo-da-Cunha, 2019).

Post mortem dissection of fresh specimens is the classical method for describing the gross anatomy. In fresh specimens, it is possible to view and measure the distribution of contents and classify their nature, i.e. liquid, solid and colour (Figure 2); note a description of changes in squid stomach contents with time in Wallace et al. (1981) and Boucher-Rodoni (1976) for caecal contents in Eledone cirrhosa. Care should be taken in interpretation as the method of killing may modify digestive tract motility (e.g., brain destruction will stimulate neurones; magnesium chloride relaxes smooth muscle; Bacq, 1934) and contractions may occur post mortem redistributing contents, particularly liquids. Guidance on dissection techniques is given in Øresand and Oxby’s (2021) illustrated guide to the dissection of the bobtail squid, Sepietta oweniana and by Guerra (2019) for several species.

Useful data can also be obtained from dissection of fixed specimens but distortion can occur and fixation of entire (larger) animals is a challenge. Freezing intact animals immediately post mortem preserves the tissue and maintains anatomical relationships.

Gross dissection provides largely descriptive, qualitative information on the morphology of the digestive system enabling comparison between species (see above). However, quantification of the wet weight of identified tissues provides data enabling further differentiation of species, assessment of the impact of different diets (e.g., natural vs. prepared/synthetic/elaborated) and investigation of potential correlations with environmental changes.

Most DS organ weight data is available for the DG; for example: six decapodiform cephalopods (Omura and Endo, 2016), O. vulgaris (García-Garrido et al., 2010; Bastos et al., 2020), Octopus maya and Octopus mimus (Linares et al., 2015; Pascual et al., 2019) with more limited data for the stomach and caecum (e.g., Omura and Endo, 2016). Detailed analysis of digestive gland weights has shown changes in the weight of the digestive gland in O. maya with season and location (Pascual et al., 2019).

Ideally, data on dry weight should also be obtained to differentiate between increase tissue mass compared to increased water content which may be pathophysiological (e.g., oedema).

Dissection is a destructive process so should only be used when there are no other methods available. For example, high-field magnetic resonance imaging and micro-computed tomography have been used for external and internal taxonomic description of a preserved specimen of a novel species of dumbo octopus (Grimpoteuthis imperator, Ziegler and Sagorny, 2021). The resolution of the digestive system was sufficient to identify all key features (e.g., presence of crop and “pancreatic” portion of the DG, paired hepatic ducts) and also to note that there was only a single posterior salivary gland and the anal flaps were absent (as was the ink sac). Importantly, collection of this type of imaging data enables construction of 3D models of the viscera so that the relationships of structures can be readily examined. High resolution ultrasound can also be used to image tissue in preserved animals but it has particular utility for anatomical and physiological studies in living animals so is discussed below.

The following are selected examples to illustrate the potential contribution of histological techniques to understanding physiology.

Motility describes the way the contraction and relaxation of the muscles of the digestive tract move the secretions (including salivary and DG ducts), the food whilst it is digested and absorbed, and any indigestible components, waste products, adherent mucus and shed epithelial cells.

Ingestion rate can be measured in experimental settings using live or frozen organisms, pieces of natural or processed (elaborated/synthetic) food and is calculated as the difference between the delivered food and that remaining for at least 2 h after being offered to the animals. The latter time can be established according to the feeding time regularly used in the laboratory depending on the species feeding habits. The percentage of leached nutrients must be measured and depends on the type of the diet. When processed food is used, leaching is measured using the shaking method (Obaldo et al., 2002). For this, 10 samples of 2 g of each diet are placed in 250-ml flasks placed in a horizontal shaker for 2 h (or the time that is estimated in each experimental condition to feed the animals) at the same temperature and water quality used experimentally. After that time, the water is filtered (pre-weighed, Whatman 1441-090 is recommended) to separate the remaining food from the leached water. Original food (N = 10) and leached samples of each diet are dried in a convection oven at 60°C for 48 h until constant weight, and then cooled in a desiccator. Dried feed samples are weighed and analysed for dry matter retention. Ingestion rate is calculated according to the following equation: Ingestion rate ( I ) = [ offered food - recovered food ] × [ 1 − ( nutrient leaching ) ] where offered and recovered food are expressed as dry weight (g) (60°C, 48 h), and unconsumed food as % of the nutrients lost during the shaking procedure (Obaldo et al., 2002). Ingestion rate (I) is expressed as g. ingested food (wet weight) considering the water content obtained after dry sampling at 60°C for 48 h. In the case of fresh or frozen food the same method is used. A review of the stomach contents and ingestion rate in cephalopods by Ibáñez et al. (Ibáñez et al., 2021) gives recommendations for obtaining data of ecological importance.

Digestion time differs between species, with temperature, dissolved oxygen and type of food, among other variables. For that reason, when evaluating the digestive processes, preliminary observations are required to determine the overall duration of digestion and hence sampling frequency. In O. vulgaris sensu stricto, O. vulgaris type II from Brazil, O. maya, and O. mimus fed fresh crab at 5% of the body weight it took ∼ 400 min to complete the digestion process (Martínez et al., 2012; Linares et al., 2015; Gallardo et al., 2017; Bastos et al., 2020). Depending on the number of animals available, it is possible to take timed samples during the process, but how many animals should be sampled to maintain a balance between 3Rs considerations and data quality? The answer depends on the type of analysis required, but usually a minimum of three animals will be necessary to obtain data that can be statistically analysed (Table 2). Also, it is essential to remember that digestive juice enzyme activity can be highly variable, especially with a mixed diet. For that reason, in this type of study it is highly recommended to use a single type of food to evaluate digestion timing (crustacea are arguably the best diet because their importance in trophic ecology of cephalopods).

Table 3 shows the variety of digestive enzymes in octopus species. Although there are probably more than a dozen types of enzyme with different roles, there is evidence suggesting that the acidic enzymes have higher activity during digestion (e.g., Ibarra-García et al., 2018). Based on Boucaud-Camou and Boucher-Rodoni (1983), Martínez et al. (2012) studied the pH of the “gastric juice” (i.e., the fluid found in the stomach but not secreted by the stomach) of O. maya during digestion of crab. They found that pH varied between 5.2 and 6, demonstrating, as was previously observed in O. sinensis (?) by Morisita (1972 a, b and c, cited by Boucaud-Camou and Boucher-Rodoni, 1983) that the main enzymes in the digestive tract of those octopus species were acidic proteinases. Until now, besides O. sinensis(?) and O. vulgaris sp. (Table 3) the presence of acidic proteinases has been demonstrated in the gastric juice of O. mimus, O. maya (Linares et al., 2015) and O. vulgaris type II (Bastos et al., 2020), and in the digestive gland of O. bimaculoides (Ibarra-García et al., 2018), Robsonella fontaniana (Pereda et al., 2009) and Enteroctopus megalociathus (Farías et al., 2010) indicating that the acidic proteinases may be the main type of digestive enzyme in octopus species. It is interesting to note that in some studies high activities of amylase are reported suggesting that some species (O. bimaculoides; O. vulgaris sensu stricto) may have the ability to digest complex carbohydrates (Table 3). Although interesting, these studies require duplication (Table 3) and the sources of complex carbohydrate identified.

It is important to evaluate the pH, temperature, substrate concentration and divalent ions that could specifically regulate digestive enzymes activity in each region of the tract (Martínez et al., 2011; Ibarra-García et al., 2018). Sampling requires the animal to be anaesthetized and killed (e.g., Roumbedakis et al., 2020). As soon as possible the digestive tract should be separated immediately into the crop, stomach, caecum and digestive gland using surgical clamps to avoid mixing the luminal contents. Once separated, the volume and weight of the chyme (the mixture of secretions [e.g., mucus and enzymes], partially digested food and ingested fluid) contained in each region should be measured, immediately frozen in liquid nitrogen, and stored until analysis (depending on the time until analysis, samples can be stored at −20, −40 or −80°C). A wealth of information on the digestive physiology of octopus can be obtained from this type of study particularly when combined with other techniques (e.g., in vitro studies of motility or preservation of tissue for IHC; see above); measurements are summarised in Figure 5.

In O. maya, O. mimus and O. vulgaris juveniles and adults, soluble nutrients (e.g., soluble protein, small peptides and amino acids) in the food move quickly to the DG and pass to the circulation becoming available for tissue metabolism, only 40–60 min after ingestion (Rosa et al., 2004; Linares et al., 2015).

The digestive physiology of octopus species and probably of other cephalopods can be divided into two phases: 1)soluble nutrients pre-digested in the prey (or elaborated food) quickly pass-through the crop, stomach and caecum to the DG and then the haemolymph to be transported to tissues; 2) more complex nutrients (e.g., myofibrillar proteins, complex lipids) are processed by the simultaneous action of the enzymes (from salivary glands and DG) when food is in the crop with mechanical degradation by concerted action of the crop and stomach, filtration in the caecum and finally entering the DG via the ducts as a suspension.

A recent histochemical study of the digestive tract of O. vulgaris detected no digestive enzyme presence in the digestive tract epithelium (Fernández-Gago et al., 2019) consistent with the view that chyme formation during the digestive process is predominantly by digestive enzymes secreted from DG with a minor contribution from salivary enzymes. This means that the changes occurring as food passes along the digestive tract are mainly because of the pulses of enzyme production in DG during the digestion phases previously described in cephalopods (Semmens, 2002; Martínez et al., 2011; Linares et al., 2015). The time to digest a meal depends on the species and ambient temperature. For example, in tropical and sub-tropical pre-adults of O. maya (26°C; 0.500 kg), O. mimus (14°C; 1.2 kg) (Linares et al., 2015) and O. vulgaris type II from Brazil (20.8°C) (Bastos et al., 2020) the complete digestive process takes 6–7 h, while in Octopus cyanea the entire digestive process was takes around 12 h at 30°C from the beginning of the meal (Boucher-Rodoni, 1973). Although digestion appears is a fast process in octopus species it is necessary to evaluate how temperature is modulating all the physiological, endocrinological and enzymatic mechanisms involved in the process to predict possible consequences in warming scenarios; a meta-analysis would be a useful first step.

Once the chyme is in the DG, biochemical process involving absorption and further digestion are activated (Boucaud-Camou and Boucher-Rodoni, 1983; Budelmann et al., 1997). In the DG of O. maya fasted for 24 h in laboratory conditions, “resting cells” characterised by few or no heterolysosomes or residual bodies are observed (Martínez et al., 2011). Additionally, an apocrine secretion and cell debris can be observed in the tubular lumen, probably resulting from catabolism in response to food deprivation for 24 h. During the first 1 and 2 h of the postprandial period (PP), digestive cells show nuclei forming a belt at the basal region of each columnar cell as the typical characteristic. Some heterolysosomes appeared inside the digestive cells. Acidophilic chyme in the tubule lumen indicates soluble nutrients in the DG. Once the digestive system in the DG is activated, heterophagosomes in the acinar cells transport amino acids to the haemolymph and then muscle and other tissues. In O. maya, O. mimus and O. vulgaris type II studies suggest that soluble protein is accumulated in muscle tissues initiating the glycosynthetic process (Gallardo et al., 2017; Bastos et al., 2020). After 4 h PP, the digestive cells store heterolysosomes (nutrients). Additionally, heterophagosomes were observed only near the apical zone of the cell, revealing the beginning of complex nutrient transport because of the mechanical digestion of more complex nutrients. Digestive cells showed a brush border (i.e., microvilli) on the apical membrane contacting the lumen. Acidophilic chyme was observed in the centre of the lumen. Peaks of glucose, cholesterol, acylglycerols and soluble protein can be observed with high digestive activity in DG cells. All these processes indicate that nutrients were transformed and transported to the tissues for energy and anabolic processes. At 6 h PP, the digestive cells have a conspicuous brush border and an empty lumen. Digestive cells appear filled with heterolysosomes, without heterophagosomes. At 8 h PP, a reduction in heterolysosomes was be observed, together with residual bodies, cell debris and apocrine secretion in the lumen. The proliferation of nuclei indicated the presence of replacement cells located in the basal lamina of the tubules signalling the end of the process (Boucaud-Camou and Boucher-Rodoni, 1983; Martínez et al., 2011; Gallardo et al., 2017; Fernández-Gago et al., 2019; Bastos et al., 2020).

Before describing the measurement of metabolic rate we first highlight one of the important parameters derived from its measurement. Many years ago it was recognized that the apparent heat increment (AHI) also termed specific dynamic action, (SDA) could be used as an indicator of the cost of mechanical and biochemical processes associated with the ingestion and assimilation of food (Nixon, 1969). Although muscular tissue is responsible for the mechanical activity, and the epithelium of the tract synthesises and secretes chitin and mucus requiring energy, the digestive gland is the main site of metabolic functions (Boucaud-Camou and Boucher-Rodoni, 1983). Hence, the AHI may result from addition of the energy used in the above processes; depending on the diet and temperature this constitutes a considerable percentage of the daily energy budget in aquatic organisms (Katsanevakis et al., 2005; Aguila et al., 2007; Noyola et al., 2013). In O. vulgaris values between 25 and 55% of food conversion efficiency were calculated. When different temperatures were compared it was observed that the peak value of AHI (SDA) of O. vulgaris was 64% at 20°C and 42% at 28°C (Katsanevakis et al., 2005) and in O. cyanea AHI was 60% of the ingested food (Nixon, 1969; Boucaud-Camou and Boucher-Rodoni, 1983). In O. maya it was also demonstrated that the AHI values are dependent on the type of diet with values that oscillate between 9 and 14% of the ingested energy when animals were fed fresh crabs or mixed diets made from fresh crab meat and other ingredients (Aguila et al., 2007; Rosas et al., 2007; Rosas et al., 2008). When temperature was investigated, the AHI values of O. maya were relatively higher in animals maintained in optimal temperature ranges (30–48% of routine metabolic rate; 22–26°C) than observed in animals maintained at 30°C (21%) suggesting that in higher temperatures the energy invested in mechanical and biochemical transformation of food is decreased due to additional costs associated with maintaining a higher metabolic rate at higher temperatures (Noyola et al., 2013). That means that depending on the species and its thermal tolerance temperatures beyond the optimal thermal range could affect the general physiological condition of the animals reducing the energy available to be invested in processing food. In larvae of coral reef, tropical fish it was observed that a temperature of 31.5°C does not affect the AHI (SDA) magnitude and duration suggesting that this species (Amphiprion percula) is well adapted to temperatures expected even in warming scenarios (McLeod and Clark, 2016).

There has been considerable debate about measurement of respiratory metabolism in aquatic animals in the light of new hypotheses related to global warming and physiological adaptation of ectotherms (Sokolova et al., 2012a; Sokolova et al., 2012b; Norin and Clark, 2016; Pörtner et al., 2017). Although the debate has focused on fish (Chabot et al., 2016a), there is a broad consensus on measurement of metabolism in aquatic organisms (Steffensen, 1989; Chabot et al., 2016b; Svendsen et al., 2016). Two types of respirometer have been used:

i) Flow-through respirometry. This measures oxygen consumption by quantifying the difference between inlet and outlet oxygen concentration and adjusting the flow of water through the respirometer to maintain a pre-set oxygen content difference. Oxygen consumption is then calculated as the product of water flow through the respirometer per unit time and the difference in oxygen concentration of the water entering and exiting the respirometer. In this respirometer the constant inflow of clean water reduces or eliminates the hypoxia, hypercapnia and nitrogenous waste issues associated with closed-system respirometry, but it introduces mixing and equilibration (washout) problems (Steffensen, 1989), although this can be corrected (Niimi, 1978).

The development of oxygen sensors to measure dissolved oxygen facilitated experimental work in this area. Although polarographic oxygen probes are still used, optical sensors are considered the best to use in physiological evaluations of metabolic rate of aquatic animals including cephalopods. Much of the data obtained in the last 50 years on octopus species used the home tank as a respirometric chamber and in the many cases as closed respirometers (Wells et al., 1983; O'Dor and Wells, 1987; Wells et al., 1996; Cerezo Valverde and Garcia Garcia, 2005; Katsanevakis et al., 2005). Although those data give useful information about the respiratory physiology of several octopus species, using optical sensors and improving the systems (i.e., Intermittent-flow or flow-through respirometry) will give more precise data as shown in O. vulgaris and O. maya (Cerezo Valverde and Garcia Garcia, 2004; Aguila et al., 2007; Noyola et al., 2013; Martínez et al., 2014; Meza-Buendía et al., 2021).

Data on oxygen consumption can be used to evaluate the energetic costs of different types of food (natural and elaborated) if the AHI is measured but also, if ingested food (I), faeces production (F), nitrogen excretion (N) and growth (P) are measured (all expressed in energy units; J day⁻¹ g⁻¹ or kg⁻¹), these values can be integrated into an energetic flow equation (Wells et al., 1996): I   =  F  +  N  +  P  +  R where assimilated energy (As) can be calculated as: As  =  R  +  P In this equation R = R_AHI + R_rut, with R_rut the metabolic rate of animals before feeding in respirometric chamber. Studies made in O. maya show stable R_rut values can be obtained in animals conditioned to respirometric chambers for 12–18 h, depending on temperature and the animal size (Roumbedakis et al., 2020; Meza-Buendía et al., 2021). When this type of data is obtained in animals fed different types of food it is possible to know how the nutritional characteristics of the food modulate the quantity of energy that is channelled to growth in comparison to mechanical and biochemical transformation of the ingested food. When pre-adult O. maya were fed Callinectes sapidus crab, R was 30% of As, but animals fed with elaborated food made from fish meal the R/As, % was between 80 and 90% indicating that a significantly proportion of the ingested energy was used to transform that diet in physiologically useful energy reducing the energy available for growth (Aguila et al., 2007). When an elaborated paste made from crab and squid was used to feed early juveniles of this species, a R/As value of 27%, was obtained. Analysis of R/A% and growth shows that lower values (∼30%) are associated with higher growth rates but diets with high R/A% (80–95%) are associated with lower growth as almost all the ingested energy is channelled to maintenance but not growth (Van Heukelem, 1976; Martínez et al., 2014).

The utility of molecular studies is well illustrated by studies investigating effects of the gastrointestinal parasite Aggregata octopiana. Changes in gene expression in tissue from discrete regions of the DS provides insights into physiology. For example, in the octopus gastric ganglion expression of selected genes correlated with increased relative expression of six genes in conditions of high infection (i.e., Sn, Nfkb2, Cckar, SCPRP, Serpinb10, Tlr3) while reduced expression was reported for others (e.g., Litaf, Mirp, Sod1, Prdx6, Gpx1, Rph3al) (Baldascino et al., 2017). The study demonstrated effects of the parasite on a neural structure at a site distant from the locus of infection in the intestinal wall. Transcriptomic analysis of haemocytes identified >500 differentially expressed transcripts in Aggregata infected octopus infected with Aggregata including genes involved in pathways such as Nuclear Factor-kB, Toll-like receptor, and Complement (e.g., Castellanos-Martínez et al., 2014) or transcriptional variations of diet, temperature, and growth-related genes in paralarvae (e.g., García-Fernández et al., 2019; Garcia De La Serrana et al., 2020).

Proteomics further characterised salivary glands and their secretions. For example, the Southern blue-ringed octopus (Hapalochlaena maculosa) is a remarkable exception with loss of proteinaceous toxins due to the presence of tetrodotoxin (Whitelaw et al., 2020). The gland has >600 genes exclusive to H. maculosa when compared to O. bimaculoides and Callistoctopus minor) providing an overall scenario of fewer and more specialized genes expressed by other octopod species, when compared to Hapalochlaena.

Although omics and molecular studies provide novel insights it is important that findings are translated where possible into understanding the physiology of the digestive system.

Relatively little is known about how the individual functions described above are controlled, coordinated with each other (e.g., the relationship between the crop, stomach and caecum and the secretory and absorptive phases of DG activity) and with the regulation of food intake. Here we briefly review selected studies to illustrate some of the applicable techniques.