Aerobic performance in tinamous is limited by their small heart. A novel hypothesis in the evolution of avian flight

Abstract

Some biomechanical studies from fossil specimens suggest that sustained flapping flight of birds could have appeared in their Mesozoic ancestors. We challenge this idea because a suitable musculoskeletal anatomy is not the only requirement for sustained flapping flight. We propose the "heart to fly" hypothesis that states that sustained flapping flight in modern birds required an enlargement of the heart for the aerobic performance of the flight muscles and test it experimentally by studying tinamous, the living birds with the smallest hearts. The small ventricular size of tinamous reduces cardiac output without limiting perfusion pressures, but when challenged to fly, the heart is unable to support aerobic metabolism (quick exhaustion, larger lactates and post-exercise oxygen consumption and compromised thermoregulation). At the same time, cardiac growth shows a crocodilian-like pattern and is correlated with differential gene expression in MAPK kinases. We integrate this physiological evidence in a new evolutionary scenario in which the ground-up, short and not sustained flapping flight displayed by tinamous represents an intermediate step in the evolution of the aerobic sustained flapping flight of modern birds.

Figure 1

Cardiac growth from embryonic age to juvenile or adult age in Red Junglefowl (Gallus gallus) in red, Chilean Tinamou (Nothoprocta perdicaria) in green, Ornate Tinamou (Nothoprocta ornata) in blue and American alligator (Alligator mississippiensis) in black. Power regression lines for each species were obtained after logarithmic transformation and Model II analysis (orthogonal regression) in Minitab 17. The power regression equation are as follows: Gg: VM = 0.0085 BM^0.892; Np: VM = 0.0186 BM^0.688; No: VM = 0.0207 BM^0.649; Am: VM = 0.0117 BM^0.737.

Figure 2

Comparative morphometry of the heart in adult specimens of Red Junglefowl (Gallus gallus, Gg), Chilean Tinamou (Nothoprocta perdicaria, Np) and Ornate Tinamou (Nothoprocta ornata, No). (a) Relative ventricular mass in males and females shown as the percentage of ventricular mass to body mass (VM:BM). (b) Mass of the right ventricle (RV) in males and females shown as the percentage of RV to VM. (c) Left and right ventricular wall thickness normalized to the diameter of the heart (WT:Ø_h) obtained from the ventricular section showing the attachment of an incipient right atrioventricular valve to the right ventricular free wall (see Material and Methods for details). (d) Normalized left and right ventricular wall thickness obtained as in panel C from ethanol preserved specimens of other tinamou species kept at the Colección Boliviana de Fauna at the Universidad Mayor de San Andrés in La Paz, Bolivia. Species nomenclature as in Table 1: Tm – Tinamus major (N = 2), Cu – Crypturellus undulatus (N = 2), Ct – Crypturellus tataupa (N = 1), Rr – Rhynchotus rufescens (N = 3), No – Nothoprocta ornata (N = 2), Npt – Nothoprocta pentlandii (N = 2), Nb – Nothura boraquira (N = 2), Nd – Nothura darwinii (N = 3). All data presented as mean and 95% confidence intervals with individual data points shown. N values as follows (in order from left to right in the different panels): a – 68,70,19,21,19,26; b – 21,17,8,8,9,15; c – 10,10,14,14,22,22. For statistical analysis for panels (a), (b) and (c) we used general linear modeling (GLM) considering species and gender or species and ventricle as factors followed by Tukey posthoc test with a customary fiduciary significant level of p < 0.05 (shown as dissimilar letters) in Minitab 17. No statistical analysis was performed for panel d.

Figure 3

Functional echocardiographic measurements of the heart in conscious birds placed in supine position under tonic immobility. Bantam chickens (Gallus gallus domesticus, Ggd) were used for comparison with the Ornate Tinamou (Nothoprocta ornata, No). (a) Left ventricular wall thickness normalized to the diameter of the heart (LWT:Ø_h) from a parasternal echocardiographic plane in which the right ventricular free wall is incipient but without a visible right ventricular chamber (see Material and Methods for details). (b) Fractional Shortening (%) of the cardiac muscle at the same plane. (c) Heart rate estimated from the time between subsequent peak systolic events in M-mode echo. All data presented as mean and 95% confidence intervals with individual data points shown (N = 5 for Ggd and N = 8 for No). Due to small sample size and an assumed lack of normality and homocedasticity of the data, permutation tests were used to test for differences between species using StatBoss (see Material and Methods for further details). No significant differences were found.

Figure 4

Functional measurements of cardiovascular variables in ketamine-xylazine anesthetized Red Junglefowl (Gallus gallus, Gg) and Ornate Tinamou (Nothoprocta ornata, No) before (Bsl – baseline) and after the administration of 3 ug kg⁻¹ of isoproterenol (Iso3). (a) Mean Arterial Pressure (MAP, mm Hg) measured from an intravascular catheter in the brachial artery. (b) Heart Rate calculated from the instantaneous pressure trace. (c) Mass specific total Cardiac Output (CO) estimated from a transit flow probe placed in the aortic arch after the splitting of the brachiocephalic arteries (see Material and Methods and Suppl. Figure 5 for details) and (d) Stroke Volume calculated from the quotient between CO and heart rate. All data presented as mean and 95% confidence intervals with individual data points shown (N = 5 for Gg and N = 4 for No). Due to small sample size and an assumed lack of normality and homocedasticity of the data, paired permutation tests were used to test for differences between species and for the effect of isoproterenol treatment using StatBoss (see Material and Methods for further details). A customary fiduciary significant level of p < 0.05 was used after compensation for multiple comparisons. Statistical differences between species but not due to treatment were seen only in panels (c) and (d) and are shown by dissimilar letters.

Figure 5

Metabolic measurements before and after a 3 min chase-and-exhaust protocol. ABC) Mass-specific oxygen consumption (VO₂) in adult individuals of (a) Red Junglefowl (Gallus gallus, Gg, N = 5), (b) Chilean Tinamou (Nothoprocta perdicaria, Np, N = 10) and (c) Ornate Tinamou (Nothoprocta ornata, No, N = 6). Open symbols show the data for the 60 min baseline measurements and closed symbols show the data for the 90 min following the chase protocol. (d) Excess post-exercise oxygen consumption (EPOC) obtained by integrating the pre- and post-curves shown as ABC. (e) Plasma lactate levels obtained using the same protocol in a separate group of individuals (Gg N = 6, Np N = 8, No N = 16). Data in A–C presented as mean and standard deviations. Data in D–E presented as mean and 95% confidence intervals with individual data points shown. For statistical analysis we used general linear modeling (GLM) considering species (D) and species/treatment (E) as factors followed by Tukey posthoc test with a customary fiduciary significant level of p < 0.05 (shown as dissimilar letters) in Minitab 17. No statistical analysis was performed for panels ABC because the integrated response is considered in panel (d).

Figure 6

Cloacal temperature before and after a 3 min chase-and-exhaust protocol in Ornate Tinamou kept at an ambient temperature of 4 °C (open blue symbols) and 25 °C (closed blue symbols). The chase-and-exhaust protocol was carried out after a baseline measurement lasting 1 h. For comparison, data on cloacal temperature in bantam chickens kept at 4 °C (open red symbols) that underwent the same protocol are shown. Data from chickens at 25 °C did not differ substantially and is shown in Suppl. Figure 2. All data presented as mean and standard deviations (N = 6 for both species). Paired permutation tests were used to compare baseline temperatures preceding the chase-and-exhaust protocols (120 min) with the temperatures 30 min after the test. Significant differences were observed only for the Ornate Tinamou (p = 0.03 in both cases) and are shown by “*” in the graph. “ns” indicate no significant difference.

Figure 7

Relative expression of the three main MAP kinase genes: ERK (a), p38 (b) and Jnk (c) and PI3K (d) in Red Junglefowl (Gallus gallus, Gg, N = 6), Chilean Tinamou (Nothoprocta perdicaria, Np, N = 12) and Ornate Tinamou (Nothoprocta ornata, No, N = 7). All values calculated in relation to the expression in Red Junglefowl after normalization against three housekeeping genes: GAPDH, β-actin and TBP. Dotted line indicates the reference expression level for Red Junglefowl. All data presented as mean and 95% confidence intervals with individual data points shown. For statistical analysis we used general linear modeling (GLM) considering species as a factor followed by Tukey posthoc test with a customary fiduciary significant level of p < 0.05 (shown as dissimilar letters) in Minitab 17.