Data bases were developed based on published reports. Among the search terms employed were the following:

This was supplemented by cross checking the cited papers in the top 15 papers identified by the initial search.

Cell volume (liver and small intestine mucosa) and cell number were calculated from the published hepatic DNA concentrations and the mass of DNA in female chickens (Mendonça et al., 2010, 2016). The number of liver cells per gram was calculated as follows: DNA concentration (mg per gram)/diploid genome mass in mg. This was simplified as DNA concentration as mg per gram × 10⁹/diploid genome mass in pg. We employed a mass of 2.15 pg for chicken diploid genome (Mendonça et al., 2010, 2016). The mass of cells was calculated as 1/number of cells per gram in pg. The volume was assumed to be numerically equal to the mass × 10⁻³. The calculations assumed that the tissue has a density of 1.0 kg L⁻¹.

Sexually mature female birds were those that were laying eggs either in domesticated species or during the breeding season for wild birds. Data for sexually immature birds included in the analysis were restricted to birds that were over 75% of mature body weight and were neither laying eggs nor incubating eggs.

Estimation of the gastro-intestinal absorptive area entailed the following:

Data (the mean from the individual studies) were analyzed by paired and unpaired Students t tests or by one-way ANOVA followed by Dunnett’s range test employing Excel software (Microsoft Excel 2019 for Mac).

Table 1 summarizes fluid dynamics in laying hens (Gallus gallus) and mallards (Anas platyrhynchos; or domesticated ducks—Anas platyrhynchos). Extracellular water, plasma, and interstitial fluid, expressed as a percent of body weight are similar in laying hens and mallards (Table 1). There appear to be differences in laying hens and mallards for intracellular water and total water expressed as a percentage of body weight.

The CO (ml kg⁻¹ min⁻¹) in laying hens (Table 1) is also very similar to that in young (broiler) chickens 192 ± (5) 5.6 ml kg⁻¹ min⁻¹ (for means of individual studies see Supplementary Table 1 in Supplementary Data). The CO is markedly greater in mallards than laying hens (Table 1).

Fluid dynamics in laying hens and mallards (Table 1) are key information for the development of PBK models. Cardiac output is decreased in starved chickens (Vogel and Sturkie, 1963). It is, therefore, possible that chemical agents that decrease voluntary feed intake will also depress cardiac output.

Both the Hct/PCV and hemoglobin of mallards (Table 1) are similar to those across avian species [Hct/PCV 44.0% based 158 species of birds; hemoglobin 14.5 g dL⁻¹ based 134 species of birds; Scanes, 2022]. In contrast, both the Hct/PCV of laying hens (Tables 1, 2) are markedly lower than that in mallards and the mean across avian species. Blood concentrations of hemoglobin are similarly low in reptiles [Crocodilia: 8.23 g dL⁻¹ (three species); Dessauer, 1970, 7.76 g dL⁻¹ (five species); Pough, 1980; Squamata: 7.77 g dL⁻¹; Dessauer, 1970, 8.8 g dL⁻¹ (Pough, 1980); and Testudines: 7.97 g dL⁻¹; Dessauer, 1970, 8.5 g dL⁻¹; Pough, 1980]. The last common ancestor being the following:

Laying hens were found to exhibit a lower Hct/PCV and hemoglobin than mallards (Table 1). This reduction in the Hct/PCV and hemoglobin concentrations would reduce the number of available binding sites for chemical agents, a finding that is likely to impact PBK modeling.

Serum/plasma concentrations of total protein and albumin are very similar in laying hens (Table 2), in mallards and across birds (total protein concentration: 39.6 across 100 avian species; 15.9 across 63 avian species; Scanes, 2022). The increase in total protein in the serum/plasma in mallards between reproductively quiescent and laying females is presumed to reflect the appearance of vitellogenin in the circulation of laying ducks.

There are at least three binding sites on human albumin, respectively, the following:

It is reasonable to assume that avian albumin has analogous binding sites. The presence of such binding sites is likely to impact PBK modelling as is the concentration of albumin in plasma/serum (Sjöholm et al., 1979; Sethi et al., 2016; also see review: Yang et al., 2014).

Plasma/serum concentration of albumin in laying hens (Table 2) and laying female mallards (discussed above) are lower than in mammals [3.59 ± (39) 0.09 g dL⁻¹; Scanes, unpublished]. This has implications for PBK modeling, as there are likely to be increases in the fraction unbound of compound for specific drugs or pesticides (Sethi et al., 2016). For instance, the percentage of unbound compound was much higher in the plasma of human neonates (diazepine: 15.4%, cyclosporine 20.0%, and deltamethrin 26.9%) than in adults (diazepine: 4.1%, cyclosporine 4.3%, and deltamethrin 10.0%); there being higher plasma concentrations of albumin in adults (2.8 g dL⁻¹) vs. neonates (4.0 g dL⁻¹).

There is a strong case for determining the unbound fraction for chemical agents in the following: (1) hen plasma, (2) immature female chicken plasma, (3) purified chicken serum albumin, and (4) vitellogenin. It is suggested that studies could employ heparinized blood from laying hens and young female chickens to determine in distribution between plasma and erythrocytes and within plasma the free unbound.

It is interesting that there is bio-concentration of vitellogenin, α₂-glycoprotein/β-livetin, triglyceride, and phospholipids into yolk but not of albumin or immunoglobulin compared to plasma (Table 3). There is also bio-concentration of α₂-glycoprotein/β-livetin and immunoglobulin compared to interstitial fluid (Table 3). The basis for the bioconcentration is receptor-mediated transport. A specific receptor has been identified that mediates transfer of both vitellogenin (Vt) and very-low-density lipoprotein (VLDL), including triglyceride and phospholipid, across the oocyte plasma membrane to fill the oocyte with yolk (chickens: Stifani et al., 1990; Barber et al., 1991; Bujo et al., 1994). Furthermore, the receptor binds α₂-glycoprotein (Jacobsen et al., 1995). There are, in addition, other receptors mediating transfer into the oocyte. What is not known is the extent to which chemical agents are transferred into the oocyte bound (or loosely associated) with Vt or VLDL and thereby pass into the oocyte by receptor-mediated transport. This would not be consistent with a simplistic view in which only “fraction unbound” chemical agents can pass into cells by passive diffusion.

Chemical agents can bind to yolk precursors in the blood and, thereby, be deposited into the oocyte (yolk). In older literature, radiolanthanum has been demonstrated to bind to yolk proteins (granules and specifically phosvitin) both in vivo and in vitro (Robinson et al., 1979). Deposition into the yolk has been examined in Japanese quail. Radiolanthanum and other radiolanthanides were administered intravenously (Robinson et al., 1980). After 18 h, 15% of the dose was located in the oocytes with ~90% of the dose being located in the eggs after 10 days (Robinson et al., 1978, 1980). Similar results were reported for other radiolanthanides (Robinson et al., 1980). This is consistent with radiolanthanides binding to the phosvitin component of vitellogenin and then receptor mediated transport across the oocyte membrane.

As might be expected, ovary and oviduct masses were much greater in laying hens than in sexually immature chickens (Table 4). In contrast, there were no differences in the relative weights of the gizzard, heart, kidneys, liver, small intestine, and spleen between those in sexually immature chickens and in laying hens (Table 4). However, as female chickens come into lay there is a 29.2% increase in body weight (Hurwitz and Bar, 1971). Thus, conclusions based on relative weights of the gizzard, kidneys, liver, and spleen (Table 4) are confounded by the increase in body weights.

Like its mammalian counterparts, the avian liver is critically important for controlling metabolism and, for many toxicants, detoxification (Zaefarian et al., 2019). The avian liver receives blood via both the hepatic artery and the hepatic portal vein (Zaefarian et al., 2019). The liver produces bile. In Galliform and Anseriform birds such as chickens and mallard ducks, the major bile acid is chenodeoxycholic acid (Hagey et al., 2010a) while in mammals, the major bile acids are cholic acid together with chenodeoxycholic acid (Hagey et al., 2010b). In addition, 23 and 15α-hydroxylated forms are found in both birds and mammals (Hagey et al., 2010b). Chemical agents can be eliminated in the bile but there is recycling with a chemical agent or its metabolite passing from the blood through the liver into the bile. The bile passes to the small intestine where the chemical agent or its metabolite is (re-)absorbed into the blood. It is suggested that the avian liver affects detoxification and excretion of chemical agents in a manner similar to that in mammals and that PBK modelling can use parameters developed for the mammalian liver.

There are some differences between the avian and mammalian kidneys. A renal portal system is present in birds including chickens and mallards but not mammals (reviewed: Akester, 1967). Closure of the left and/or right renal portal valves diverts blood from the left/right external iliac veins to renal portal veins in chickens, mallards, and other avian species (Akester, 1967). Blood flow can also be diverted from the kidney to the central circulation (vena cava; Burrows et al., 1983). Blood flow in the coccygeo-mesenteric vein can flow in either direction allowing blood flow into the hepatic portal system (Akester, 1967). The valves in the renal portal system are controlled by the sympathetic nervous system (Burrows et al., 1983).

It is reasonable to conclude that there is renal intestinal recycling of toxicants in mallards and chickens. Recycling encompasses a chemical agent or its metabolite passing from the blood, filtered into nephrons, and hence into the urine. This passes from the cloaca to the colon and caeca by retrograde peristalsis and then is absorbed into the blood stream.

There are marked similarities between the gastrointestinal tract between mammals and the laying hen or mallards with an esophagus, the equivalent of a stomach, a small intestine, and a colon. There are several major differences between the anatomy of the gastrointestinal tract of birds compared to mammals including the presence of the following: (1) A sac-like organ off the esophagus—the crop; (2) A separate glandular and muscular stomach, namely, the proventriculus and gizzard; (3) Two ceca; and (4) The absence of teeth.

Absorption and reabsorption of nutrients or chemical agents predominantly occur in the small intestine. In the present analysis, the absorptive areas in chickens and mallard ducks were very similar (Table 6). Moreover, they were with the range for mammals (Table 6). This suggests that the characteristics of absorption of chemical agents in these avian species is similar to that in mammals, although the specificity and kinetics of active transport in birds needs to be included in any analysis.

The lack of a difference between jejunal villus height in young broiler chickens and adult laying hens (Table 5) is in contrast to a relationship, albeit modest (R² 0.24) between log villus height and log body weight across multiple avian species (Ricklefs, 1996).

Absorption and reabsorption of chemical agents in the urine may also occur in the colon and ceca due to retrograde flow and colonic anti-peristalsis (Lai and Duke, 1978; Duke, 1989). It has been demonstrated that drugs can be administered as an enema in birds (day-old chicks: Marietto-Gonçalves). Moreover, rehydration can be accomplished by rectal administration of a hypotonic solution (pigeons: Ephrati and Lumeij, 1997) supporting both retrograde peristalsis and absorption in the colon and perhaps also the ceca.

There is broadly both very similar numerical values between the preset study together with complementarity between the present analyses and those of Wang et al. (2020) with multiple parameters reported in one but the other (Tables 7–9). The former was despite the greater number of studies employed in the present analysis (Table 7). An exception to this is the markedly greater relative weights of the gizzard and small intestine in the present report and that of Wang et al. (2020) (Table 7). This difference may reflect the greater number of studies employed in the present analysis together with the extremely low relative organ weights in the study of Martínez et al. (2015). Moreover, in the present meta-analysis, there were similar relative weights of the gizzard and small intestine in sexually immature females and laying hens (Table 4; Wang et al., 2020).

Turning to the issue of hematocrits, there was only a small difference in hematocrits (Table 7). There were considerably more replicate studies employed in the present analysis (10 studies) compared to Wang et al. (2020) (five studies). However, this difference is unlikely to influence PBK analysis as the fraction unbound is present in the plasma not the erythrocytes.

Blood flow to the ovary reported in the present report differed from that in Wang et al. (2020). This would indicate a lower reliability of either estimate with this being supported by the high coefficient of variation (Table 8). Moreover, a number of studies were excluded from the present analysis because blood flow to ovarian follicles was only expressed relative to follicular tissue weight (e.g., Hrabia et al., 2005; Rząsa et al., 2008). Unfortunately, these could not be included in the analyses.

Physiological and anatomical data on laying hens and mallards/domestic ducks (Anas platyrhynchos) have been rigorously compiled and subjected to analysis. As a comprehensive compendium, these data provide one of the critical bases for the development of PBK models for laying hens and mallards.