PLoS Computational Biology
Public Library of Science
Tracheal branching in ants is area-decreasing, violating a central assumption of network transport models
Volume: 16 , Issue: 4
Doi: 10.1371/journal.pcbi.1007853
•
•
• Altmetric

### Notes

Abstract

A fundamental assumption of models of the transport of substances through networks of tubes, such as circulatory systems in animals and vascular systems in plants, is that the total cross-sectional area of the tubes remains constant irrespective of the branching level, or that it increases slightly in the direction from the largest to the smallest tubes. One large tube should have the same or a slightly smaller area than the sum of the next two tubes after a branching. The assumption of such a pattern underpins one of biology’s most influential ideas–the metabolic theory of ecology. Surprisingly, the assumption has never been systematically examined for insects–the planet’s most diverse group of animals which deliver oxygen to and remove carbon dioxide from their bodies using a network of tubes known as tracheae. Until recently, it has been technologically very challenging to do so. Here, we use x-ray synchrotron tomography to overcome this challenge. We show that tracheal branching in 20 species of ants does not follow this pattern. Rather, cross-sectional area reduces in an inwards direction. We then use modelling to show that such a pattern facilitates outward CO2 release, a process more challenging for insects than moving oxygen inwards. Our work suggests that much still needs to be done to understand the fundamental assumptions underlying network transport models and how they apply more generally across life–especially in the context of why metabolic rate scales with body size.

Aitkenhead, Duffy, Devendran, Kearney, Neild, Chown, and Savage: Tracheal branching in ants is area-decreasing, violating a central assumption of network transport models

## Introduction

Mechanistic explanations for the scaling of energy and mass flows with organism mass have proliferated in the last two decades. The long-term goal has been to find a general mechanism to explain the tendency for quarter-power scaling of metabolic rate as well as departures from such scaling, which would provide a powerful theoretical basis for inferring how energy and mass flow at scales below and above the individual [14]. A range of explanations has been proposed for both inter-specific and intra-specific scaling patterns, including quantitative theories incorporating the role of surface-area/volume dynamics of storage pools and the overhead costs of growth, and arguments about the relative roles of supply and demand processes in the context of metabolic activity levels [1,59].

The most controversial and high-profile ideas in recent times are transport network-based models [1,7,1012]. These models make a series of assumptions about the physical characteristics of transport networks (e.g. animal circulatory and plant vascular systems), with the overall assumption that they constrain resource delivery and hence metabolism [10,12]. An advantage of such physically-explicit models is that they make testable predictions about underlying processes and structures [9]. Perhaps for this reason, they have generated substantial debate. A wide range of works has examined empirical evidence for their assumptions and predictions, and the theory underpinning them, noting that these models do not apply in all circumstances, and often conclude that other approaches have greater support [1318].

The most influential of the transport network-based approaches, the metabolic theory of ecology (MTE) [1,2], assumes that in the branching of plant vessels and insect tracheae (the main transport route for O2 to insect tissues and for CO2 efflux from the body [19]), the total cross-sectional area remains constant across all levels (Da Vinci’s rule). In mammals this pattern applies to large vessels, with a transition to an increase in total cross-sectional area of smaller vessels (also known as Murray’s law) [1,18]. The summed cross-sectional area of the child branches is either the same as or larger than the cross-sectional area of the parent branch (Fig 1). Support for these assumptions and the consequences of their relaxation have been well explored for mammals and plants [10,12,2022].

Fig 1
The metabolic theory of ecology assumes that in branching of plant vessels and insect tracheae, cross-sectional area is preserved across branching levels (DaVinci’s Rule) [1]. In mammals, large vessels are area-preserving with a transition to area-increasing (Murray’s Law) as vessel size decreases [1]. By contrast, Nunome [32] proposed that insect tracheae may be area-reducing (Nunome’s Pattern).Diagrammatic representation of transport-network branching following each of the three main branching schematics.

What the situation is for insects remains poorly known. The MTE assumes that in insects tracheal branching is area-preserving and gas exchange is dominated by diffusion [1,23]. Many studies have investigated insect gas exchange and demonstrated convection is common and can dominate gas exchange [24,25]. By contrast, the empirical evidence on which the area-preserving assumption of the MTE is based comes from 100-year old observations of a single caterpillar [26] (S1 Text). Only a few other investigations of the relationship between the radii of parent and child tracheal branches have been undertaken, but they are not widely appreciated. Therefore, the general principle of the way in which network architecture contributes to gas exchange in the Earth’s largest group of terrestrial animals remains untested [27].

We examined this fundamental assumption about insect tracheal network structure by conducting a review of the limited available information and then complementing it with a systematic investigation of cross-sectional area change during abdominal tracheal branching in 165 individuals of 20 species of ants using synchrotron x-ray microtomography [28,29]. Specifically, we tested the prediction that in the equation relating parent to child branches:

${r}_{i}^{\alpha }={r}_{ja}^{\alpha }+{r}_{jb}^{\alpha }$
where ri and rj are the radii of the parent and child vessels, respectively [21], α should be ~2 for area-preserving branching or ~3 for area-increasing branching, as assumed by the MTE and other theoretical studies [1,23], and acknowledging that variation may be continuous between these values in some circumstances. We also determined whether relationships between parent and child branches changed in concert with changes in radius size of the primary level tracheae across all of the species examined, by regressing the left- and right-hand sides of Eq 1 against each other. Finally, we modelled CO2 transport in the system, determining what the consequences are of different values of α, because tracheal systems are typically confronted by differences in the ease of transport of CO2 relative to transport of O2 [30,31].

## Results

### Tracheal architecture

The ant tracheal system was readily reconstructed from the tomography, revealing the ubiquitous presence of air sacs across the species investigated (Fig 2). Solving for α resulted in a median of α = 1.01 for level 1 to 2, and α = 1.01 or 1.03 for level 2 to 3. Data were typically right-skewed (Skewness for Level 1–2: 2.03; L2-3: 0.43; L2-3b: 2.41). After log10 transformation to account for the skew [21], α differed statistically from the expected values of 0.30 (log of 2) or 0.48 (log of 3) (Level 1–2: 0.3: t = -34.7, df = 164, 0.48: t = -58.0, df = 164; L2-3: 0.3: t = - 41.2, df = 157; 0.48: t = -65.3, df = 157; L2-3b: 0.3: t = - 25.6, df = 151; 0.48: t = -42.3, df = 151; p < 0.0001 in all cases), but was either identical or close to 0 (95% C.I.: Level 1–2: 0.02 to 0.05; L2-3: -0.02 to 0.01; L2-3b: 0.0 to 0.05), which accords with an untransformed α = 1.

Fig 2
Results of the reconstruction of x-ray tomography of a Camponotus suffusus minor (a-c), and Myrmecia fulvipes (d-f). (a, d) Volume render of the body and internal airspaces (cyan). (b, e) Horizontal planar slice of the abdomen with air sacs (AS) and tracheal branches (white vertical line) indicated, the mounting syringe (S) is visible around the edges of the images. Additional examples of the raw slices for other ant genera can be found in S3 Fig. (c, f) Volume render of the internal airspace showing the air sac and associated trachea branches, coloured by their respective level as in Fig 1 (blue, green, red for tracheal levels 1, 2, 3, respectively, as in Fig 1). Note perspective distortions arising from presenting a three-dimensional volume in two dimensions. The radii of each branch shown for C. suffusus (c) were measured as 29.15 μm (blue), 13.87, 10.11, 5.59 μm (green), and 6.35, 8.02, 5.40, 5.24 μm (red). The radii of each branch shown for M. fulvipes (f) were measured as 32.83 μm (blue), 17.67, 14.44 μm (green), and 11.78, 8.13, 6.19, 6.61 μm (red).

The regression analyses revealed no difference with size, and strong linear relationships, though these did not accord with the assumptions of α = 2 or 3, but rather with the expectation that α = 1 across the size spectrum [see Materials and Methods] (body length range among individuals: 4–24 mm) (Fig 3A and 3B and S1 Table). We also tested for a species-level effect, but found that it was small, resulting in slopes that were generally indistinguishable from those in the OLS regressions (S2 Table).

Fig 3
(a) Empirical outcomes (slopes ± standard error) of Ordinary Least Squares (OLS) regression forced through the intercept (sum of radii of child branchesα ~ radius of parent branchα where α = 1, 2 or 3 for Nunome’s, DaVinci’s or Murray’s laws, respectively; left hand points, solid lines) and Major Axis (Model II) regression (right hand points, dashed lines) for levels 1 to 2, and 2 to 3 (full results in S1 Table). Shaded boxes indicate upper and lower confidence intervals of each model fit and error bars indicate the standard error of the forced intercept models. A slope of approximately 1 is expected if the relationship fits observed data. (b) Relationship between parent and child tracheal branch radii for levels 1–2 (green dots) and 2–3 (blue dots), with the black fitted line indicating an ordinary least squares slope of ~1 (α = 1), whereas the yellow and orange lines indicate expectations for relationships with area-conserving (α = 2) and area-increasing (α = 3) branching, respectively. (c) Modelled whole-organism CO2 flux at each tracheal level (level 5 is tissue level, level 1 is spiracle) for area-conserving (yellow line), area-increasing (orange line), or area-reducing (black lines) transport networks. Each of the three black lines represents a model system with either no air sac (continuous solid line) or an air sac at either the 3rd or 4th level (indicated by the dashed line-segment), after which hydraulic resistance decreases and efflux is projected rather than modelled explicitly. Tracheal branch lengths of the modelled system ranged from 211 μm at the deepest level (level 5) to 1569 μm at level 1 (levels 4, 3, 2 had branch lengths of 352, 586, and 976 μm, respectively). The tracheal radius of level 5 was set at 3 μm. Whole-organism flux was modelled with a partial pressure difference of 6 kPa (see S1 Fig for 4 kPa and 8 kPa) and assuming 6 pairs of abdominal spiracles. Horizontal grey lines indicate the flux needed to meet the metabolic requirements of an average ant at rest or walking (S5 Table). Where flux values are below these activity lines, flux of the system is inferred to be insufficient to maintain activity by diffusion alone.Predictions and empirical data for the relationship between parent and child branch radii and their modelled consequences for CO2 efflux.

Therefore, in the ant abdominal tracheal systems examined here, total cross-sectional area is not preserved, nor does it increase, from parent to child branches. Rather, cross-sectional area declines: the summed radii of branches at the lower level equals the radius of the branch at the higher level. We call this ‘Nunome’s pattern’, after early work referring to this form of area change with branching [32].

### CO2 transport modelling

Our modelling (S2 Text) uses tracheal network data from the current investigations and partial pressure variation data from the literature [31,33]. Furthermore, we assume that gas exchange is dominated by diffusion in ants at rest, in accordance with empirical data [34] and the assumptions of the MTE [1]. The modelling outcomes indicate that tracheal architecture with an inwards decline in cross-sectional area with branching (which means area-increasing branching outwards), delivers the highest potential CO2 flux. This CO2 flux is in keeping with empirical data on ant metabolic requirements (Fig 3C and S1 Fig). By contrast, area-preserving branching and Murray’s law cannot deliver the required CO2 efflux.

A model of convection presuming that convection is periodic, as is true of many insects [24], indicates that an additional 0.241 μl h-1 of CO2 flux per spiracle can be achieved when the network is characterised by inwards area-reducing branching, under the assumption of 66% volume evacuation at a frequency of 0.26 Hz (tracheal compression parameters based on empirical observations [35]). This value contributes to the background flux provided by the diffusive system (Fig 3C) and, as flux from this model scales linearly with evacuation frequency, flux gains through convection are expected to increase as needed as individuals become more active.

## Discussion

Our findings contrast strongly with those of Krogh [26], the foundation on which contemporary network models of metabolic ecology are based [1], that cross-sectional area is conserved (S1 Text). They are, however, in keeping with overlooked studies on an adult dragonfly [36], in part also with data on a water bug and silkworm [32,37], and with recent data for tracheae of Drosophila melanogaster [38]. For the 30 species of insects that have been examined for cross-sectional area variation among tracheal branch levels, in 22 of them total cross-sectional area of measured tracheae declines with branching, in six cross-sectional area is preserved, and in two species the pattern is variable (S3 Table).

Current theory dictates that either area-preserving or area-increasing branching results in the required resource delivery by tubular transport networks [1,12,18]. Yet, the ants examined here and some other insects have evolved an entirely different branching pattern in their tracheal system. In consequence, such a transport network should provide a physiological advantage proximally, and ultimately a fitness benefit. What that advantage might be is not immediately obvious given long-standing statements that any system capable of delivering an oxygen supply to the tissues should be adequate for the removal of CO2 [19]. Those statements are typically predicated, however, on the basis of tissue solubility differences between the two gasses, rather than also on the way they are transported to and from metabolically active areas through the tracheal system. Oxygen is generally delivered from the spiracles along the tracheae to the finest tracheole branches which lie within a few micrometres of the mitochondria where it is required [19,31]. By contrast, CO2 accumulates within the insect and is buffered in the haemolymph. It then moves across the tracheal walls back into the tracheal lumina across the full length of the tracheal system, rather than just at the finest tracheole branches [30,39]. As a consequence of this situation, and partial pressure differences between intratracheal and external O2 and CO2, respectively, tracheal systems are confronted by too little transport of CO2 relative to transport of O2 [30,33], and are sensitive to much smaller experimental changes in the partial pressure of CO2 than of O2 [31]. In consequence, tracheal systems are likely to be optimised for CO2 transport outwards.

Our model of diffusion (S2 Text) reveals why optimisation for outward CO2 transport is likely the case. Nunome’s pattern delivers the highest potential CO2 flux: enough to meet the metabolic demands of individual ants. Area-preserving branching and Murray’s law either are less adequate, or entirely unable to do so, respectively. Thus, they seem less likely to be found in insects, especially in the case of Murray’s law, because such branching patterns cannot deliver the required CO2 efflux, accounting for the empirical data demonstrating that Murray’s law is rare in insects (S3 Table).

The ant tracheal system includes air sacs, as both our investigations (Fig 2) and others [27] have revealed. They are likely to be involved in some form of convective gas exchange as has been demonstrated for many other insects [24,25], but are not explicitly included in the network geometry considered by the MTE. Our model of convection indicates that considerable additional CO2 flux per spiracle can be achieved when the network is characterised by Nunome’s pattern, with further potential increases associated with increasing organismal activity. Where external CO2 concentrations are high, such as in underground nests or other locations with CO2 accumulation [40,41], Nunome’s pattern may be the only architecture that enables sufficient gas exchange (Fig 3C). Under the most extreme circumstances, however, ants are compelled to switch to anaerobic metabolism [42].

Although we recognise that gas flux in the tracheoles might have greater complexity than we have indicated here [43], our modelling and empirical data provide an initial demonstration of why tracheal network architecture in ants follows Nunome’s pattern. Our data for the more proximal tracheal branches also show that such branching patterns are not simply restricted to more distal elements of the system as suggested recently [38]. They demonstrate clearly why, in insects, tracheal architecture is unlikely to conform to Murray’s law. That is, the system has evolved to meet the demands of both O2 influx and CO2 efflux. The ca. 133 million-year separation between the ants measured here and their last known common ancestor [44] (S2 Fig) suggests that branching in accordance with Nunome’s pattern may be ubiquitous across all ant species. Further empirical investigations, which are now within relatively easy reach given the availability of x-ray microtomography [28,29], coupled with additional modelling, will reveal whether Nunome’s pattern is most typical for insects which encounter high CO2 environments, while Da Vinci’s rule or some more complex arrangement [38] dominates in others. Irrespective, two major empirical assumptions of the metabolic theory of ecology, that branching is either area-preserving or area-increasing, and that gas exchange is predominantly by diffusion, do not apply routinely to insects, the most diverse group of animals globally.

What are the implications of our findings for the mechanistic role of branching networks in theories of metabolic scaling? It has been raised previously that the generality of the mechanistic role of branching networks in such theories is limited by the absence of closed circulatory systems in many groups [45]. Until the implications of varying architecture of branching systems, like that described here, are explored further, the generality of network transport-based models will remain further reduced. Explorations for some aspects of architectural variation have already begun [1012]. The findings presented here indicate that broadening this exploration to encompass area-reducing architecture is an intuitive next step as network-based transport models, and the MTE in particular, continue to explain a wide range of ecological phenomena [46,47]. Examining tracheal architecture in ants in concert with metabolic rate scaling and cell size variation would likely prove especially beneficial given previous outcomes from investigations of the latter [14].

## Materials and methods

We used synchrotron x-ray microtomography [28,29] to measure the cross-sectional area of the abdominal tracheae of 165 individuals from 20 species of ants. We use ants as the focal taxon because: (i) the characteristics of gas exchange and metabolic scaling have been relatively well characterized for them, including demonstration of a predominance of diffusion at rest, (ii) substantial size variation is found both within and among species, and (iii) because they are an ecologically important group of insects [14,34,48,49].

### Collection and initial treatment

Individual ants of twenty species in the genera Camponotus (8 spp.), Rhytidoponera (6 spp.) Myrmecia (4 spp.), Polyrachis (1 sp.) and Iridomyrmex (1 sp.) were collected from sites in Victoria and Queensland, Australia. Samples from Victoria were kept alive in 100 ml vials on moistened plaster of Paris substrate until their return to the laboratory. Ants were kept in these vials for up to four days at 20°C in a Panasonic MLR-353H-PE climate chamber (Panasonic Healthcare Co., Ltd, Sakata) with a 12:12 L:D cycle. They were then exposed to iodine vapour for 7 days to improve contrast for imaging [50], which resulted in death. Ants from Queensland were exposed to iodine vapour within 8 hours of collection such that preserved specimens could be returned to our laboratory without compromising inter-state biosecurity protocols relating to transport of live insects. Species were identified post mortem using relevant keys [51,52], with subsequent specialist verification (S4 Table). Up to 12 ants were loaded into customised 10 ml plastic syringes in preparation for synchrotron x-ray tomography at the Australian Synchrotron.

### Tracheal imaging and measurements

Synchrotron x-ray image data were collected on hutch-B of the imaging and medical beamline (IMBL) at the Australian Synchrotron. The IMBL custom-designed ‘Ruby’ detector, comprising a PCO.edge sensor equipped with a Nikon Micro-Nikkor 105 mm/f 2.8 macro lens and a scintillator with a 200 μm terbium-doped gadolinium oxy-sulphide screen [53], was used for imaging. The distance between sample and detector was 50 cm. Each syringe containing ants was mounted between the source and detector and scanned at an energy of 35 KeV. X-ray projections were taken every 0.1˚ over an 180˚ rotation. Individuals were scanned a single time only, with scans lasting from 3–6 minutes per individual. Resolution of the images (7.6 μm) was determined by measuring an object of known size (16.05 mm), measured with Mitutoyo-Absolute Series 500 digital callipers (instrument error ± 0.02 mm; Mitutoyo, Kawasagi).

Raw tomography data were reconstructed using the XLI-CT workflow V0.9.6 software (CSIRO, Clayton, https://www.ts-imaging.net/), where vertical stacking was required the overlap was 110 pixels; no measurements were made in areas of overlap. Individual ants were then manually segmented from resultant image stacks using the freehand cropping tool in Fiji image analysis software [54]. Individual stacks were then binarized using the triangle method [55] and the binarized image was inverted to allow for modelling of the airspace rather than the animal tissue. Internal airspaces were then mapped by selecting a single point in the abdominal tracheal system and using the Find Connected Regions Fiji plugin. This process produced a second binary image stack of only the internal airspaces. This internal-only image stack was imported into Avizo 9.0.0 (FEI Visualization Sciences Group, Oregon, http://www.vsg3d.com/) and a volume render of the image stack was created to enable measurement of the airspace in three-dimensional space (Fig 2).

Measurements were made on a single abdominal branch on the left-hand side of each animal (Fig 2). The diameter of each branch was measured as the largest distance between the internal tracheal walls of the primary, secondary and tertiary branches (Figs 1 and 2) using Avizo’s 3D length measuring tool. Values were multiplied by a scaling factor of 7.6 to convert them from distance in pixels to μm. Bifurcations (100/165 measured branches) and less commonly three (61/165) and four (4/165) branches were found in the tracheal architecture. Branches with three or more sub-branches were found exclusively in the genus Camponotus . Given previous assessments that some asymmetry does not alter assumptions of area-preserving or area-increasing branching [12] we included measurements of all branch-types in our analyses. All tracheal diameter measurements were made manually by the first author (IA) and then halved to represent radii. Measurements were validated by remeasuring the full dataset and comparing first and second measurement values: across the full data set of 1258 measurements, no measures differed statistically by more than 0.15 μm (below the nominal resolution of the system of 7.6 μm), with a mean difference of 0.05 μm. The original measurements were used for analysis.

### Statistical analysis

The assumptions of area-preserving or area-increasing tracheal network branching were examined in two ways. First, we estimated the exponent (α) in the equation:

${r}_{i}^{\alpha }={r}_{ja}^{\alpha }+{r}_{jb}^{\alpha }$
where ri is the radius of the branch at parent level i and rj the radius of the bifurcating branches a and b at subsidiary level j [21], using the uniroot function in R (v3.5.0) [56], with an expansion to include additional branches in the case of more complex branching patterns (e.g. trifurcations). In some insects, the relationship between features among levels and in different parts of the tracheal system can change [36,37]. We therefore used data for all individuals from all species to separately estimate α across step-wise levels, from level 1 to level 2, and then from level 2 to level 3. In the latter case, we examined both sets of branches (i.e. branches a and b on Fig 1) to ensure the pattern of branching is consistent. For illustration we also solved Eq 2 for all bifurcating branches of individuals across all levels simultaneously. The resulting distribution of α in each case was right-skewed, thus we transformed α (log10) so that empirical values could be reliably tested against expected values (0.30 and 0.48 for α of 2 or 3, respectively) using a t-test in R.

Next, we determined whether relationships between parent (as X) and the summed area of child branches (as Y) changed in concert with changes in radius size of the primary level tracheae across all of the species examined. Using empirical data, the slope of a relationship between the left-hand and right-hand terms of Eq 2 closest to 1 (assuming the intercept passes through the origin), under either area-decreasing (α = 1), area-preserving (α = 2), or area-increasing branching (α = 3), will indicate the branching pattern with the greatest empirical support. To examine whether this was the case, we used model I regression (Ordinary Least Squares), with the intercept forced through the origin, implemented in R, and using Major Axis (Model II) regression because of equal measurement error expected in both the dependent and independent variables [57], using the lmodel2 library in R. In each case we excluded outliers that were clearly outside the size range of most of our data (level 1 to 2: 6 individuals; level 2a to 3a: 9 individuals; level 2b to 3b: 13 individuals). Outlier removal and missing data meant that a total of 165, 159, and 153 individuals were analysed in each case. Although comparison among species was not our primary focus, species level effects were nevertheless tested for by including species as a random term in a generalised linear model implemented in the nlme package of R.

### CO2 transport modelling

A simple model of steady-state diffusive gas exchange was used to elucidate an optimised design for effective outward transport of CO2 . The diffusion model was based on the modified version of Fick’s law of diffusion as detailed in Kestler [58]:

$J=\frac{A}{L}K{\mathrm{\Delta }pp}_{,{CO}_{2}}$
where, J = volumetric flux of CO2 (in m3 s-1), A = tracheal cross-sectional area (in m2), L = branch length (in m), K = Krogh’s constant (assumed to be 1.428104 × 10−10 m2 s-1 Pa-1 ) [58] and Δpp,CO2 is the partial pressure difference of CO2 (in Pa). To obtain the volumetric flux of CO2 in μl h-1, we multiplied J by a conversion factor 3.6 × 1012 (considering; 3600 s per hour and 1 × 109 μl per m3). We treated each tracheal branch level as a fluidic channel in a serial configuration. As two child branches converge and feed into the parent branch upstream (i.e. from Level 5 to Level 1), they influence CO2 transport, akin to hydraulic resistance impeding fluid flow in a typical fluidic system. Accommodating for CO2 influx into the tracheal network from the tissue and integrating along the length, we derived an analytical expression for the complete system (for each spiracle inwards, making the simplifying assumption that there’s little connectedness among spiracles), consisting of 5 bifurcating tracheal branch levels.

To derive (full derivation detailed in S2 Text) the diffusive gas transport, we first assumed the radius of the deepest branch (i.e. r5) and the corresponding branch lengths. A model ant with tracheal branch lengths of L1: 1569 μm, L2: 976 μm, L3: 586 μm, L4: 352 μm, L5: 211 μm (total tracheal length of 3694 μm) and tracheal radius at the deepest level (i.e. r5 ) 3 μm was then assumed. Level 5 tracheole radii of this size are in keeping with the available data on insects [36,39]. Second, we assumed that the effective diffusive flux per unit area, φ from the tissue into the tracheal branch which is known to occur at level 3–5 [39] is a constant of arbitrary value. Finally, we assumed a CO2 partial pressure difference between the deepest branch level (i.e. level 5) and the external environment to be 6 kPa [33,49]. Based on the assumptions made and using Eq 3, the effective outward flux of CO2 could be evaluated (S2 Text). Assuming the corresponding branch lengths are held constant, we investigated the effective outward volumetric flux for α as 1 (area reducing), 2 (area conserving) and 3 (area increasing) respectively (Fig 3). The influence on net outward flux due to the presence of air sacs, which are distributed across ant tracheal systems (Fig 2) [27], was also modelled. An air sac, which acts to expel the entire air cavity thereafter (i.e. pressure release) was included at the end of the tracheal branch at either level 4 or level 3 of the modelled system. Values were summed across six pairs of abdominal spiracles in alignment with our observations of number of spiracles in the ants examined. Diffusion-based models for high CO2 environments (i.e. lower Δpp,CO2 of 4 kPa) and high activity (i.e. higher Δpp,CO2 of 8 kPa), with all other parameters unchanged, were also created to examine the effects of an altered CO2 diffusion gradient on achievable flux.

In addition, a simplified convection model was used to understand the convective contribution of CO2 efflux when the ants undergo periodic breathing. The model assumed a 66% evacuation of the entire tracheal airway with every exhalation. The frequency of tracheal contraction was assumed to be 0.26 Hz [35], whilst the assumptions used in the diffusive model were retained (i.e. the branch lengths, partial pressure difference and r5 assumed to be the same). The local partial pressures throughout the airway were evaluated as per the diffusive steady-state case (S2 Text) to accommodate for the resultant varying local concentrations of CO2 in the tracheal network.

To compare with modelled estimates of CO2 flux, empirical data on ant metabolic rates were used to estimate CO2 flux (S5 Table). Thirty-eight estimates (from 21 species) of metabolic rate (based on measures of CO2 production or O2 consumption) under a range of temperatures (20–43°C) for ants undertaking different activities (resting, walking or running workers) were used. Values were converted to CO2 flux rate following Lighton [59]. Based on these data mean empirical CO2 flux rates of: (1) Resting 10.63 μl.h-1, (2) Walking/running 115 μl.h-1, were used, acknowledging that thermal variation in rates has been included implicitly by using measures made at different temperatures. Tissue-level CO2 flux rate was calculated as CO2 flux rate divided by 70% of mass (assuming that 30% of tissue is metabolically relatively inactive–such as the chitin exoskeleton).

## Acknowledgements

We thank members of the ChownLab for assistance during the scanning. Jon Harrison and Roger Seymour provided insightful comments on a previous version of the manuscript. Doug Glazier and Jake Socha provided valuable comments during the review process.

## References

1

GB West, JH Brown, BJ Enquist. . A general model for the origin of allometric scaling laws in biology. Science. 1997; 276: , pp.122–126. , doi: 10.1126/science.276.5309.122

2

JH Brown, JF Gillooly, AP Allen, V Savage, GB West. . Toward a metabolic theory of ecology. Ecology. 2004; 85: , pp.1771–1789.

3

J van der Meer. . Metabolic theories in ecology. Trends in Ecology and Evolution. 2006; 21:, pp.136–140. , doi: 10.1016/j.tree.2005.11.004

4

J Maino, MR Kearney, RM Nisbet, SALM Kooijman. . Reconciling theories for metabolic scaling. Journal of Animal Ecology. 2014; 83: , pp.20–29. , doi: 10.1111/1365-2656.12085

5

DS Glazier. . A unifying explanation for diverse metabolic scaling in animals and plants. Biological Reviews. 2010; 85: , pp.111–138. , doi: 10.1111/j.1469-185X.2009.00095.x

6

DS Glazier. . Beyond the '3/4-power law': variation in the intra- and interspecific scaling of metabolic rate in animals. Biological Reviews. 2005; 80: , pp.611–662. , doi: 10.1017/S1464793105006834

7

JR Banavar, et al. A general basis for quarter-power scaling in animals. Proceedings of the National Academy of Sciences of the USA. 2010; 107: , pp.15816–15820. , doi: 10.1073/pnas.1009974107

8

SALM Kooijman. Dynamic Energy Budget Theory for Metabolic Organisation. Cambridge: Cambridge University Press; 2010.

9

JF Harrison. . Do performance-safety tradeoffs cause hypometric scaling in animals. Trends in Ecology and Evolution. 2017; 32: , pp.653–664. , doi: 10.1016/j.tree.2017.05.008

10

VM Savage, EJ Deeds, W Fontana. . Sizing up allometric scaling theory. PLoS Computational Biology. 2008; 4: , pp.e1000171, doi: 10.1371/journal.pcbi.1000171

11

DD Smith, et al. Deviation from symmetrically self-similar branching in trees predicts altered hydraulics, mechanics, light interception and metabolic scaling. New Phytologist. 2014; 201: , pp.217–229. , doi: 10.1111/nph.12487

12

AB Brummer, VM Savage, BJ Enquist. . A general model for metabolic scaling in self-similar asymmetric networks. PLoS Computational Biology. 2017; 13: , pp.e1005394, doi: 10.1371/journal.pcbi.1005394

13

J Kozłowski, M Konarzewski, AT Gawelczyk. . Cell size as a link between noncoding DNA and metabolic rate scaling. Proceedings of the National Academy of Sciences of the USA. 2003; 100: , pp.14080–14085. , doi: 10.1073/pnas.2334605100

14

SL Chown, E Marais, JS Terblanche, CJ Klok, JRB Lighton, TM Blackburn. . Scaling of insect metabolic rate is inconsistent with the nutrient supply network model. Functional Ecology. 2007; 21: , pp.282–290.

15

CR White, MR Kearney, PGD Matthews, SALM Kooijman, DJ Marshall. . A manipulative test of competing theories for metabolic scaling. American Naturalist. 2011; 178: , pp.746–754. , doi: 10.1086/662666

16

AG Hirst, DS Glazier, D Atkinson. . Body shape shifting during growth permits tests that distinguish between competing geometric theories of metabolic scaling. Ecology Letters. 2014; 17: , pp.1274–1281. , doi: 10.1111/ele.12334

17

D Glazier. . Rediscovering and reviving old observations and explanations of metabolic scaling in living systems. Systems. 2018; 6: , pp.4.

18

19

RF Chapman. The Insects: Structure and Function. 4th edn. Cambridge: Cambridge University Press; 1998.

20

VM Savage, et al. Hydraulic trade-offs and space filling enable better predictions of vascular structure and function in plants. Proceedings of the National Academy of Sciences of the USA. 2010; 107: , pp.22722–22727. , doi: 10.1073/pnas.1012194108

21

CA Price, SJ Knox, TJ Brodribb. . The influence of branch order on optimal leaf vein geometries: Murray's law and area preserving branching. PLoS One. 2013; 8: , pp.e85420, doi: 10.1371/journal.pone.0085420

22

LP Bentley, et al. An empirical assessment of tree branching networks and implications for plant allometric scaling models. Ecology Letters. 2013; 16: , pp.1069–1078. , doi: 10.1111/ele.12127

23

X Zheng, et al. Bio-inspired Murray materials for mass transfer and activity. Nature Communications. 2017; 8: , pp.14921, doi: 10.1038/ncomms14921

24

MW Westneat, et al. Tracheal respiration in insects visualized with synchrotron X-ray imaging. Science. 2003; 299: , pp.558–560. , doi: 10.1126/science.1078008

25

EP Snelling, R Duncker, KK Jones, EP Fagan-Jeffries, RS Seymour. . Flight metabolic rate of Locusta migratoria in relation to oxygen partial pressure in atmospheres of varying diffusivity and density. Journal of Experimental Biology. 2017; 220: , pp.4432–4439. , doi: 10.1242/jeb.168187

26

A Krogh. . Studien über tracheenrespiration. II. Über gasdiffusion in den tracheen. Pflügers Archiv. 1920; 179: , pp.95–112.

27

JJ Socha, TD Förster, KJ Greenlee. . Issues of convection in insect respiration: insights from synchrotron x-ray imaging and beyond. Respiratory Physiology and Neurobiology. 2010; 173: , pp.S65–S73. , doi: 10.1016/j.resp.2010.03.013

28

JJ Socha, F De Carlo. . Use of synchrotron tomography to image naturalistic anatomy in insects. Developments in X-Ray Tomography VI, Proceedings of SPIE. 2008; 7078: , pp.70780A1–70780A7.

29

MW Westneat, JJ Socha, WK Lee. . Advances in biological structure, function, and physiology using synchrotron x-ray imaging. Annual Review of Physiology. 2008; 70: , pp.119–142. , doi: 10.1146/annurev.physiol.70.113006.100434

30

JS Terblanche, HA Woods. . Why do models of insect respiratory patterns fail?Journal of Experimental Biology. 2018; 221: , pp.jeb130039, doi: 10.1242/jeb.130039

31

PGD Matthews. . The mechanisms underlying the production of discontinuous gas exchange cycles in insects. Journal of Comparative Physiology B. 2018; 188: , pp.195–210.

32

J. Nunome. Studies on the respiration of the silkworm. Part III. On the air current of respiration. Journal of Sericological Science Japan. 1951; 20: , pp.111–127.

33

SK Hetz, TJ Bradley. . Insects breathe discontinuously to avoid oxygen toxicity. Nature. 2005; 433: , pp.516–519. , doi: 10.1038/nature03106

34

JRB Lighton, T Fukushi, R Wehner. . Ventilation in Cataglyphis bicolor: Regulation of carbon dioxide release from the thoracic and abdominal spiracles. Journal of Insect Physiology. 1993; 39: , pp.687–699.

35

JJ Socha, WK Lee, JF Harrison, JS Waters, K Fezzaa, MW Westneat. . Correlated patterns of tracheal compression and convective gas exchange in a carabid beetle. Journal of Experimental Biology. 2008; 211: , pp.3409–3420. , doi: 10.1242/jeb.019877

36

T Weis-Fogh. . Diffusion in insect wing muscle, the most active tissue known. Journal of Experimental Biology. 1964; 41: , pp.229–256.

37

WH Thorpe, DJ Crisp. . Studies on plastron respiration II. The respiratory efficiency of the plastron in Aphelocheirus. Journal of Experimental Biology. 1947; 24: , pp.270–303.

38

J Harrison, et al. Developmental plasticity and stability in the tracheal networks supplying Drosophila flight muscle in response to rearing oxygen level. Journal of Insect Physiology. 2018; 106: , pp.189–198. , doi: 10.1016/j.jinsphys.2017.09.006

39

A Schmitz, SF Perry. . Stereological determination of tracheal volume and diffusing capacity of the tracheal walls in the stick insect Carausius morosus (Phasmatodea, Lonchodidae). Physiological and Biochemical Zoology. 1999; 72: , pp.205–218. , doi: 10.1086/316655

40

M Bollazzi, LC Forti, F Roces. . Ventilation of the giant nests of Atta leaf-cutting ants: does underground circulating air enter the fungus chambers?Insectes Sociaux. 2012; 59: , pp.487–498.

41

MG Nielsen, K Christian, PG Henriksen, D Birkmose. . Respiration by mangrove ants Camponotus anderseni during nest submersion associated with tidal inundation in Northern Australia. Physiological Entomology. 2006; 31: , pp.120–126.

42

MG Nielsen, KA Christian. . The mangrove ant, Camponotus anderseni, switches to anaerobic respiration in response to elevated CO2 levels. Journal of Insect Physiology. 2007; 53: , pp.505–508. , doi: 10.1016/j.jinsphys.2007.02.002

43

SM Simelane, S Abelman, FD Duncan. . Gas exchange models for a flexible insect tracheal system. Acta Biotheoretica. 2016; 64: , pp.161–196. , doi: 10.1007/s10441-016-9278-z

44

EP Economo, N Narula, NR Friedman, MD Weiser, B Guenard. . Macroecology and macroevolution of the latitudinal diversity gradient in ants. Nature Communications. 2018; 9: , pp.1778, doi: 10.1038/s41467-018-04218-4

45

L von Bertalanffy. . Quantitative laws in metabolism and growth. The Quarterly Review of Biology1957; 32: , pp.217–231. , doi: 10.1086/401873

46

T Jonsson. . Metabolic theory predicts animal self-thinning. Journal of Animal Ecology. 2017; 86: , pp.645–653. , doi: 10.1111/1365-2656.12638

47

RF Hechinger, KL Sheehan, AV Turner. . Metabolic theory of ecology successfully predicts distinct scaling of ectoparasite loads on hosts. Proceedings of the Royal Society B. 2019; 286: , pp.20191777, doi: 10.1098/rspb.2019.1777

48

B Hölldobler, EO Wilson. The Ants. Cambridge, MassBelknap Press of Harvard University Press; 1990.

49

50

A Boyde, FA McCorkell, GK Taylor, RJ Bomphrey, M Doube. . Iodine vapor staining for atomic number contrast in backscattered electron and x-ray imaging. Microscopy Research and Technique. 2014; 77:, pp.1044–1051. , doi: 10.1002/jemt.22435

51

K Ogata, RW Taylor. . Ants of the genus Myrmecia Fabricius: a preliminary review and key to the named species (Hymenoptera: Formicidae: Myrmeciinae). Journal of Natural History. 1991; 25:, pp.1623–1673.

52

AJ McArthur. A key to Camponotus Mayr of Australia In: RR Snelling, BL Fisher, PS Ward, editors. Advances in Ant Systematics (Hymenoptera: Formicidae): Homage to E. O. Wilson– 50 Years of Contributions. Logan: Memoirs of the American Entomological Institute2007; 80: , pp.290–351.

53

C Hall, et al. Detectors for the Imaging and Medical Beam Line at the Australian Synchrotron. Journal of Instrumentation. 2013; 8: , pp.C06011–C06011.

54

J Schindelin, et al. Fiji: An open-source platform for biological image analysis. Nature Methods. 2012; 9: , pp.676–682. , doi: 10.1038/nmeth.2019

55

GW Zack, WE Rogers, SA Latt. . Automatic measurement of sister chromatid exchange frequency. Journal of Histochemistry and Cytochemistry. 1977; 25: , pp.741–753. , doi: 10.1177/25.7.70454

56

R Core Team. R: A language and environment for statistical computingVienna, Austria: R Foundation for Statistical Computing2018 Available from: https://www.R-project.org/

57

P Legendre, L Legendre. Numerical Ecology. 2nd Edition. Amsterdam: Elsevier Science; 1998.

58

P Kestler. Respiration and respiratory water loss In: KH Hoffmann, editor. Environmental Physiology and Biochemistry of Insects (ed K. H. Hoffmann). Berlin: Springer; 1985, pp.137–183.

59

JRB Lighton. Measuring Metabolic RatesA Manual for Scientists. Oxford: Oxford University Press; 2008.

This paper was transferred from another journal. As a result, its full editorial history (including decision letters, peer reviews and author responses) may not be present.

16 Dec 2019

Dear Dr Chown,

Thank you very much for submitting your manuscript 'Tracheal branching in ants is area-decreasing, violating a central assumption of network transport models' for review by PLOS Computational Biology. Your manuscript has been fully evaluated by the PLOS Computational Biology editorial team and in this case also by independent peer reviewers. The reviewers appreciated the attention to an important problem, but raised some substantial concerns about the manuscript as it currently stands. While your manuscript cannot be accepted in its present form, we are willing to consider a revised version in which the issues raised by the reviewers have been adequately addressed. We cannot, of course, promise publication at that time.

Please note while forming your response, if your article is accepted, you may have the opportunity to make the peer review history publicly available. The record will include editor decision letters (with reviews) and your responses to reviewer comments. If eligible, we will contact you to opt in or out.

Your revisions should address the specific points made by each reviewer. Please return the revised version within the next 60 days. If you anticipate any delay in its return, we ask that you let us know the expected resubmission date by email at ploscompbiol@plos.org. Revised manuscripts received beyond 60 days may require evaluation and peer review similar to that applied to newly submitted manuscripts.

(1) A detailed list of your responses to the review comments and the changes you have made in the manuscript. We require a file of this nature before your manuscript is passed back to the editors.

(2) A copy of your manuscript with the changes highlighted (encouraged). We encourage authors, if possible to show clearly where changes have been made to their manuscript e.g. by highlighting text.

(3) A striking still image to accompany your article (optional). If the image is judged to be suitable by the editors, it may be featured on our website and might be chosen as the issue image for that month. These square, high-quality images should be accompanied by a short caption. Please note as well that there should be no copyright restrictions on the use of the image, so that it can be published under the Open-Access license and be subject only to appropriate attribution.

Before you resubmit your manuscript, please consult our Submission Checklist to ensure your manuscript is formatted correctly for PLOS Computational Biology: http://www.ploscompbiol.org/static/checklist.action. Some key points to remember are:

- Figures uploaded separately as TIFF or EPS files (if you wish, your figures may remain in your main manuscript file in addition).

- Supporting Information uploaded as separate files, titled Dataset, Figure, Table, Text, Protocol, Audio, or Video.

- Funding information in the 'Financial Disclosure' box in the online system.

While revising your submission, please upload your figure files to the Preflight Analysis and Conversion Engine (PACE) digital diagnostic tool, https://pacev2.apexcovantage.com PACE helps ensure that figures meet PLOS requirements. To use PACE, you must first register as a user. Then, login and navigate to the UPLOAD tab, where you will find detailed instructions on how to use the tool. If you encounter any issues or have any questions when using PACE, please email us at figures@plos.org.

To enhance the reproducibility of your results, we recommend that you deposit your laboratory protocols in protocols.io, where a protocol can be assigned its own identifier (DOI) such that it can be cited independently in the future. For instructions see here

We are sorry that we cannot be more positive about your manuscript at this stage, but if you have any concerns or questions, please do not hesitate to contact us.

Sincerely,

Van M Savage

Guest Editor

PLOS Computational Biology

Stefano Allesina

Deputy Editor

PLOS Computational Biology

A link appears below if there are any accompanying review attachments. If you believe any reviews to be missing, please contact ploscompbiol@plos.org immediately:

The reviewers both appreciate the contributions of this paper and recognize the incredible value of the empirical data and the amount of work it took to obtain these measurements. The connection to allometric theory is also viewed as important. Indeed, the common theme among the reviewers is that there should be more discussion of the details and debates about metabolic theory--though for somewhat different reasons and points. I think the general advice is good. Test acknowledging limitations of allometric theory, that debates exists, the variation exists and possible sources of this variation, and how this all connects to the data in the present paper would be welcome. However, I think this can be kept fairly brief via modifications to existing text and perhaps a few additional paragraphs throughout. I don't think the main themes of the paper needs to change, and I think leading with the main results and dominant views of metabolic theory is appropriate. All other comments by the referees should be addressed and changes made to the paper where appropriate, including correction of typographical errors.

Reviewer's Responses to Questions

The authors show that the tracheal transport networks in 20 species of ants are area decreasing, rather than area preserving or slightly increasing, as assumed by resource-transport-network models at the core of the influential metabolic theory of ecology (MTE). Technically, this is a remarkable piece of work. The authors’ findings and their implications for O2 uptake and CO2 release are both useful and noteworthy because they apply to the most diverse major taxon of life, the insects.

However, I have two major concerns, which I believe that the authors should consider.

1) It would be helpful if the authors explicitly addressed the question of whether and how the geometry and physics of transport networks affect the body-mass scaling of metabolic rate. As the authors know, this is a controversial topic, but in my opinion, they do not give sufficient context so that readers can fully appreciate the many dimensions of this controversy and the relevance of their own findings. I am not suggesting a long discussion about this. However, it would be helpful if readers were made aware that:

(a) Closed resource-transport networks, assumed by the MTE, do not occur in all organisms. Many groups of animals and plants do not have closed resource-transport networks, and some do not have any anatomical transport networks at all. Therefore the MTE, as originally proposed, has limited applicability.

(b) Insects have air sacs in their tracheal transport networks, which departs from the network geometry specified by resource-transport-network models at the core of the MTE.

(c) Resource-transport-network models of metabolic scaling currently have no direct mechanistic empirical support, unlike some other models based on surface-area exchange of resources/wastes, and size-related changes in cell size/number, various resource-demanding processes, and the relative amounts of tissues with varying metabolic intensities (reviewed in Glazier 2014, 2018a).

(d) Several lines of evidence contradict predictions of resource-transport-network models of metabolic scaling, or are not explained by these models (see e.g., Chown et al. 2007; Glazier 2014, 2015, 2018a; Hirst et al. 2014; Glazier et al. 2015; Harrison 2017).

2) It would be helpful if the authors considered whether the area-decreasing branching geometry of tracheal transport networks in ants helps to explain their metabolic scaling, if at all. In particular, how do the authors’ findings help explain the diverse metabolic scaling relationships shown by various ant species, as reported previously by one of the authors and his colleagues (Chown et al. 2007). Can this variation in metabolic scaling (slopes ranging between 0.56 and 1.28) be explained by parallel variation in the geometry of their transport networks? If not, the claim that resource-transport-network models of metabolic scaling have no direct mechanistic empirical support still stands. No mention is made of the claim by Chown et al. (2007) that the cell-size model better explains the metabolic scaling of ants than do resource-transport-network models.

In the last sentence of their summary the authors state: “Our work suggests that much still needs to be done to understand the fundamental assumptions underlying network transport models and how they apply more generally across life – especially in the context of virtually ubiquitous metabolic scaling.” However, nowhere in the main text do the authors discuss the relevance of their findings for our understanding of metabolic scaling in an explicit way.

Author summary, L 4, 12: Please clarify what is meant by “inwards direction”. Do you mean from the external spiracles towards the inner body core?

Author summary, L 14-15: What is meant by “virtually ubiquitous metabolic scaling”, Are you referring to the 3/4-power law, or merely to the fact that metabolic rate scales with body mass regardless of the slope. Metabolic scaling may be log-log linear or curvilinear with slopes varying from near 0 to > 1. Whatever way the authors view this, how do they think that the geometry of transport networks affects metabolic scaling relationships, if at all? Evidence is growing that transport networks may not exert supply limits on resting metabolic rate, as assumed by the MTE (see e.g., Glazier 2015; Harrison 2017). Some investigators are now arguing that effects of resource demand should be considered in addition to or instead of resource-supply limits (e.g., Glazier 2014, 2018b; Harrison 2017). The question boils down to whether the geometry of transport networks are primary causes or secondary responses to metabolic scaling.

Methods (p 12): In OLS regression analyses, it is assumed that there is a dependent (Y) and independent variable (X). Is ri or rj the X variable, and if so what is the justification?

Fig. 2b, e: Please make sure that the features designated in these pictures are more easily visible to readers.

Fig. 3a: Are the OLS “outcomes” slopes? If so, please say so. Also please make clear what the X and Y variables are for these regressions. From reading the text (e.g. on P 4), I assume that the variables are ri and rj (radii of the parent and child vessels), but I am not sure which was considered X versus Y. I am having difficulty matching up the data shown in panel ‘a’ with those shown in panel ‘b’.

References:

Chown, S. L., Marais, E., Terblanche, J. S., Klok, C. J., Lighton, J. R. B., & Blackburn, T. M. (2007). Scaling of insect metabolic rate is inconsistent with the nutrient supply network model. Functional Ecology, 21(2), 282-290.

Glazier, D. (2015). Body-mass scaling of metabolic rate: what are the relative roles of cellular versus systemic effects? Biology, 4(1), 187-199.

Glazier, D. (2018a). Rediscovering and reviving old observations and explanations of metabolic scaling in living systems. Systems, 6(1), 4.

Glazier, D. S. (2018b). Resource supply and demand both affect metabolic scaling: A response to Harrison. Trends in Ecology & Evolution, 33(4), 237-238.

Glazier, D. S., Hirst, A. G., & Atkinson, D. (2015). Shape shifting predicts ontogenetic changes in metabolic scaling in diverse aquatic invertebrates. Proceedings of the Royal Society B: Biological Sciences, 282(1802), 20142302.

Harrison, J. F. (2017). Do performance–safety tradeoffs cause hypometric metabolic scaling in animals?. Trends in Ecology & Evolution, 32(9), 653-664.

Hirst, A. G., Glazier, D. S., & Atkinson, D. (2014). Body shape shifting during growth permits tests that distinguish between competing geometric theories of metabolic scaling. Ecology Letters, 17(10), 1274-1281.

Reviewer #2: see attachment

**********

Have all data underlying the figures and results presented in the manuscript been provided?

Large-scale datasets should be made available via a public repository as described in the PLOS Computational Biologydata availability policy, and numerical data that underlies graphs or summary statistics should be provided in spreadsheet form as supporting information.

Reviewer #1: Yes

Reviewer #2: Yes

**********

PLOS authors have the option to publish the peer review history of their article (what does this mean?). If published, this will include your full peer review and any attached files.

If you choose “no”, your identity will remain anonymous but your review may still be made public.

Reviewer #1: Yes: Douglas S. Glazier

Reviewer #2: Yes: Jake Socha

18 Feb 2020

31 Mar 2020

Dear Dr. Chown,

Thank you very much for submitting your manuscript "Tracheal branching in ants is area-decreasing, violating a central assumption of network transport models" for consideration at PLOS Computational Biology. As with all papers reviewed by the journal, your manuscript was reviewed by members of the editorial board and by several independent reviewers. The reviewers appreciated the attention to an important topic. Based on the reviews, we are likely to accept this manuscript for publication, providing that you modify the manuscript according to the review recommendations.

Reviewer 1 had no additional comments, and Reviewer 2 had a few final comments that the authors should consider in finalizing their manuscript. Both reviewers were very pleased with the revision and with the content and quality of the paper overall.

Please prepare and submit your revised manuscript within 30 days. If you anticipate any delay, please let us know the expected resubmission date by replying to this email.

[1] A letter containing a detailed list of your responses to all review comments, and a description of the changes you have made in the manuscript. Please note while forming your response, if your article is accepted, you may have the opportunity to make the peer review history publicly available. The record will include editor decision letters (with reviews) and your responses to reviewer comments. If eligible, we will contact you to opt in or out

[2] Two versions of the revised manuscript: one with either highlights or tracked changes denoting where the text has been changed; the other a clean version (uploaded as the manuscript file).

Thank you again for your submission to our journal. We hope that our editorial process has been constructive so far, and we welcome your feedback at any time. Please don't hesitate to contact us if you have any questions or comments.

Sincerely,

Van M Savage

Guest Editor

PLOS Computational Biology

Stefano Allesina

Deputy Editor

PLOS Computational Biology

***********************

A link appears below if there are any accompanying review attachments. If you believe any reviews to be missing, please contact ploscompbiol@plos.org immediately:

Reviewer's Responses to Questions

Reviewer #1: Great piece of work! The authors have effectively dealt with all of my concerns. I look forward to the day when direct comparisons between the geometry of resource-transport networks and the scaling of metabolic rate are made across individual species, a huge but necessary task for evaluating the validity of resource-transport network models of metabolic scaling.

Reviewer #2: I apologize to the authors for my delay. The Covid-19 crisis hit the US in the past few weeks, and as with many of us, I scrambled to rework my classes and deal with other things as a consequence.

The authors have made a great number of changes to improve the manuscript, and I have no further major comments. I do have a few minor suggestions below. I stand by my original assessment that this manuscript is an exciting piece of work, and I applaud the authors for their efforts in going above and beyond to provide a great amount of detail, useful scholarship for the field. Congratulations.

Page 2

>“Interpretation of the tracheal system as one optimized for the release of carbon dioxide, while readily catering to oxygen demand, explains the branching pattern.”

In the abstract here, having not read the study, it’s not clear if mechanical optimization or evolutionary optimization is intended. As these mean different things and are analyzed differently, it would be helpful to indicate that here. (I assume mechanical is meant.)

>“Our results, together with widespread demonstration that gas exchange in insects includes, and is often dominated by, convection, indicate that for generality, network transport models must include consideration of systems with bidirectional flow.”

The directionality doesn’t matter, per se. The same argument could apply for grasshoppers, which partially use unidirectional flow.

Page 4

>“models do not apply in all circumstances, and often concluding that other approaches have”

Change ‘concluding’ to ‘conclude’.

Page 8

>”In consequence, tracheal systems are likely to be optimised for CO2 transport outwards.”

Same comment about mechanical optimization applies here.

Page 9

>”Until the implications … is explored further,”

Change to ‘are explored’.

Page 11

Thank you for supplying details of the imaging and providing sample slices in the supplement; it is all useful to see. One more question: what is the nominal resolution of the system? The value of 0.15 µm difference described below should be at or below the possible resolution.

Also, as there are very few studies of tracheal systems using synchrotron tomography, I think it would be useful to point to the first one (Socha and De Carlo 2008), because the methods of preparation described here are different (and were also used in Harrison et al 2018). Of course no strict need to do this; I don’t mean to insist that you cite my paper, by any means.

Page 14

>”A simple model of steady state diffusive gas exchange”

Supplement 1

I apologize for my original comments about editorial errors in the document. Somehow when I was reading it I missed the header, indicating that it was a translation. Clearly I was reading too quickly, and I thought that the text described a comparison to his original study. I actually think it’s a great service to the field to provide the translation in this paper, so I thank you. (Many years ago, I believe that I sat down with a German speaker to discuss the paper, but I didn’t have a written translation to work with.) For completeness, it would be nice to provide the name of the person who translated it as credit at the top, as well as the full original citation and translated title.

Supplement 6

Just a side note in response to the comment, “We note that Krogh did not supply units, and that ratios have no units.” Ratios are not necessarily dimensionless; for example, speed is a ratio that has units of length per time.

**********

Have all data underlying the figures and results presented in the manuscript been provided?

Large-scale datasets should be made available via a public repository as described in the PLOS Computational Biologydata availability policy, and numerical data that underlies graphs or summary statistics should be provided in spreadsheet form as supporting information.

Reviewer #1: Yes

Reviewer #2: Yes

**********

PLOS authors have the option to publish the peer review history of their article (what does this mean?). If published, this will include your full peer review and any attached files.

If you choose “no”, your identity will remain anonymous but your review may still be made public.

Reviewer #1: Yes: Douglas S. Glazier

Reviewer #2: Yes: Jake Socha

Figure Files:

While revising your submission, please upload your figure files to the Preflight Analysis and Conversion Engine (PACE) digital diagnostic tool, https://pacev2.apexcovantage.com. PACE helps ensure that figures meet PLOS requirements. To use PACE, you must first register as a user. Then, login and navigate to the UPLOAD tab, where you will find detailed instructions on how to use the tool. If you encounter any issues or have any questions when using PACE, please email us at figures@plos.org.

Data Requirements:

Please note that, as a condition of publication, PLOS' data policy requires that you make available all data used to draw the conclusions outlined in your manuscript. Data must be deposited in an appropriate repository, included within the body of the manuscript, or uploaded as supporting information. This includes all numerical values that were used to generate graphs, histograms etc.. For an example in PLOS Biology see here: http://www.plosbiology.org/article/info%3Adoi%2F10.1371%2Fjournal.pbio.1001908#s5.

Reproducibility:

To enhance the reproducibility of your results, PLOS recommends that you deposit laboratory protocols in protocols.io, where a protocol can be assigned its own identifier (DOI) such that it can be cited independently in the future. For instructions see http://journals.plos.org/ploscompbiol/s/submission-guidelines#loc-materials-and-methods

1 Apr 2020

6 Apr 2020

Dear Dr. Chown,

We are pleased to inform you that your manuscript 'Tracheal branching in ants is area-decreasing, violating a central assumption of network transport models' has been provisionally accepted for publication in PLOS Computational Biology.

Before your manuscript can be formally accepted you will need to complete some formatting changes, which you will receive in a follow up email. A member of our team will be in touch with a set of requests.

Please note that your manuscript will not be scheduled for publication until you have made the required changes, so a swift response is appreciated.

IMPORTANT: The editorial review process is now complete. PLOS will only permit corrections to spelling, formatting or significant scientific errors from this point onwards. Requests for major changes, or any which affect the scientific understanding of your work, will cause delays to the publication date of your manuscript.

Should you, your institution's press office or the journal office choose to press release your paper, you will automatically be opted out of early publication. We ask that you notify us now if you or your institution is planning to press release the article. All press must be co-ordinated with PLOS.

Thank you again for supporting Open Access publishing; we are looking forward to publishing your work in PLOS Computational Biology.

Best regards,

Van M Savage

Guest Editor

PLOS Computational Biology

Stefano Allesina

Deputy Editor

PLOS Computational Biology

***********************************************************

15 Apr 2020

PCOMPBIOL-D-19-01679R2

Tracheal branching in ants is area-decreasing, violating a central assumption of network transport models

Dear Dr Chown,

I am pleased to inform you that your manuscript has been formally accepted for publication in PLOS Computational Biology. Your manuscript is now with our production department and you will be notified of the publication date in due course.

The corresponding author will soon be receiving a typeset proof for review, to ensure errors have not been introduced during production. Please review the PDF proof of your manuscript carefully, as this is the last chance to correct any errors. Please note that major changes, or those which affect the scientific understanding of the work, will likely cause delays to the publication date of your manuscript.

Soon after your final files are uploaded, unless you have opted out, the early version of your manuscript will be published online. The date of the early version will be your article's publication date. The final article will be published to the same URL, and all versions of the paper will be accessible to readers.

Thank you again for supporting PLOS Computational Biology and open-access publishing. We are looking forward to publishing your work!

With kind regards,

Laura Mallard

PLOS Computational Biology | Carlyle House, Carlyle Road, Cambridge CB4 3DN | United Kingdom ploscompbiol@plos.org | Phone +44 (0) 1223-442824 | ploscompbiol.org | @PLOSCompBiol

Citing articles via
https://www.researchpad.co/tools/openurl?pubtype=article&doi=10.1371/journal.pcbi.1007853&title=Tracheal branching in ants is area-decreasing, violating a central assumption of network transport models&author=Ian J. Aitkenhead,Grant A. Duffy,Citsabehsan Devendran,Michael R. Kearney,Adrian Neild,Steven L. Chown,Van M. Savage,Stefano Allesina,Van M Savage,Stefano Allesina,Van M Savage,Stefano Allesina,Van M Savage,Stefano Allesina,Van M Savage,&keyword=&subject=Research Article,Biology and Life Sciences,Organisms,Eukaryota,Animals,Invertebrates,Arthropoda,Insects,Hymenoptera,Ants,Biology and Life Sciences,Organisms,Eukaryota,Animals,Invertebrates,Arthropoda,Insects,Earth Sciences,Atmospheric Science,Atmospheric Chemistry,Greenhouse Gases,Carbon Dioxide,Physical Sciences,Chemistry,Environmental Chemistry,Atmospheric Chemistry,Greenhouse Gases,Carbon Dioxide,Ecology and Environmental Sciences,Environmental Chemistry,Atmospheric Chemistry,Greenhouse Gases,Carbon Dioxide,Physical Sciences,Chemistry,Chemical Compounds,Carbon Dioxide,Physical Sciences,Mathematics,Geometry,Radii,Computer and Information Sciences,Network Analysis,Physical Sciences,Physics,Classical Mechanics,Continuum Mechanics,Fluid Mechanics,Fluid Dynamics,Biology and Life Sciences,Anatomy,Animal Anatomy,Spiracles,Medicine and Health Sciences,Anatomy,Animal Anatomy,Spiracles,Biology and Life Sciences,Zoology,Animal Anatomy,Spiracles,Biology and Life Sciences,Anatomy,Respiratory System,Spiracles,Medicine and Health Sciences,Anatomy,Respiratory System,Spiracles,Engineering and Technology,Equipment,Particle Accelerators,Synchrotrons,