The Interrelationships Between Rodent Respiration in a Burrow

The Interrelationships between Rodent Respiration in a Burrow Environment and the Physical Ventilation of the Burrow System: a Chapter in the Physiological Ecology of Sundevall's Jird ( Meriones crassus )

Thesis submitted in partial fulfillment of the requirements for the degree of “DOCTOR OF PHILOSOPHY”

Inbal BricknerBraun

Submitted to the Senate of BenGurion University of the Negev

3 July, 2013

Beer Sheva

Thesis submitted in partial fulfillment of the requirements for the degree of "DOCTOR OF PHILOSOPHY"

Inbal BricknerBraun

Submitted to the Senate of BenGurion University of the Negev

Approved by the advisors

______

Approved by the Dean of the Kreitman School of Advanced Graduate Studies

______

3 July, 2013

Beer Sheva

This research was done under the supervision of Professors Berry Pinshow and Pedro Berliner

The Mitrani Department of Desert Ecology and the Wyler Department for Dryland Agriculture. The Jacob Blaustein Institute for Desert Research

ResearchStudent's Affidavit when Submitting the Doctoral Thesis for Judgment

I, Inbal BricknerBraun, whose signature appears below, hereby declare that (Please mark the appropriate statements):

X I have written this Thesis by myself, except for the help and guidance offered by my Thesis Advisors.

X The scientific materials included in this Thesis are products of my own research, culled from the period during which I was a research student.

___ This Thesis incorporates research materials produced in cooperation with others, excluding the technical help commonly received during experimental work. Therefore, I am attaching another affidavit stating the contributions made by myself and the other participants in this research, which has been approved by them and submitted with their approval.

Date: 3 July 2013 Student's name: Inbal BricknerBraun

Signature: ______

ACKNOWLEDGEMENTS

First and foremost I thank my advisers, Berry Pinshow and Pedro Berliner, for their boundless support and sharing of ideas; they taught me how to ask questions and how to search for answers; the values they instilled in me as a researcher will continue to guide me and be a part of me for the rest of my life. Berry, you have been my mentor, my friend and together with Hanna, my family. Pedro, your enjoyment of scientific discussion has been an inspiration to me.

My special thanks to Ishai Hoffman who has been involved in my project from the beginning; thank you for your friendship and for all that you have taught me during long, enjoyable hours in the technical lab and in the field. My special thanks also to Daniel Zucker for being a friend and my research partner; together we explored ideas during long hours, day and night in the lab and in the field, over many cups of coffee. I thank Irina Khokhlova for teaching me everything I know about animal maintenance, breeding and handling; thank you for your encouragement and generosity.

Big thanks to my friends and lab members over the past six years: to Carmi Korin, for several enjoyable evenings catching fruit bats in Beer Sheva and for patiently teaching me how to take blood samples from bats; to Marshall McCue for teaching me how to figure out how things work in the lab; to Itzick Vatnick who held my hand when I took the first steps in the world of acid base – thank you for your friendship, to Miri BenHamo and Shai Pilosof who provided me with invaluable statistical advice; to Edward Westen for his suggestions and advice; and to Cynthia Downs for exchange of knowledge and skills.

I thank Amos Ar for advising about my blood acidbase balance experiments, Marc Goldberg for technical support and for keeping the technical workshop door open, Rami Mosley for sharing his excellent creative ideas and David Klepatch for providing information from the Desert Meteorology Weather Station.

Most importantly, I thank my loving parents for always being there and Avi, my husband, for his love, support and friendship – for everything, really….

Contents Abstract ...... 3 General Introduction ...... 6 Study animal ...... 8 PART 1: THE RESPIRATORY ENVIRONMENT INSIDE THE BURROW OF SUNDEVALL'S JIRD...... 9 1.1 Introduction ...... 9 Research goals ...... 11 1.2 Methods ...... 13 Description of artificial burrow, type 1 ...... 13 Description of artificial burrow, type 2 ...... 14

Measurements of [CO 2] inside the burrows ...... 15 Measurements of wind speed ...... 15 Measurements of temperature ...... 16

Injection of CO 2 into the nest chamber ...... 16 1.2.1 Experiment 1. Mechanisms for burrow ventilation ...... 16 Experimental settings ...... 17 1.2.2 Experiment 2. Windinduced ventilation ...... 18

1.2.3 Experiment 3. The effect of the presence of an animal on nest chamber [CO 2] . 19

1.2.4 Experiment 4. [CO 2] in the nest chamber while young were being raised ...... 20 1.3 Results ...... 23 1.3.1 Experiment 1. Mechanisms for burrow ventilation ...... 23 1.3.2 Experiment 2. Windinduced ventilation ...... 27

1.3.3 Experiment 3. The effect of the presence of an animal on nest chamber [CO 2] . 30

1.3.4 Experiment 4. [CO 2] in the nest chamber while young were being raised ...... 32 1.4 Discussion ...... 40 Mechanisms of burrow ventilation ...... 40 Windinduced ventilation ...... 41

The effect of the presence of an animal on nest chamber [CO 2] ...... 42

[CO 2] in the nest chamber while young were being raised ...... 44

PART 2: PHYSIOLOGICAL RESPONSES OF SUNDEVALL’S JIRDS TO HIGH AMBIENT

CO 2 CONCENTRATION ...... 47 2.1 Introduction ...... 47 A brief introduction to acidbase balance ...... 47 The oxygen dissociation curve and the Bohr effect ...... 49

Acidbase balance when ambient [CO 2] is high ...... 50 Blood buffering capacity ...... 50 Ventilatory Regulation of AcidBase Balance ...... 51 Renal Compensation of AcidBase Balance ...... 52 Research goals ...... 52 2.2 Methods ...... 54 Animal maintenance ...... 54 The fat sand rat...... 54 Preparation of gas mixtures ...... 55

2.2.1 Experiment 1. CO 2 tolerance ...... 55 2.2.2 Experiment 2. Oxygen dissociation curve and Bohr effect ...... 56 2.2.3 Experiment 3. Blood buffer capacity ...... 58 2.2.4 Experiment 4. Behavioral response and physiological costs in jirds inspiring air

with high FCO 2 ...... 59

2.2.5 Experiment 5. The effect of sequential and acute changes in inspired FCO 2 on the energy expenditure of jirds...... 61 2.3 Results ...... 64 2.3.1 Experiment 1. CO 2 tolerance ...... 64 2.3.2 Experiment 2. Oxygen dissociation curve and Bohr effect ...... 64 2.3.3 Experiment 3. Blood buffer capacity ...... 65 2.3.4 Experiment 4. Behavioral response and physiological costs in jirds inspiring high

FCO 2 ...... 67

2.3.5 Experiment 5. The effect of sequential and acute changes in inspired FCO 2 on energy expenditure in jirds...... 69 2.4 Discussion ...... 74 CO 2 tolerance ...... 74 Oxygen dissociation curve and Bohr effect ...... 74 Blood buffer capacity ...... 76

Behavioral response and physiological costs in jirds inspiring high FCO 2 ...... 78 SUMMARY ...... 82 References ...... 85 Appendix ...... 97 104 ...... תקציר

Abstract Burrows are the living environment of species of many taxa, ranging in size from ants to aardvarks. Animals use burrows to avoid predation, find protection from adverse environmental conditions, raise young, and hoard food. It has been suggested that burrow geometry, together with the presumably low permeability of soils to gases, and the respiration

of the burrow's tenants, may bring about fractional concentrations of CO 2 (FCO 2) and O 2

(FO2) that are, respectively, above and below those in the free atmosphere. In mammals, persistent breathing of air containing high FCO 2 causes chronic respiratory acidosis, so burrow dwellers are either physiologically adapted to tolerate high FCO 2 in their environment, or alternatively, the architecture of burrows allows for their ventilation and prevents the built up of CO 2. I investigated ecological and physiological aspects of the life of a semifossorial desert rodent, Sundevall's jird, Meriones crassus Sundevall, 1842. The first part of my study was an investigation of the respiratory environment inside the burrow of jird, specifically the respiratory environment which the animal experiences in its immediate vicinity and what affects it. To begin with, I characterized the means by which an open ended burrow system is ventilated, and have offered a new explanation for the underlying mechanism of wind induced ventilation. I found that burrow ventilation occurs by several mechanisms, the importance of which varies even throughout the day, depending on ambient conditions and especially wind speed. At wind speeds above 2 m/s burrows are likely to be well ventilated by eddies, irregular gusts of air caused by turbulence in the boundary layer, that penetrate into the burrows from different entrances, and move atmospheric air to different depths of the main tunnels. Although eddies do not penetrate directly into parts of the burrow that branch out from the main tunnels ( e.g., nest chamber), I suggest that these parts remain well ventilated due to a large, and almost constant, gas concentration gradient maintained between

them and main tunnel. This gradient facilitates the diffusion of CO 2 into the main tunnel where it is removed by the next eddy. Even at low wind speeds (below 2 m/s) some eddies were still detected in the main tunnel. In addition, at low wind speeds gas diffusion in still air through the burrow and into the soil may become an important mechanism for gas exchange, and in that case, soil porosity may be imperative when considering the ventilation of dead end spaces, deep and far from entrances.

Next I examined how CO 2 concentration [CO 2] in a specific location in the burrow is affected by the whereabouts of the animal and showed that the location of the animal’s nose

(the direct source for CO 2) with respect to the CO 2 probe greatly affects the measured [CO 2] and the rate of increase/decrease in [CO 2]. Studies of respiratory gas concentrations in burrows seldom include information about the location of the animal within the burrow system at the time of sampling, probably due to the technical difficulties of accessing burrows nondestructively (Roper et al., 2001). My results indicate that knowledge of the whereabouts of the animal within the system is critical to the analysis of results, since the respiratory gas concentrations in the immediate vicinity of the animal are very different than those in the burrow system as a whole.

Finally, in order to assess what are the levels of [CO 2] that jirds cope with inside their burrows, I introduced pregnant female jirds into artificial burrows, and after the females gave birth, I measured [CO 2] and temperature in several locations along the burrow as well as wind speed at the surface, and ambient temperature ( Ta) for the duration of the 3weeks lactation period. My findings suggest that the CO 2 levels, which jirds commonly experience in their burrows, do not exceed 2 2.5%. These levels of CO2 are not considered as physiologically challenging for rodents. Only in one burrow that was completely impermeable to gas exchange with the soil environment, and in which gas exchange with the atmosphere could only occur via the entrances, [CO 2] reached over 7%. It is of note that the dam and pups in this burrow lived "normally" for the entire three weeks of measurement, and the pups were all in good health when trapped after the experiment.

In the second part of the study I showed that jirds voluntarily tolerated high CO 2 (7%) in

the air they breathe, and showed no preference for CO 2free air when given the choice. Then I continued to investigate several physiological traits that may enable jirds to tolerate high

inspired fractional concentrations of CO 2 (FCO 2) in their living environment. First I

examined the blood O 2transport properties of jirds in the context of its burrow's ostensibly

high CO 2/hypoxic environment. Contrary to expectation, P50 of jirds is similar to, or even slightly higher than, that predicted for an animal of its body size by the allometric equation of SchmidtNielsen and Larimer (1958) and their Bohr effect falls within typical mammalian values, suggesting that jirds do not have increased Hb affinity for O 2 or enhanced O 2 + unloading at the tissues. Next, I examined the buffering capacity, [H ], [HCO 3 ] and PCO 2 of the blood of jirds and compared them to these of other, burrowing and nonburrowing mammals (from the literature) and with those of the fat sand rat, Psammomys obesus , a Negev Desert rodent of similar body size with close taxonomic affinity to jirds, that may also be exposed to high FCO 2 in their natural environment. Jirds have a slightly higher blood buffering capacity than nonburrowing mammals; however, their buffering capacity is 4

considerably lower than that of the semifossorial fat sand rat and other completely fossorial species, implying that jirds do not have enhanced noncarbonic blood buffering abilities.

Next, I examined the ventilatory response in jirds inspiring air with high FCO 2 by

comparing the increase in their respiration frequency, when breathing CO 2 free air and air with CO 2 = 7%, with that of other mammalian species. I also compared the activity, resting metabolic rate (RMR) and mean metabolic rate ( MR) of jirds inspiring air with CO 2 = 7% to that of jirds breathing CO 2free air. A reduced ventilatory response to air containing high

FCO 2 has been measured in several fossorial and semifossorial species, but jirds did not show reduced ventilatory response when they breathed air containing high FCO 2.

Furthermore, in high CO 2 environment jirds showed recognizable signs of discomfort; they

were more active and slept less than jirds in CO 2free air and as a result increased their MR.

In addition, the RMR of jirds increased significantly when inspiring air containing high FCO 2

suggesting that jirds have to pay an energetic cost in a high CO 2 environment. Finally, I measured the metabolic rate (MR) of jirds during sequential step changes in

inspired FCO 2 (from CO 2 = 0 to CO 2 = 1%, 2%, 4%, 7% and again 0%) and during acute changes in inspired FCO 2 (from CO 2 = 0 to CO 2 = 7% and again 0%). Changes in inspired

FCO 2, whether sequential or acute, did not affect the MR of Sundevall's jirds.

Although Sundevall's jirds are apparently tolerant of high CO 2 in the air they breathe, I found no specific physiological traits to suggest that jirds are especially physiologically adapted to a high CO 2 environments. I conclude that Sundevall's jirds do not naturally

encounter high CO 2 in their burrow environment, as the architecture of their burrows and

their ventilation by the penetration of eddies prevents the built up of CO 2 in them.

Key words: Sundevall's jird, Meriones crassus , fossorial mammal, burrow, burrow

ventilation, P50 , Bohr effect, blood buffer capacity, metabolic rate

General Introduction

General Introduction Burrows are the living environment of species of many taxa, ranging in size from ants (Hölldobler and Wilson, 1990) to aardvarks (WhittingtonJones, 2007). Animals use burrows to avoid predation, find protection from adverse environmental conditions (Walsberg, 2000), raise young, and hoard food (Peinke and Brown, 2005). The presence of burrows affects soil hydrology and dynamics, vegetation patterns, and animal community diversity at scales that range from microsites to landscapes (Greene and Reynard, 1932, Kinlaw, 1999, Whitford and Kay, 1999, Villarreal et al ., 2008). The earliest evidence of burrowing by terrestrial vertebrates goes back as early as 250 million years to the lower Triassic (Damiani, 2003). Burrow building probably had strong adaptive value for the cynodont lineage that ultimately led to mammals (Damiani, 2003). While most mammal species use shelter of some sort ( i.e., the shade of a tree, the lee of a rock) either daily or seasonally, many species construct very elaborate, complex burrows (Reichman and Smith, 1990, Whitford and Kay 1999, Turner 2000, 2001) making them important burrow excavators in the animal kingdom today (Kinlaw, 1999). Rodents are the largest order of mammals, and many species are adapted to life underground. They can be solitary ( e.g., Cape mole rat, Georychus capensis ; Lovegrove and Jarvis, 1986) or social ( e.g., blacktailed prairie dog, Cynomys ludovivianus ; Hoogland, 1995), occupy one burrow for a lifetime ( e.g., great gerbil, Rhombomys opiums ; Randall et al., 2000), or use several burrows ( e.g., springhares, Pedetes capensis ; Peinke and Brown, 2005). Completely fossorial rodents differ from semifossorial ones morphologically, anatomically, and physiologically (Ebensperger and Bozinovic, 2000). Completely fossorial rodents ( e.g., Middle East blind mole rat, Mendelssohn and Yom Tov, 1999) inhabit closed burrow systems; they forage underground and almost never surface, except to disperse. Their burrow systems are structurally different from the openended burrows of semifossorial rodents ( e.g., Gerbils, Gerbillus spp ., Mendelssohn and Yom Tov, 1987) that spend much time foraging on the surface, usually at night (Walsberg, 2000). The architecture of openended burrows varies among and within species. The variety of functions of burrows is associated with their structural complexity, represented by a number of functionally different chambers, connected by a system of tunnels (Degen, 1997). For example, the burrow system of the African ice rat (Otomys sloggetti robertsi ), comprises several interlinking tunnels, numerous burrow entrances and 1 2 nest chambers. They reach an average maximum depth of 26.27 ± 7.7 cm (Hinze et al., 2006). The burrows of kangaroo rats ( Dipodomys spectabilis ) are described as a complicated labyrinth of tunnels. The ejection 6

General Introduction

of refuse and soil from the burrow builds up a mound which has 6 12 entrance holes in it, each with a diameter of 10 14 cm. there are 3 4 large chambers used as storage and a single nest chamber that ends blindly at the depth of about 65 cm, about 0.5 1 m beyond the periphery of the mound itself (Vorhies and Taylor, 1922). Indian desert jirds (Meriones hurrianae ) are crepuscular. They congregate in colonies and inhabit complicated and extensive burrow system with as many as 30 surface openings, which reach a maximum depth of 35 45 cm (Goyal and Ghosh, 1993). The solitary, nocturnal Indian gerbils ( Tatera indica ) occupies simple 'Y' shaped burrows with one or two surface openings and one blind branch for nesting. Their burrows' maximum depth changes seasonally from 35 cm in the winter to 4550 cm in the summer (Goyal and Ghosh, 1993). It has been suggested that burrow geometry, together with the presumably low permeability of soils to gases, and the respiration of the burrow's tenants, may bring about

fractional concentrations of CO 2 (FCO 2) and O 2 (FCO 2) that are, respectively, above and below those in the free atmosphere ( e.g., Studier and Baca, 1968, Chapman and Bennett, 1975, Withers, 1978, Wilson and Kilgore, 1978, Maclean, 1981). Although in the literature burrows are often described as environments of high FCO 2 (Baudinette, 1974, Ar et al., 1979, Boggs et al., 1984, Ar, 1987, BarIlan et al., 1985) it is not clear from published material if this is indeed the case. Extremely low concentrations of O 2 and/or high concentrations of CO 2 have been reported in occupied artificial burrows (e.g., Arieli, 1979 for Middle East blind mole rats Spalax ehrenbergi ; Kuhnen, 1986 for golden hamsters Mesocricetus auratus ).

Conversely, other studies of the gaseous environment of burrows report O 2 and CO 2 levels that only slightly differ from atmospheric concentrations, both in occupied artificial burrows (e.g., Soholt, 1974 for Merriam's kangaroo rat Dipodomys merriami ) and in natural burrows (e.g., Roper et al., 2001 for mole rats Georhychus capensis and Cryptomys damarensis ). In addition, studies of burrow respiratory gas concentrations seldom specify the location in the burrow from where gas samples were drawn in relation to the location of the animal within the burrow system at the time of sampling, so that actual information about the immediate respiratory environment of an animal inside its burrows is missing. In mammals, persistent breathing of air containing high FCO 2 causes chronic respiratory acidosis (Douglas et al., 1979, Schaefer, 1982). Therefore, burrow dwellers are either physiologically adapted to tolerate high FCO 2 in their environment ( e.g., Marder and BarIlan, 1975, Arieli et al., 1977, Arieli and Ar, 1979, Ar, 1987), or alternatively, the architecture of burrows is such that their ventilation prevents buildup of CO 2. Since natural selection has shaped the configuration of rodent burrows for millions of years, they can be regarded as part of the rodent's "extended 7

General Introduction

organism", while their configuration can be considered as part of its "extended physiology"(Turner 2000). Elucidating which of the two alternatives is the dominating mechanism has a strong bearing on our understanding of the interactions between rodents and their environment. I studied ecological and physiological aspects of the life of a semifossorial desert rodent, Sundevall's jird, Meriones crassus Sundevall, 1842. The first part of my study was an investigation of the respiratory environment inside their burrows, specifically the respiratory environment which the animal experiences in its immediate vicinity and what affects it. The second part of this study was an investigation of the physiological attributes that enable

Sundevall’s jirds to tolerate high inspired FCO 2 in the context of the burrow's ostensibly high

[CO 2] environment.

Study animal

Sundevall's jird is one of the most widespread species of Gerbillidae, ranging from Morocco to Pakistan, and occurring in extremely arid environments (Krasnov et al., 1996a). In Israel, jirds are widespread throughout the Mediterranean region, in the Golan Heights and in all except rocky habitats in the central Negev Desert, where they are especially abundant in sand dunes and loess valleys (Mendelssohn and YomTov, 1987, Krasnov et al., 1996a, Krasnov et al., 1996b). The species is considered colonial in some habitats (Harrison and Bates, 1991); however, in Israel it is solitary (Mendelssohn and YomTov, 1987). Sundevall's jirds are semifossorial; they inhabit openended burrows, with 2 12 round openings with a diameter of 4.5 8 cm (Shenbrot et al., 2002), which they sometimes block from inside (Mendelssohn and YomTov, 1987). The length of the burrow tunnels ranges between 200 1000 cm, with 0 3 chambers at a maximum depth of 30 60 cm (Shenbrot et al., 2002). These rodents forage nocturnally and are primarily granivorous (Harrison and Bates, 1991), but shift seasonally to a diet comprising considerable amounts of green vegetation (Degen et al., 1997).

Part 1 Introduction

PART 1: THE RESPIRATORY ENVIRONMENT INSIDE THE BURROW OF SUNDEVALL'S JIRD

1.1 Introduction An animal’s immediate respiratory environment inside a burrow system is the result of several factors combined, which can be divided into three main categories: (1) the resident animal, (2) the physical variables of the burrow system, and (3) the external environmental conditions. The first category includes factors such as the size of the animal (Wilson and Kilgore, 1978, Withers, 1978), its degree of activity and its physiological state, all of which affect its metabolic rate, for example, digging (Ebensperger and Bozinovic, 2000), or lactating (Degen et al, 2011). The animal's location within the burrow system (Wilson and Kilgore, 1978) and the amount of time it spends in any specific place also influence its immediate environment, as does the number of occupants (Studier and Baca, 1968). The second category includes the geometry of the burrow, namely the number of openings, length of tunnels, presence and shape of chambers, depth, etc. (Maclean, 1981), and the properties of the soil, i.e., soil type and porosity, whether the pores are dry or water filled, the presence of physical or biological soil crusts, etc. (Arieli, 1979). The shape of the burrow can also be affected by the animal’s activity, for example when an animal plugs one or more of the burrow’s entrances (Reichman and Smith, 1990, Roper and Kemenes, 1997), or drags nesting material into the tunnels, creating a physical barrier. The third category includes above ground wind velocity (Studier and Baca, 1968) and air temperature (Nikol’skii and Savchenko, 2002, Ganot, 2012). To the best of my knowledge, there are no reports in the literature of measurements of the immediate respiratorygas environment of a semifossorial rodent inside its burrow. No study of burrow respiratory gas concentrations specifies the location in the burrow from where gas samples were drawn (for examples see table 3 in Roper et al., 2001), neither do any report the location of the animal within the burrow system at the time of sampling. Since diffusion of gases through soils is probably too slow to adequately vent gases exhaled by resident rodents, respiratory gas concentrations may rapidly become quite different from atmospheric in an animal's immediate vicinity, rather than in the burrow system as a whole. This view is supported by Wilson and Kilgore's (1978) model that predicts that only about three body lengths on either side of the animal are of importance to its steady state gas exchange. In addition, existing studies do not include information about ambient wind conditions or burrow ventilation. This information is important to understand how ventilation takes place, 9

Part 1 Introduction

especially in localities in the burrow where diffusion may be the primary mechanism of gas exchange between the burrow lumen and the environment. Gas exchange between a burrow and the atmosphere can occur by means of convection, involving an external force, such as wind (Olszewski and Skoczeń, 1965, Vogel et al., 1973, Roper and Moore 2003), by pistonlike movements induced by the movement of the animal within the burrow (Ar A. and Piontkewitz Y. 1992, Ar et al., 2004), by thermal convection as a result of temperature gradients between the burrow and the environment (Kleineidam et al., 2001, Turner, 2001, Ganot et al., 2012), and by diffusion of gas in still air through the burrow and into the soil (Wilson and Kilgore, 1978). The relative contribution of each mechanism to the total gas exchange of openended burrows has not been examined. For example, the contributions of forced convection and thermal convection depend on the geometry of the burrow and on the presence and size of the animal, as well as on properties of the wind for the former and temperature differences for the latter. Diffusion of gas through the soil occurs constantly, through a network of pores that forms a continuous system connecting the surface to the deeper layers of the soil (Hillel, 1998). Wilson and Kilgore (1978) developed a mathematical model in which they considered gas diffusion to be the sole mechanism for exchange and showed that in that case soil porosity is the most important variable affecting the rate of respiratory gas transfer in (O 2) and out (CO 2) of the burrow, more than depth of the burrow or size of the animal. Windinduced convection is most commonly thought of as unidirectional. When the air flows above the ground surface, it enters the burrow through one entrance and leaves it through another. This process can take place due to air pressure differences between burrow entrances as a result of dissimilarity in shape, height or diameter among the entrances. The principle of unidirectional flow was demonstrated using smoke visualization in natural burrows of blacktailed prairiedogs, Cynomys ludovicianus , (Vogel et al., 1973) whose burrows typically have many entrances, with mounds of different forms surrounding each. However, most burrows do not have this kind of architecture. Furthermore, the model for unidirectional flow is based on the notion of laminar air flow at the ground surface, while in reality air flow at the surface is turbulent. The atmospheric boundary layer is the lowest part of the atmosphere and its behavior is mainly affected by its interaction with the ground surface. Flow near the surface encounters obstacles that change the wind speed and introduce random vertical and horizontal velocity components at right angles to the main direction of flow (Dalgliesh and Boyd, 1962). As a result, physical quantities in this layer such as flow, velocity, temperature and moisture 10

Part 1 Introduction

display rapid fluctuations – characteristic of turbulence flow – and strong vertical mixing between layers. Turbulence in the boundary layer consists of irregular gusts of air called eddies (Stull, 1988). ZuckerMilwerger (2013) investigated the pattern of air movements around the two entrances of an artificial burrow. He did pulsechase experiments using propane as a tracer gas and found that propane left the burrow, alternating from one entrance to the other. This pattern is different than expected if flow is unidirectional and, in light of the turbulent nature of air flow in the boundary layer, led me to hypothesize that the underlying mechanism for the ventilation of burrows is by penetration of eddies that convey atmospheric air to the depths of the burrow from both entrances. Interestingly, to my knowledge, there is but a single report in the literature that reports episodic windinduced turnover of air in the single opening, horizontal burrows of the European beeeater, Merops apiaster (White et al. 1978). Other than that, the effect of eddies on burrow ventilation has not been studied.

Research goals

My goal was to obtain a picture of the immediate respiratory environment of a semi fossorial desert rodent in an openended burrow system and to examine the factors that affect this environment. In order to do so I examined some aspects of each of the abovedescribed categories, namely the resident animal, the physical variables of the burrow system, the external environmental conditions, separately and together. I focused on the gas exchange between the nest chamber of a burrow and the atmosphere. The nest chamber is typically a part of the burrow that is deep, deadended, and far from the entrances, but where an animal may spend extended periods of time (for example, when rearing young), and is therefore a location where CO 2 may accumulate, resulting in a high [CO 2] environment for the animal. First I examined the hypothesis that burrow ventilation occurs by several mechanisms working in parallel, whose importance varies depending on ambient conditions and especially wind speed. I tested the predictions that (1) at low surface wind speeds, the predominant mechanism for the ventilation of the nest chamber is gas diffusion through still air into the burrow and through the soil, and (2) when wind speed at the surface is higher, windinduced convection through the openings of the burrow is the predominant mechanism for the ventilation of the nest chamber. I then examined the underlying mechanism by which ventilation by windinduced convection occurs. I tested the prediction that eddies penetrate into the burrow from both entrances, and convey atmospheric air to different depths.

Part 1 Introduction

Next I asked how the resident animal affects its own immediate respiratory environment. I

examined the hypothesis that [CO 2] inside the nest chamber is affected by the proximity of the animal to the chamber. I tested the prediction that [CO 2] inside the nest chamber increases as the animal enters the area close to the nest chamber. Finally, in order to find out what are

some of the most extreme [CO 2] that a Sundevall's jird may encounter within their burrow, I monitored [CO 2] inside nest chambers of burrows while females were present with their broods during the threeweek lactation period. In addition I examined how [CO 2] in the nest chamber was affected by the behavior of the female and by ambient conditions ( e.g., temperature and wind speed).

Methods

1.2 Methods In order to carry out my studies I built two types of artificial burrows based on available data. Since I was interested specifically in the environment of the nest chamber and its ventilation, I chose to treat the chamber and its adjacent openings as a distinct unit. The simplified burrows which I built were designed according to the configuration of a natural burrow of Sundevall's jird in loess habitat, described in figure 2 in Shenbrot et al. (2002). They consist of a main tunnel and a corridor leading to a nest chamber (figure 1 and figure 2) and each entrance is at a distance of about 100 cm from the nest chamber.

Description of artificial burrow, type 1

I built four burrows from 7 mm galvanized mesh, designed according to the configuration of a natural burrow of Sundevall's jird as described by Shenbrot et al., (2002). Each burrow had two entrances, one at each end of a 2 m long main tunnel, with a spherical nest chamber, 13 cm in diameter, in the middle, connected to the former by a 10 cm long corridor (figure 1).

Figure 1. Schematic diagram of artificial burrow, type 1. The figure is not drawn to scale.

The diameter of the burrow entrances, main tunnel and entry corridor was 5 cm. The total volume of each burrow was 5,340 cm 3. The nest chambers were made from two kitchen sieves joined together to form a sphere. The base of the sphere was 60 cm below the soil surface. Each burrow was equipped with two plastic guide tubes (d = 2 cm) for the insertion of CO 2 sensors; one in the nest chamber and the other at the junction of the main tunnel and

Methods

the corridor (henceforth referred to as “the junction”). The nest chambers were also equipped with aluminum tubes (d = 1.6 cm) for the insertion of a camera. All tubes could be plugged with rubber stoppers. Each entrance burrow had a 10 cm Perspex tube (d = 5 cm) inserted into it, allowing it to be plugged with a rubber stopper. The walls of three of the burrows were permeable to gasses, the hardware cloth being wrapped with medical gauze to permit free movement of gases between the burrow and the

soil. In order to study the effect of gas movement by convection alone on [CO 2], three layers of Saran TM plastic wrap and a thick layer of silicon rubber were added to the fourth burrow, making its walls impermeable to gases . The four burrows were buried on my study site on the Sede Boqer campus of Ben Gurion University of the Negev, side by side and parallel to one another, 1 m apart, in sieved loess soil packed to a bulk density of 1.3 g/cm 3, as reported for natural loess soil in the Negev Desert (Shafran, 2005). The axis of the burrows was 300° northwest 120° southeast, parallel to the prevalent wind direction (ZuckerMilwerger, 2013). When I began this project I did several exploratory measurements and I experimented with different measuring techniques and equipment, and with a various artificial burrow configurations. Initially, I injected a gas mixture of 15% CO 2, 85% N 2 into a permeable burrow and into an impermeable one, similar to those described above, but with four

entrances each, and measured the rate at which CO 2 decreased in the nest chamber under the same conditions described in experiment 1 in this thesis (p. 17 under "experimental

settings"). When the entrances of the burrows were open, [CO 2] inside the nest chambers of both burrows decreased rapidly. Within ~15 min [CO 2] was down to 2% and in ~30 min it was not measurably different from the atmospheric concentration. These results led me to the idea that convective ventilation of the burrow may be the dominant mechanism preventing

the buildup of CO 2. In order to test my hypothesis, I further simplified the design of the burrows, by building them with only 2 entrances, suitable to investigate the possibility that natural ventilation plays a role in reducing buildup of CO 2 inside the nest chamber.

Description of artificial burrow, type 2

I built a transparent, artificial burrow from Perspex tubes (d = 5 cm). The burrow had two entrances, one at each end of a horizontal main tunnel 2 m long, with a spherical nest chamber in the middle, at the end of a 10 cm corridor (figure 2). The nest chamber was built as described above. The nest chamber and the connections between the main tunnel and the corridor were covered with a layer of silicon rubber, sealing the burrow completely, so that

Methods the only gas exchange between the burrow and the environment could occur through its entrances. The nest chamber was equipped a plastic guide tube (d = 2 cm) for the insertion of

CO 2 sensor (figure 2).

Figure 2. Schematic diagram of the transparent artificial burrow, type 2.

Measurements of [CO 2] inside the burrows

In the following experiments [CO 2] was measured with insitu infrared CO 2 sensors (CARBOCAP® GMT221 and CARBOCAP® GMT222, Vaisala, Finland) that do not require drawing air samples from the burrow thus avoiding the potential disruption of inside natural air movement . Data was collected on a laptop PC using Microsoft Windows Hyper Terminal.

The measurement range of the GMT221 is 0 1% CO 2 and the range of the GMT222 is 0

10% CO 2. The accuracy of both analyzers at 25 °C and 1 atm. is 1.5% of the range + 2% of reading. Response times of the GMT221 and GMT222 are 20 and 30 seconds, respectively.

Measurements of wind speed

Above ground wind speed and direction were measured 10 m away from the burrows at 0.4 m above the ground with an ultrasonic anemometer (81000, R.M Young Company, USA). According to the manufacturer, the anemometer has a measurement range of 0 40 m/s, a resolution of 0.01 m/s, accuracy of ± 1% rms (root mean square). For complete details see http://www.youngusa.com/products/6/3.html . Wind direction is reported by the direction

Methods

from which it originates. Other wind data was obtained from the Blaustein Institutes for Desert Research meteorological station, located 250 m to the west of the study site (Jacob Blaustein Institutes for Desert Research Desert Meteorology Weather Station, where measurements are made 10 m above ground. http://bidr.bgu.ac.il/BIDR/research/phys/meteorology/default.asp

Measurements of temperature

All thermocouples were calibrated in the laboratory to ± 0.1 °C against a mercuryinglass thermometer with calibration accuracy traceable to the U.S. National Institute of Standards and Technology (NIST) [Catalog numbers 21001 (8 °C to 32 °C) and 21002 (25 °C to 55 °C), Taylor Instrument Co. Rochester, New York]. The transmitters were placed in a water filled glass container in a temperature controlled, circulating water bath (ThermoHaake DC10 and V26, Karlsruhe, Germany). Measured at five temperatures (0, 10, 20, 30 and 40 °C), and individual regression equations of TC temp vs. thermometer temperature were calculated. . Other temperature data was obtained from the Blaustein Institutes for Desert Research meteorological station, located 250 m to the west of the study site (Jacob Blaustein Institutes for Desert Research Desert Meteorology Weather Station, where measurements are made 10 m above ground. http://bidr.bgu.ac.il/BIDR/research/phys/meteorology/default.asp

Injection of CO 2 into the nest chamber

For the slow injection of CO 2 into the nest chambers I used a modified, microcontrolled doublesyringe pump (DN Infusion Pump, Holland). I substituted the usual pair of syringes with two pneumatic pistons (0.39 liter, Baccara, Geva, Israel) in order to be able to inject for

as long as 2 h. The pistons were filled with pure CO 2 from a tank. Refilling, after 2 h, took less than 30 seconds. A threeway stopcock allowed alternation of the flow from the gas tank between the burrow and the pistons.

1.2.1 Experiment 1. Mechanisms for burrow ventilation To assess the relative contributions of convection and diffusion to the ventilation of the burrow’s nest chamber, I used the rate of decrease in [CO 2] inside the chamber of a type 1 burrow (figure 1) as a proxy for burrow ventilation. I did a simulation experiment in which I measured the change in [CO 2] inside the chambers of permeable and impermeable burrows as

a function of time and external wind speed, while simulating the CO 2 output of an adult

Sundevall's jird at rest inside the chamber. This was done by injecting CO 2, at a calculated

Methods

rate, continuously into the center of the nest chamber through a BevALine tube® (d = 1.6 mm) at the end of which there was a thin 6.5 cm long stainless steel tube projecting into the

center of the brood chamber (figure 1) that could not be chewed by the jirds. CO 2 was injected at a flowrate of 2.5 ml/min. This flow rate was chosen because it is close to that of

the CO 2 production of an adult female Sundevall's jird respiring inside the chamber based on my own measurements of resting metabolic rates in jirds, described below (in section 2.3.4 of the thesis). Concurrently, I measured [CO 2] in the upper part of the chamber with an infrared

CO 2 sensor (GMT222) and measured surface wind speed and direction 10 m away from the burrows at 0.4 m above the ground with an ultrasonic anemometer (81000). [CO 2] inside the nest chamber was sampled every minute. Wind speed and direction were measured every second and averaged each minute.

Experimental settings

[CO 2] inside the nest chamber was measured for the following conditions: 1. Impermeable burrow with entrances open. 2. Permeable burrow with: a) entrances plugged with rubber stoppers; b) open entrances;

c) open entrances, and no injection of CO 2 into the nest chamber (control).

When the entrances of the impermeable burrow were open, the decrease in CO2 concentration could only be the result of windinduced convection. When the entrances of the permeable burrow were plugged, gas diffusion in still air through the burrow and into the soil, was the only possible mechanism available for [CO 2] exchange in the nest chamber. When the entrances of the permeable burrow were open, both convection and diffusion could take place, as is usually the case in natural burrows. The fourth condition served as control for baseline [CO 2] levels in an unoccupied burrow. Measurements were made within a 30 days period in May and June 2012. Measurements under each of the conditions 1, 2a and 2b were made on seven nonconsecutive days, while measurements under condition 2c were made on five nonconsecutive days. On any single day measurements were done in a single burrow (permeable or impermeable), this is due to constraints on equipment used in this experiment. Each of the measurements series in a single burrow lasted 8 hours and was done between 0800 1600 h. Half hourly averages were

calculated for [CO 2], wind speed and direction. To the east of the study site (95° 190° from true North) there were several prominent structures that could have caused changes in wind

Methods patterns. Therefore, intervals during which wind blew from their direction were omitted from analysis. It should however be noted that the dominant wind direction throughout the year, is northwest.

1.2.2 Experiment 2. Windinduced ventilation To examine the underlying mechanism of windinduced convection by the irregular penetration of eddies, I quantified the number of eddies that penetrated to different depths in the burrow, the corridor and the nest chamber, at different times of day as a function of wind speed. The method for identifying eddy penetration was based on the change of temperature inside the burrow. Eddy penetration results in in the sudden appearance of a mass of air that causes local changes in air temperatures, provided there is a temperature difference between the inside of the burrow and air temperature above the ground surface. I positioned eight fine (45 SWG) type T thermocouples (TC) along an impermeable, type 1 burrow (figure 3). I used a water level, consisting of a long PVC tube with its ends protruding from the burrow exits to level the entrances of the burrow. Six TCs were positioned along the main tunnel, with their tips in the center of the tunnel; TCs 1SE, 2SE and 3SE were located at 33 cm intervals from the southeastern burrow entrance and TCs 1NW, 2NW and 3W were located at 33 cm intervals from the northwestern burrow entrance. One TC was located in the middle of the corridor leading to the nest chamber, with its tip at the center of the corridor and one TC was located at the nest chamber, with its tip at the center of it. The temperatures of all TCs were measured at a rate of 5 Hz and the data was collected to a data logger (Campbell scientific 23X, Logan UT, USA).

Methods

Figure 3. Schematic diagram of a type 1 impermeable burrow with fine thermocouples (45 SWG). A water level was used to level the entrances of the burrow. The figure is not drawn to scale.

This setup allowed me to quantify the number of eddies that penetrated to different depths in the burrow. A penetration event was defined as a change of more than 0.5 ºC in the temperature of the TCs located in the corridor and in the nest chamber and a 1 ºC change in the temperature of the TCs in the main tunnel. Measurements were made in November 2012 on four nonconsecutive days for 24 h at a time. The number of eddy penetrations per 10 min was calculated using data collected between 0900 1700 h, when the difference between ambient and burrow temperature was high enough to detect eddies. Ten minute averages of wind speed were obtained from the Blaustein Institutes for Desert Research meteorological station, located 250 m to the west of the study site.

1.2.3 Experiment 3. The effect of the presence of an animal on nest chamber [CO 2]

To investigate how [CO 2] in the nest chamber is affected by the proximity of an animal to it, I monitored the change in [CO 2] inside the nest chamber of a type 2 burrow (figure 2) while a Sundevall’s jird was moving freely about it, and recorded the amount of time that the animal spent close to the nest chamber and the amount of time that it spent far away from it.

To account for the effect of wind on [CO 2] inside the nest chamber, the burrow was placed on a table inside a room with no windows or air conditioning, wind speed being negligible for all practical purposes. A female Sundevall’s jird was introduced into the burrow and both entrances were blocked with a 7 mm mesh door to prevent her from leaving. While the animal was inside the burrow, I recorded its movements and behavior with a video camera

Methods

mounted on a tripod directly above it. I marked the length of the tube in 10 cm intervals and recorded the time the animal spent within 30 cm of the junction between the main tunnel, and the corridor on either side of it (figure 2) and the time that the animal spent further than 30 cm from the junction in either direction. In addition, I noted the animal’s position, if it was active or inactive, and the direction and location of its nose when it was inactive. The jird spent 7 hours in the burrow system, and did not fully enter the nest chamber in any of the

events. [CO 2] was sampled every 0.35 seconds at the top of the nest chamber with an infra

red CO 2 sensor (GMT221).

1.2.4 Experiment 4. [CO 2] in the nest chamber while young were being raised

In order to assess what are the actual [CO 2]s that jirds experience inside their burrows, and what external and internal variables affect these concentrations, I used four type 1 burrows (figure 1), three of which were permeable and one impermeable. I introduced pregnant female jirds into the burrows and, after the females gave birth, I measured [CO 2], burrow temperature, Ta and wind speed at the surface during the 3week lactation period.

[CO 2] was measured every 1min in the upper part of the nest chamber and at the roof of

the junction between the main tunnel and corridor, with the infrared CO 2 sensors (respectively GMT221 and GMT222). Burrow temperatures were measured in six places in the burrow with TypeT TC's held by small plastic tubes secured to the ceiling; one TC was in the ceiling of the nest chamber, and five TCs were positioned, at 33 cm intervals, along the main tunnel, with the middle TC at the junction (figure 1). Ten minute temperature averages were collected using the Loggernet program and a data logger (21X Campbell Scientific, Logan UT, USA). I was unable to apply the method of recording temperature changes in order to measure ventilation (described in section 1.2.2) because it requires that the TCs be positioned with their tips in the center of the tunnel, where the jirds would most likely damage them. Tenminute averages of wind speed and Ta measurements were obtained from the standard meteorological station, located on the Sede Boqer Campus of BGU. To allow the jirds to forage above ground, the entrances of the type 1 burrow all opened into bird cages (figure 4). One entrance opened into a small cage (38.5 x 25.5 x h 37 cm) and the other into a large cage (87 x 46.5 x h 70 cm). The floor of the cages had been removed and replaced with 7 mm galvanized steel mesh to prevent the jirds digging their way out. The four burrows were buried parallel to one another on the Sede Boqer campus of Ben Gurion University of the Negev.

Methods

Figure 4. Schematic diagram of a type 1 burrow with exits into bird cages, to allow the jirds to forage above ground and to prevent their escape. Figure is not drawn to scale. See text for details.

One pregnant female Sundevall’s jird was introduced into each burrow. Animals were taken from the breeding colony maintained by Prof. Boris Krasnov and Dr. Irina Khokhlova on the Sede Boqer Campus of BGU (see details in section 2.2 of this thesis). On day zero, 6 females from the colony were paired with 6 males with which they had already successfully bred 6 months earlier. On day 7 the males were removed from the females' cages. From day 7 through day 14 the females were weighed every other day. On day 14 the four females who gained the most body mass were introduced into the 4 burrows. The gestation time of Sundevall’s jird is 22 24 days (Mendelssohn and YomTov, 1987); on day 14 a female, if pregnant, would be at either a minimum of 7 days or a maximum of 14 days into her pregnancy, allowing her between 8 to 17 days to become accustomed to her new burrow residence before giving birth to a litter. Females were introduced into the burrows on September 1st, 2012 between 17:00 and 17:30 in the evening. Each female was placed in the large bird cage and found its way into the burrow within 20 minutes. In the large cages the jirds were provided with nesting materials (long dry blades of grass and paper towels) and millet seeds ad libitum . Freshly cut alfalfa leaves were provided daily. Starting on day 21 the nest chambers were inspected daily

Methods with videoprobe camera (SSVR710 Snakescope TM , Advanced Inspiration Technologies, China) through the designated tube to check for the presence of pups. I began measurements on the first day on which pups were observed inside the nest chamber (the day of birth). Measurements were made in only one burrow per day. Each measurement lasted for 23 hours from 22:00 to 21:00 on the following day.

Results

1.3 Results

1.3.1 Experiment 1. Mechanisms for burrow ventilation The data presented in figure 5 are halfhour averages of wind speed and nest chamber

[CO 2], measured in permeable and impermeable type 1 burrows (figure 1) under three conditions. When convection was the only possible mechanism for the ventilation of the nest chamber (figure 5A), the half hour average [CO 2] decreased exponentially with wind speed (Spearman rank order correlation coefficient 0.802, p < 0.001), so that at low wind speed

(> 2 m/s) small changes resulted in large decreases in chamber [CO 2]. At wind speeds > 3 m/s further increases did not affect [CO 2] in the chamber. The maximum [CO 2] measured in the impermeable burrow (under condition 1) was 25,660 ppm, six times the atmospheric

[CO 2] of ~390 ppm (NOAA, 2012). This value was measured when wind speed was ~1 m/s blowing from ~50 ° northeast. When diffusion in still air through the burrow and into the soil was the only possible mechanism for the ventilation of the nest chamber (figure 5B), as predicted, chamber [CO 2] was independent of wind speed. Nest chamber [CO 2] ranged between 8600 10400 ppm. When both mechanisms were active (figure 5C), [CO 2] decreased linearly with wind speed from 8,600 ppm at 1 m/sec to 2,800 ppm at 3.9 m/sec

(Spearman rank order correlation coefficient 0.718, p < 0.001). When I did not inject CO 2 into the nest chamber, thereby simulating an unoccupied burrow, atmospheric [CO 2] (NOAA,

2012) was 2.91 standard deviations less than the mean [CO 2] in the nest chamber (606.07 ppm) (figure 6). There was a strong correlation between the patterns of wind speed (Spearman rank order correlation ρ = 0.816, p < 0.001, figure 7) and wind direction (angular correlation coefficient = 0.708, p = 0.014, figure 8) 1 on days when convection through the openings was the only mechanism for ventilation and when both convection and diffusion took place at the same time. Therefore it is possible to compare between the two data sets which were collected on different days.

1 These analyses were done in R (R development core team, 2011), with the package "circular" (Jammalamadaka and Sarma, 1988, Jammalamadaka and SenGupta, 2001). 23

Results

Figure 5. Halfhour averages of CO 2 concentrations [CO 2] inside the nest chamber of artificial Sundevall’s jird burrows as a function of wind speed during continuous injection of CO 2 into the nest chamber. A decrease in [CO 2] is a measure for the ventilation of the nest chamber as a result of the following mechanisms: (A) convection through the openings of the burrow, (B) gas diffusion in still air through the burrow and into the soil, and (C) convection and diffusion taking place at the same time. The empty circles ( ) in 5A are data points for one day of measurements (28.5.2012) and are consistently lower than the other data points measured in the same set up. When convection was the only mechanism for ventilation (5A), external wind speed had a substantial effect on [CO 2] in the nest chamber; [CO 2] was 5 6 times higher in the nest chamber at low wind speed (< 2 m/s) than it was at high wind speed (> 3 m/s). When gas diffusion was the only mechanism for ventilation (5B), external

wind speed did not affect chamber [CO 2]. When both mechanisms for ventilation were simultaneously enabled (5C) [CO 2] at low wind speed

was considerably lower than when ventilation occurred by convection alone. At wind speeds > 3 m/s, the [CO 2] in both cases was similarly low.

Results

Figure 6. Control measurments of CO 2 concentration [CO 2] inside the nest chamber of an artificial Sundevall’s jird burrow. CO 2 was sampled every 1 min on five nonconsecutive days between 0800 – 1600 h. The red line is atmospheric [CO 2] (NOAA, 2012), 2.909 standard deviations lower than the mean [CO 2] in the nest chamber (606.07 ppm).

Figure 7. Half hour averages of wind speed for 14 days of measurements in which [CO 2] inside the nest chambers of artificial Sundevall’s jird burrows was measured as a function of

ambient wind speed, during continuous injection of CO 2 into the nest chamber. Decreasing [CO 2] (washout) was used as a measure of nest chamber ventilation. Full circles denote the means of seven days of measurement in which ventilation resulted only from convection through the openings of the burrow. Empty circles denote the means of seven days of measurement in which ventilation resulted from a combination of winddriven convection and diffusion through still air through the burrow and into the soil. Values are means ± SD. There is a strong correlation between the patterns of wind speed on the different days of measurement (Spearman rank order correlation, ρ = 0.816, p < 0.00).

Results

Figure 8. Halfhour averages of wind direction for the 14 days on which [CO 2] inside the nest chambers of artificial Sundevall’s jird burrows was measured as a function of ambient wind

speed, during continuous injection of CO 2 into the nest chamber. Decreasing [CO 2] (washout) was used to measure the ventilation of the nest chamber as a result of either winddriven convection through the openings of the burrow, or of simultaneous convection and diffusion in still air through the burrow and into the soil. The patterns of wind direction on the different days of measurement are correlated (angular correlation coefficient = 0.708, p = 0.014).

Results

1.3.2 Experiment 2. Windinduced ventilation I examined the potential effects of eddy penetration into burrows on the burrow's ventilation regime. The data analysis was made for data series obtained between 0900 h and 1700 h, a period during which air temperature outside is higher than that inside the burrows. Each eddy is essentially a mass of air that due to its quick movement does not change its properties (temperature, water vapor concentration, etc.) thereby conveying the average properties prevailing in the layer of origin of the eddy into the burrow as a result of which there is sequential, momentary rise in temperatures starting at the first TC location, closest to the entrance, and advancing along the tunnel deeper into the burrow (figure 9).

Figure 9. Temperatures at six locations along the main tunnel of an impermeable, type 1 artificial Sundevall's jird burrow as a function of time , illustrating penetration of an eddy during daytime. The eddy moves atmospheric air that is warmer than burrow air into the burrow causing a sequential, momentary rise in temperatures along the main tunnel, starting at the first location, 1NW, closest to the northwestern entrance (blue line) and advancing deeper along the tunnel to locations 2NW (light blue line), 3NW (green line), 3SE (yellow line), 2SE (orange line) and 1SE (red line). The black arrows indicate the successive peak temperatures of the thermocouples and are indicative of eddy movement.

Results

Figure 10. Temperatures in six locations along the main tunnel of an impermeable, type 1 burrow as a function of time , illustrating the irregular nature of ventilation by eddies; The eddy moves atmospheric air that is warmer than burrow air into the burrow causing a sequential, momentary rise in temperatures along the main tunnel. The penetration of eddies alternates between the two entrances. For the same time sequence; the top figure shows a penetration of an eddy from the northwestern entrance, starting at location 1NW (blue line), and advancing deeper along the tunnel to locations 2NW (light blue line) or and 3NW (green line). The bottom figure shows a penetration of an eddy from the southeastern entrance starting at location 1SE (red line), and advancing deeper along the tunnel to locations 2SE (orange line) and 3SE (yellow line).

Eddies penetrated the burrow from either one of the two entrances (figure 10) but never simultaneously. The duration of each eddy and the intervals between eddies varied. Some eddies penetrated partially into the burrow while others moved through it and exited via the other opening. The reason for this behavior was probably the result of the different speeds of the eddies, which were not measured. The number of ventilation events which were recorded

Results in the main tunnel of the burrow, (at sites 3NW and 3SE in figure 3) was calculated as the total number of peak temperatures recorded irrespective of the opening through which the eddy entered the burrow. The frequency of events increased with wind speed up to 4.5 m/s (figure 11A); above that speed, penetration decreased over the range of speeds measured. More eddies reached the northwestern part of the burrow (locations 3NW) than the southeastern part of the burrow (3SE). This was probably due to the fact that in the geographic region where measurements were made, northwest is the dominant wind direction throughout the year and despite the alternating nature of penetrations more eddies entered from that direction. Considerably fewer eddies reached the corridor leading from the main tunnel to the nest chamber and no eddies were detected inside the nest chamber (figure 11B).

Figure 11. The average number of ventilation events per 10 minutes at locations 3NW, 3SE, corridor and nest chamber of an artificial Sundevall's jird burrow as a function of wind speed intervals, between 0900 h and 1700 h. The average number of ventilation events (A) 99 cm from the southeastern (full circles) and northwestern (empty circles) entrances and (B) in the middle of the corridor (full circles) and in the nest chamber (empty circles). Values are averages ± SD.

Results

1.3.3 Experiment 3. The effect of the presence of an animal on nest chamber [CO 2]

Figure 12. [CO 2] in the nest chamber of an artificial, type 2 Sundevall’s jird burrow (grey jagged line), with an adult female present, and her behavior (blue and red line) as a function of time. The white areas are times when the jird was near the nest chamber, namely < 30 cm on either side of the junction between the main tunnel and the corridor. The lightly shaded areas are times when the animal was far from nest chamber. Blue on the behavior line indicates when the animal was active and red indicates when it was inactive.

As the animal moved away from the site where [CO 2] was measured, the [CO 2] immediately decreased (figure 12, gray shaded areas). An increase in [CO 2] occurred only when the animal moved closer to the area of measurement (figure 12, white areas). From the data used to prepare figure 12, I calculated clearance constants and the rate of increase in

[CO 2] as related to the location of the animal close to, and far from, the nest chamber. Three periods in which the animal was far from the nest chamber and [CO 2] decreased by over 50%

(marked with lower case letters, ac), were used to calculate [CO 2] clearance constants by exponential fitting (table 1). Six sections in which the animal spent over 20 minutes in the area close to the nest chamber (marked with upper case letters, AF) were used to calculate a the rate and increase of [CO 2] (tables 2).

Results

Table 1. Clearance constants for the elimination of CO 2 in the nest chamber of an occupied artificial Sundevall’s jird burrow. [CO 2] decreased when the animal left the area defined as ‘close to the nest chamber’; these events are marked with lower case letters in figure 9. See text for further details.

Clearance Section on constant figure 9 (sec 1)

0.002 a

0.001 b

0.001 c

Table 2. Rates of increase in [CO 2] inside the nest chamber of an occupied artificial Sundevall’s jird burrow and the values used for its calculation. [CO 2] increased when the animal was active or inactive inside the area defined as ‘close to the nest chamber’; these events are marked with upper case letters in figure 9. See text for further details.

Rate of [CO 2] Time Section on decrease (ppm) (Hours) figure 9 (ppm/sec) 4,450 1.02 A 8.5 18,160 1.46 Not active 15,510 1.73 B 8.16 33,730 2.35 Not active 10,080 3.47 C 4.97 17,840 3.87 Not active 14,800 3.93 D 7.89 23,900 4.25 Not active 10,290 4.59 E 11.25 21,310 4.87 Active 21,310 4.87 F 1.49 23,360 5.25 Not active

Results

1.3.4 Experiment 4. [CO 2] in the nest chamber while young were being raised

During September 2012, I measured [CO 2] and temperature in three permeable burrows and one impermeable burrow for 19 days. Females gave birth to litters of 3 5 pups (table 3).

Table 3. The number of pups which were raised in the nest chamber of each burrow, their date of birth and the number of days of measurements that were done in each burrow

Burrow type and Litter Number of days Birthdate number size of measurements Permeable burrow 1 3 12/9 3 Permeable burrow 2 4 5/9 6 Permeable burrow 3 4 8/9 3 Impermeable burrow 5 14/9 7

The pups were all born inside the nest chambers and did not leave them for the first eight days of their lives. The chambers were padded with nesting materials, and probably some nesting material was present in the corridors and main tunnels. Three to four days after parturition all the females plugged one of the entrances to the burrow with vegetation and nesting materials; in all cases the entrance that did not open into the cage with the food. From day 8 onward, the pups in the impermeable burrow were sometimes found in the main tunnel; the pups in the permeable burrows were almost always found in the nest chamber. On several events, when I inserted the fiberoptics probe of the camera that was equipped with a soft white light into the chamber, the pups made their way out of the nest chamber into the corridor or tunnel, sometimes with the encouragement of their mother. All pups survived the three weeks of measurement. Measurements were made in the four burrows sequentially. For each burrow, day 1 was the day of parturition. Day 0 was a day of measurements made in the impermeable burrow four days before parturition.

Results

Figure 13. The average minimum and maximum [CO 2] (bottom and top numbers in box, respectively) in three permeable and one impermeable artificial Sundevall’s jird burrow, and the absolute maximum value of [CO 2] in each burrow (on the upper whisker).

[CO 2] in the permeable burrows ranged between 2,000 22,000 ppm, namely, 5 50 fold that in the free atmosphere. The highest [CO 2]s were recorded in the impermeable burrow; the maximum was 82,945 ppm; 213 fold that in the free atmosphere (figure 13). In both types of burrow, the nest chamber [CO 2] fluctuated periodically (figures 14A and 15A) with an average frequency of 0.9 ± 0.24 peaks per hour in the three permeable burrows (n = 9), and an average frequency of 1.16 ± 0.21 peaks per hour in the impermeable burrow (n = 7). [CO 2] at the junction followed a similar pattern (figures 14A and 15A). I think that these fluctuations are related to the animal’s movements in and out of the nest chamber, see rationale for movement below. The daily pattern of wind speed was similar on all days of measurement (figures 14B and 15B), with an average maximum of 7.3 ± 0.21 m/sec in the afternoons between 1700 h and 1800 h, and an average minimum of 1.73 ± 0.33 m/sec during the night and in the early hours of the morning, between 0300 h and 0800 h.

Results

The daily pattern of Ta was similar on all days of measurement (figures 14C and 15C), with an average maximum of 29.96 ± 0.05 °C in the early afternoon between 1200 h and 1600 h, and an average minimum of 19.42 ± 0.8 °C at night and in the early hours of the morning, between 0200 h and 0600 h. The temperatures inside the nest chamber remained fairly constant throughout the three weeks of measurements (min 27.84 °C and max 32.27 °C) with small fluctuations (± 1.5 °C) during any single day. The minimum and maximum temperatures measured inside the nest chamber and at the junction are shown in table 4.

Table 4. The average minimum and maximum air temperatures measured outside and inside the nest chamber and the junction of the permeable and impermeable artificial Sundevall's jird burrows. See text for details.

Location Maximum Minimum temperature °C temperature °C Permeable Nest chamber 33.08 28.75 burrows junction 31.57 28.03 Impermeable Nest chamber 30.35 28.83 burrow junction 29.10 27.50

The daily measurements of [CO 2] inside the nest chambers of the permeable (figure 16A

C) and impermeable (figure 17A) burrows provide a sense of the daily cycle of [CO 2] that the dam and her pups experienced inside the nest chambers and highlight the resemblance of the three permeable burrows to one another, and also the differences between them and the impermeable burrow. The average values (figure 16D and figure 17B) filter out the

fluctuations in [CO 2] that are related to the animal’s movements in and out of the nest chamber, and provide a clearer picture of the daily [CO 2] cycle in the burrow along with the daily pattern of external wind speed.

The daily pattern of [CO 2] in the impermeable burrow was different from one day of measurements to the next, although the daily pattern of wind speed was the same. For example, on day 12 (figure 18A) [CO 2] reached a (very) high maximum value of 81,910 ppm around 0600 h hand then dropped sharply from 0700 0800 h. Subsequently, [CO 2] fluctuated between 5,000 ppm and 50,000 ppm. On day 16 (figure 18B), from 2300 1200 h,

[CO 2] was continuously high, with relatively small fluctuations, and reached a maximum of

65,000 ppm. [CO 2] decreased between 1200 1600 h and then increased again, reaching a maximum of 51,200 ppm around 1800 h.

Results

Figure 14. A typical example of 23 h of measurement in a permeable artificial Sundevall’s jird burrow (permeable burrow 2 on day 16). (A) 1 min measurements of

[CO 2] inside the nest chamber (black line) and at the junction between the main tunnel and the corridor (grey line). (B) 10 min averages of wind speed. (C) 10 min averages of temperature measurements outside the burrow (red line), in several locations along the main tunnel (orange, yellow, blue and purple lines), including the junction between the tunnel and the corridor (turquoise line), and in the nest chamber (green line). See text for details.

Results

Figure 15. A typical example of 23 h of measurement in an impermeable artificial Sundevall’s jird burrow (impermeable burrow on day 12). (A) 1 min measurements of

CO 2 concentration inside the nest chamber (black line) and at the junction between the main tunnel and the corridor (grey line). (B) 10 min averages of wind speed. (C) 10 min averages of temperature outside the burrow (red line), in several locations along the main tunnel (orange, yellow and blue lines) including the junction between the tunnel and the corridor (turquoise line) and in the nest chamber (green line). See text for details.

Results

Figure 16. [CO 2] as a function of time of day inside the nest chambers of three permeable, artificial Sundevall's jird burrows on 12 different days of measurement (AC), and the means ± SD of these measurements (D). (A) Permeable burrow 1 (B) Permeable burrow 2 (C) Permeable burrow 3

(D) Average ± SD [CO 2] (black and gray lines respectively) and average windspeed during all days of measurement (dotted line). Day 1 is the day of the parturition. The purpose of this figure is to provide a sense of the daily cycle of [CO 2] that the dam and her pups experienced inside the nest chambers and highlight the resemblance of the three permeable burrows to one another. The average values (D) filter out the fluctuations in

[CO 2] that are related to the animal’s movements in and out of the nest chamber, and provide a clearer picture of the daily [CO 2] cycle in the burrow along with the daily pattern of external wind speed.

Results

Figure 17. (A) [CO 2] as a function of time of day inside the nest chamber of an impermeable artificial Sundevall's jird burrow on 7 different days, (B) the means ± SD of these measurements (black and gray lines respectively), and mean windspeed during all days of measurement (dotted line). Day 1 is the day of the parturition; day 0 was a day of measurements made in the impermeable burrow 4 days before parturition.

The purpose of this figure is to provide a sense of the daily cycle of [CO 2] that the dam and her pups experienced inside the nest chamber of the impermeable burrow. The average values

(B) filter out the fluctuations in [CO 2] that are related to the animal’s movements in and out

of the nest chamber, and provide a clearer picture of the daily [CO 2] cycle in the burrow along with the daily pattern of external wind speed.

Results

Figure 18. [CO 2] inside the nest chamber of an impermeable artificial Sundevall's jird burrow (solid line) and wind speed (dotted line) as functions of time of day, on (A) day 12 and (B) day 16.

Discussion

1.4 Discussion To understand the immediate respiratory environment of a semifossorial desert rodent in a nest chamber of an openended burrow system, I first identified the factors that affect this environment and divided them into three categories: (1) the physical variables of the burrow system, (2) external environmental conditions, and (3) the resident animal, which is the source of CO 2 inside the burrow. The two distinct paths for CO 2 to leave the burrow are through the burrow’s openings and via the soil surrounding the burrow.

Mechanisms of burrow ventilation

To test my first hypothesis, that burrow ventilation occurs by several mechanisms working concurrently, the predominant one depending on ambient conditions, particularly wind speed,

I used changes in nest chamber [CO 2] as an indication of the degree of burrow ventilation. I measured [CO 2] as a function of ambient wind speed in burrows that were ventilated either by wind induced convection or by diffusion of gases in the still air of the burrow and through the soil, or by both together. I did not expect thermal convection to play an important role in the ventilation of the nest chamber during daytime, when I made these measurements in unoccupied burrows, because in the Negev Highlands daytime air temperatures in unoccupied burrows are significantly lower than ambient air temperatures (Shenbrot et al., 2002; my own unpublished data). I simulated a breathing animal inside the nest chamber, and thus I was able to neutralize confounding factors such as the animal’s location and its degree of activity, which would constantly change if the burrow was occupied. When convection was the only mechanism for ventilation (figure 5A), external wind speed had a substantial effect on [CO 2] in the nest chamber; [CO 2] was 5 6 times higher in the nest chamber at low wind speed (< 2 m/s) than it was at high wind speed (> 3 m/s). The empty circles ( ) in figure 5A are data points for one day of measurements (28.5.2012) which are consistently lower than the other data points measured in the same set up, but still follow the same pattern of decrease. No satisfactory explanation could be found to explain the low values of this set of points. As expected, when gas diffusion was the only mechanism for ventilation (figure 5B), external wind speed did not affect chamber [CO 2]. When both mechanisms for ventilation were simultaneously enabled (figure 5C), as I assume is the case in natural burrows, [CO 2] at low wind speed was considerably lower than when ventilation occurred by convection alone. At wind speeds > 3 m/s, the [CO 2] in both cases was similarly low.

Discussion

There is apparently no one "most important" mechanism for burrow ventilation, but rather burrow ventilation occurs by several mechanisms the importance of which varies, even throughout the day, depending on ambient conditions and especially wind speed. At low wind speeds – for example, at night and in the early morning hours – diffusion in still air through the burrow and into the soil may be the dominant mechanism for ventilation, and soil porosity becomes an important factor when considering the ventilation of deadended spaces, deep and far from entrances. The permeability of the soil to gas diffusion may change due to the presence of water that completely or partially clogs the pores that permeate the soil (Hillel, 1998). Further, reduction in pore diameter and pore size distribution of the upper soil layer may be caused by the impact of raindrops that create an abiotic crust (Carmi and Berliner, 2009, Morin and Van Winkel,1996), and/or by its colonization by bacteria and fungi (reviewed by Belnap and Troxler, 2006) that form a biological crust, thereby reducing soil diffusivity. My results show that at wind speeds above 2 m/s, burrows are likely to be well ventilated

and, that as long as the burrow's openings are not plugged, accumulation of CO 2 may not present a major problem for the resident animal, even in cases when soil porosity is reduced. To the best of my knowledge there are no studies that include information about surface wind speed at the time of sampling respiratory gases in animal burrows. For openended burrow systems this information is critical, and reports of mean values of burrow gas concentration may have less scientific value.

Windinduced ventilation

My finding, that burrow ventilation is correlated with external wind speed, is consistent with that of Roper and Moore (2003) who showed that wind induced air movements penetrated to a distance of at least several meters in badger setts, and were strongly, and positively correlated with external wind speed. Roper and Moore suggested that if air moves into the entrances of badger setts it must do so by direct penetration of wind currents, and not by suction of air into the burrow due to pressure differences between burrow entrances as in the case of burrows of blacktailed prairiedogs (Vogel et al., 1973). However, they did not explain what mechanism underlies this phenomenon. Counting the number of eddies that penetrated to different depths and locations in the burrow as a function of external wind speed supports the prediction that eddies penetrate the burrow from both entrances (figure 10), and move atmospheric air to different depths (figure

Discussion

11). While the number of eddies that penetrated the main tunnel of the burrow increased with wind speed for winds of up to 4 4.5 m/s (figure 11 AC), at higher wind speed the number of eddy penetrations decreased. Although I detected relatively few eddies inside the corridor leading from the main tunnel into the nest chamber, and none inside the nest chamber (figure 11D), I suggest that these parts of the burrow remain well ventilated due to large, and almost constant, gas (CO 2 and perhaps O 2) concentration differences, which are maintained between the main tunnel and the nest chamber, since each eddy brings atmospheric air into the tunnel. This gradient facilitates the diffusion of CO 2 from the chamber into the main tunnel where it is removed by the next eddy. By this mechanism locations in the burrow that are deep and far from entrances, even

though not directly ventilated by wind, have relatively low [CO 2] accumulation. When external wind speed is very low, for example at night, the animal movement induced mixing of air within the burrow ( i.e. the socalled piston effect) could be of importance for ventilation, as well as thermal convection and diffusion through the entrances and the soil.

The effect of the presence of an animal on nest chamber [CO 2]

To investigate how [CO 2] in the nest chamber is affected by the proximity of the animal to it, I monitored the change in [CO 2] inside the nest chamber of an impermeable, transparent burrow (figure 2), while a Sundevall’s jird was moving about it freely, and recorded the amount of time that the animal spent close to the nest chamber and the amount of time that it spent far away from it.

[CO 2] inside the nest chamber began to decrease immediately when the animal moved away from the area defined as ‘close to the nest chamber’ (figure 12). The average 1 elimination time constant was 0.0013 sec (table 1). [CO 2] inside the nest chamber began to increase immediately when the animal came into the area ‘close to the nest chamber’. The wide range in these rates (table 2) is apparently due to the specific location of the animal within the area ‘close the nest chamber’ and whether the animal is active or not.

The location of the animal’s nose with respect to the CO 2 probe greatly affects [CO 2]. For example, on events A, B and D (figure 12 and table 2) the animal was at rest with its nose at the junction (figure 19, case I) and the rate of increase in [CO 2] in all these cases was similar and relatively high (7.89 8.50 ppm/sec). On events C and F (figure 12 and table 2) although the animal was within the area defined as “close to the nest chamber” it was at rest with its

Discussion nose located as far as 10 cm from the junction (figure 19 case II); accordingly the rate of increase in [CO 2] was relatively low (respectively 4.95 and 1.48 ppm/sec). In event E (figure 12 and table 2) the animal was active within the ‘area close to the nest chamber’ and spent some time inside the corridor leading to the nest chamber (figure 19, case III); in that case the rate of increase in [CO 2] was the highest (11.25 ppm/sec) measured.

Figure 19. The location of the animal’s nose with respect to the measuring device located in the burrow’s nest chamber. The location of the

animal’s nose with respect to the CO 2 probe greatly affects CO 2 concentration ([CO 2]). (I) When the animal was at rest with its nose at the

junction the rate of increase in [CO 2] was relatively high (7.89 8.50 ppm/sec). (II) When the animal was at rest within the area defined as “close to the nest chamber” with its nose located as far as 10 cm from the junction the

rate of increase in [CO 2] was relatively low (1.48 4.95 ppm/sec). (III) When the animal was active within the ‘area close to the nest chamber’ and spent some time inside the corridor leading to the nest chamber the rate of

increase in [CO 2] was the highest (11.25 ppm/sec) measured.

Discussion

Studies of respiratory gas concentrations in burrows seldom include information about the location of the animal within the burrow system at the time of sampling, probably due to the technical difficulties of accessing burrows nondestructively (Roper et al., 2001). My results indicate that knowledge of the whereabouts of the animal within the system is crucial, since the respiratory gas concentrations in the immediate vicinity of the animal are very different than those in the burrow system as a whole (Turner, 2000). Roper et al., (2001) suggested

that reported mean values for withinburrow FO2 and FCO 2 may not reflect the respiratory environment that the animal actually experiences.

[CO 2] in the nest chamber while young were being raised

Ultimately, to elucidate the actual, most extreme levels of [CO 2] that jirds might encounter

inside their burrows, taking into account many of the relevant factors that affect [CO 2], I introduced four pregnant female jirds into 4 artificial burrows (figure 4) and, after they gave birth (table 3), I measured [CO 2] and temperature in several locations along the burrow, as

well as external wind speed and Ta for the duration of the 3week lactation period. I consider the three permeable burrows to be close representations of natural burrow

systems. [CO 2] in these burrows ranged between 2,000 22,000 ppm, 5 50 fold that in the

free atmosphere (figure 13). The most extreme [CO 2] values were recorded in the impermeable burrow with a maximum value of 82,945 ppm, 213 times that in the free atmosphere (figure 13). Although this burrow was very different from a natural burrow, measurements from the impermeable burrow afford an idea of the most extreme [CO 2] that an animal might encounter. These conditions could prevail in parts of the burrow in cases of extreme reduction in soil porosity, for example after a flood. The fact that a female jird lived for three weeks in these conditions, and successfully raised a brood of five pups (table 3) suggest that although such [CO 2] may not be common in rodent burrows, Sundevall’s jirds are adapted to cope with them physiologically, as will be further addressed in part two of this thesis.

[CO 2] values measured in this experiment were considerably higher than those measured in the first experiment, where the animal’s presence in the nest chamber was simulated by injecting CO 2. There are two reasons first, the simulated metabolic rate used in the first experiment was a resting metabolic rate that is as much as four times lower than the metabolic rate of a lactating female Sundevall's jird (Kam et al., 2003, Degen et al., 2011).

Also, the pups in the chamber, especially as they grow, add to the total CO 2 in the nest chamber. Second, nesting materials were present in the nest chambers, the corridors leading

Discussion

to the chambers, and probably in part of the tunnels as well. In addition, on most days of measurement, one of the entrances to each of the burrows was plugged by the mother. This situation more likely reflects conditions in natural burrows of Sundevall’s jirds. Nevertheless, since natural burrows of jirds have several entrances (Shenbrot et al., 2002) and females are not known to plug them all at once, the question of air flow and ventilation of the tunnels through two openings remains relevant.

The daily pattern of change in [CO 2] in the permeable burrows was very different from that in the impermeable burrow (figures 12 and 13). However, in both burrow types, nest chamber [CO 2] fluctuated periodically and [CO 2] at the junction followed suit. I suggest that these fluctuations are directly related to the female’s movements in and out of the nest chamber as shown in experiment 3 (figure 12). When the female was inside the nest chamber

with the pups, [CO 2] in the chamber increased, and when the female left the nest chamber

[CO 2] decreased.

The daily pattern of change in [CO 2] in the nest chambers of the three permeable burrows

(figure 16D) suggests that [CO 2] generally peaks between 0300 h and 0600 h, and again between 1000 h and 1200 h. Between the peaks (0600 h and 1000 h), there is a period of lower [CO 2]. This implies that the daily period of lower chamber [CO 2] may occur due to the activity pattern of the females. Sundevall’s jirds, like many other rodents species, are crepuscular in their foraging activity (Degen, 1997). This pattern of daily lull in chamber

[CO 2] is most conspicuous in permeable burrow 2 on days 5, 11, 16 and 18 (figure 16B), and in permeable burrow 3 on days 1 and 4 (figure 16C). For several hours during the morning,

on each of these days, [CO 2] in the nest chamber is low and with relatively few fluctuations.

The steady, low, [CO 2] in the nest chamber indicates that during this time period the female is far from the chamber and I assume that at this time of day she was outside “foraging” in the feeding cage, or resting in the main tunnel far from the nest chamber. From the daily pattern

of change in [CO 2] in the nest chamber of the impermeable burrow (figure 17B) it appears that the female spent the night in the nest chamber with the pups and daytime away from the chamber.

The daily patterns of Ta were similar on all measurement days ( e.g., figures 14C and 15C).

Ta peaked in the early afternoon and reached a minimum at daybreak. Inside the burrow the temperatures became more constant with distance from the entrances. This is a well documented phenomenon (Kennerly, 1964, Nikol’skii and Khutorskoi, 2001, Shenbrot et al., 2002). The temperatures inside the nest chamber remained fairly constant throughout the three weeks of measurements with small fluctuations during any single day. Unlike the case 45

Discussion

of unoccupied M. crassus burrows (Shenbrot et al., 2004), temperatures inside the nest chambers in this study were higher than Ta all day long, suggesting that the presence of the female and litter caused a considerable rise in temperature in their immediate environment due to their metabolic heat production. The steep temperature gradient between an occupied chamber and the ambient air could potentially promote and enhance ventilation by thermal convection (Kleineidam et al., 2001, Turner, 2001, Ganot et al., 2012). However, it appears that this is not an important mechanism for ventilation in jird burrows, since maximum [CO 2] values were reached at night before sunrise, in spite of the steep gradient between ambient and burrow temperatures. The daily patterns of wind speed were similar on all days of measurement (e.g., figures 14B and 15B). Wind speed reached its maximum in the late afternoon and minimum before

dawn. The average values of wind speed and [CO 2] inside the nest chamber of the impermeable burrow shown on figure 17B suggests that there is a strong, negative correlation between [CO 2] in the nest chamber and wind speed, however, this is probably misleading.

Figure 18 shows the [CO 2] inside the nest chamber of the impermeable burrow and the external wind speed on two different days of measurement (day 12 and day 16). The daily pattern of wind speed is very similar on both days, however, the daily pattern of [CO 2] is very different. On day 12 (figure 18A), [CO 2] reached an extreme high of 81,910 ppm, while on day 16 (figure 18B), despite the similar pattern of wind speed as on day 12, the maximum was only 65,000 ppm. Regardless of any change in external wind speed, on day 12 [CO 2] in the nest chamber dropped sharply around 0700 h; a drop that I suggest may have been brought about by the female leaving the nest chamber, while on day 16 [CO 2] remained high until around noon, and then declined slowly until 1600 h when it increased again reaching close to its maximum values, despite a very high afternoon wind speed of > 7 m/s. Comparing the two days of measurement in the impermeable burrow brings me to conclude that the resident animal, its locality, activity and its direct effect on burrow geometry ( e.g., by plugging entrances and dragging nesting material into the burrow) is a more important in the ventilation of the burrow system than ambient wind.

It is of note that despite the very high [CO 2] in the nest chamber the female and pups in the impermeable burrow lived "normally" for the entire three weeks of measurement and were all

in good health when trapped after the experiment. Such levels of inspired [CO 2] produce respiratory distress in humans and other terrestrial mammals (Davenport, 1974, Dejours, 1981, Schaefer, 1982, Halpern et al., 2004). The physiological relevance of my findings is examined in the second part of this thesis. 46

Part 2 introduction

PART 2: PHYSIOLOGICAL RESPONSES OF SUNDEVALL’S JIRDS TO HIGH

AMBIENT CO 2 CONCENTRATION

2.1 Introduction The concentrations of respiratory gases inside the burrows of mammals have received much attention in the literature; the main tenets being that the burrow's architecture, the low permeability of soils to gases, the presence of soil microorganisms and the respiration of the burrow's tenants bring about FCO 2 and FCO2 that are above and below atmospheric levels, respectively (Studier and Baca, 1968, Chapman and Bennett, 1975, Withers, 1978, Wilson and Kilgore, 1978, Maclean, 1981). In mammals, breathing air containing high FCO 2 results in hypercapnia, a condition where the partial pressure of CO 2 (PCO 2) in the blood is elevated, + leading to respiratory acidosis, namely, increased blood [H ] and ([HCO 3 ] (Douglas et al., 1979, Schaefer, 1982). There are many reports that burrowing rodents are more tolerant of hypercapnia than comparable surfacedwelling species ( e.g., Marder and BarIlan, 1975, Arieli et al., 1977, Arieli and Ar, 1979, Ar, 1987). Indeed, I found that Sundevall’s jirds are able to tolerate [CO 2] as high as 70,000 80,000 ppm in the air they breathe, for prolonged periods of time (Part 1 of this thesis). In this part of my research I examined several physiological traits that, theoretically, would enable jirds to cope with such high [CO 2]. As far as I know there are no reports in the literature of all these traits in a single species in the context of its own burrows. In what follows, gas concentrations are presented as fractional concentrations, which is the prevalent representation in the relevant physiological literature. For example:

70,000 ppm = 7% CO 2, which equals an FCO 2 of 0.07

A brief introduction to acidbase balance

(Summarized from Brandis, 2008, SiggaardAndersen, 2005, Boron and Boulpaep, 2008, Davenport, 1974 and Hill et al., 2012) Acids that are produced daily by metabolic processes in the body are either excreted or metabolized in order to maintain homeostasis. The various acids produced by the body are classified as respiratory (volatile) acids and metabolic (fixed) acids. The term ‘respiratory acid’ usually refers to CO 2, which in itself is not an acid in the BronstedLowry system (as it does not contain a H + cation, so it cannot be a proton donor), but can instead be thought of as representing a potential precursor from which an equivalent number of carbonic acid (H2CO 3)

Part 2 introduction

molecules can be created. Dissolved CO 2 combines with water to form H 2CO 3 and the latter + dissociates into bicarbonate (HCO 3 ) and hydrogen ion (H ). + CO 2 + H 2O ↔ H 2CO 3 ↔ H + HCO 3

CO 2, being a volatile acid, can be expelled via the lungs. All other acids produced by the body are fixed and are excreted by the kidney. CO 2 is an end product of metabolism; it diffuses, along a partial pressure gradient from the cells into the blood plasma. A negligible fraction remains dissolved; the remainder being either hydrated, in a fast reaction catalyzed by capillary membranebound carbonic anhydrase (CA) and is buffered by plasma proteins, or diffuses into the red blood cells (RBC). In the RBC a considerable fraction of CO 2 combines with Nterminal amino groups of hemoglobin (Hb) to form carbamates, and the rest + + undergoes CAcatalyzed hydration to form HCO 3 and H . The H ions are mostly buffered by the imidazole group of histidine. Some of the HCO 3 is carried in RBCs to the lungs where + the hydration reaction is reversed and HCO 3 recombines with H to form CO 2 that can diffuse into the alveoli along a partial pressure gradient. The remaining HCO 3 diffuses out of the RBC into the plasma; the rise in intracellular [HCO 3 ] results in a change in the Donnan ratio (which determines the distribution of ions between the RBC and the plasma) causing chloride ions (Cl ) to move from the plasma into the RBC while HCO 3 is transported in the other direction. This process is known as the chloride shift (Hamburger, 1918). Cl in the RBC binds reversibly to Hb mitigating the change in [H +] that occurs during gas transport (Prange et al., 2001, Westen and Prange, 2003). Hb has two distinct configurations, oxyhemoglobin and deoxyhemoglobin. In the pulmonary capillaries, Hb molecules are "relaxed" and bind to as many as four O 2 molecules, becoming oxyHb. In the tissue capillaries, Hb gives up O 2 becoming deoxyHb, a CO 2 carrier and a buffer. HbO2 binding affinity is inversely related both to acidity and to PCO 2. + At the tissues, an increase in blood PCO 2 and the increase in blood [H ] causes Hb to

undergo a configuration change to its taut form; the oxyHb giving up O 2 where it is needed. + In the lungs a decrease in PCO 2 and in [H ] causes Hb to relax and reconfigure into its + oxygenated form, and stimulating it to release more H and to bind more O 2. This phenomenon is known as the Bohr effect (Bohr et al., 1904). The effect facilitates O 2 loading at the lungs and its unloading in the tissues, particularly in tissues where it is most needed

(e.g., muscles); when a tissue’s metabolic rate increases, its CO 2 production increases,

causing Hb to release more O 2.

Part 2 introduction

The oxygen dissociation curve and the Bohr effect

The Oxygen Dissociation Curve (ODC) is an important tool for understanding how the blood carries and releases O 2. Hb does not give up all its bound O 2 in the tissues; on average it gives up about half. Thus the half saturation of Hb is a convenient, and the conventional,

way of describing the location of the ODC, and the corresponding O 2 pressure ( P50 ) is considered the average unloading pressure (SchmidtNielsen, 2002). The ODCs of mammalian blood are related to body size, so that the blood of smaller

mammals has a higher unloading pressure for O 2 (SchmidtNielsen, 2002). Small animals

have a higher mass specific rate of O 2 consumption ( O2) than large animals (Kleiber, 1932).

As a consequence it follows that their tissues must be supplied with O 2 at a higher rate. This can be only be achieved by a steeper diffusion gradient from capillary to cells, by (1) an increase in capillary density in the tissue (SchmidtNielsen and Pennycuick, 1961) and (2)

increased O 2 unloading pressure in the tissues ( i.e., a high P50 ). SchmidtNielsen and Larimer

(1958) showed that P50(7.4) (PO2 (mmHg) at 50% O 2 saturation of blood at pH 7.4 at 37 °C) is inversely related to the animal's body mass ( mb) in grams by the allometric equation 0.054 P50(7.4) = 53.34 mb . However, in the literature a wide range of P50 values has been reported, even among species that are phylogenetically closely related and have similar habits

and habitats. Lahiri (1975) found good correlation (r = 0.91) between predicted P50(7.4) and mb only for mammals of under 200 g. Forman (1954) reported P50 values of 12 rodent species between 2.93 kPa and 4.89 kPa (22 mmHg and 36.7 mmHg), for rodents ranging in mb from 10 540 g (appendix, table 16).

Body mass (mb) is only one of the many factors that affect O 2 transport ( e.g., Forman, 1954, Hall, 1966, Lahiri 1975, SchmidtNielsen, 2002). The Hb molecule of various animals is species specific and there are examples of its functional adaptation to different

environmental needs or living conditions. In environments where ambient PO2 is low and animals are exposed to hypoxia, such as at high altitudes and in closed burrow systems, or in

diving animals and all mammalian embryos, the P50 is relatively low (Hall et al., 1936, Banchero et al., 1971, Lechner 1976, Clausen and Ersland, 1968, Clausen and Ersland, 1971, Hall, 1966, Ar et al., 1977, Boggs et al., 1984, Wei et al., 2006, van Aardt et al., 2007). The increased Hb affinity for O 2 as reflected by low P50 permits the Hb to become fully saturated with O 2 at a partial pressure at which the blood of other mammals is only partially saturated.

An increased affinity of O 2Hb has also been shown to be associated with increased Bohr

Part 2 introduction

effect, which facilitates better unloading of CO 2 at the tissues (Riggs, 1960, Messier and Schaefer, 1973, Ar et al., 1977).

Acidbase balance when ambient [CO 2] is high

The extent of increase in [H +] in the blood and other body fluids when an animal is exposed to high FCO 2 in the air it inspires is determined by the effectiveness of three compensatory/regulatory mechanisms, (1) the capacity of the blood and tissue buffer systems,

(2) the magnitude of the ventilatory response that modulates arterial partial pressure of CO 2 (PaCO 2), and (3) the efficiency of renal compensation by alteration in HCO 3 excretion

(Darden, 1972, Lai et al., 1973, Chapman and Bennett, 1975). When PaCO 2 rises all these mechanisms are set in motion (Lai et al., 1973); they proceed simultaneously, but on different time scales that range from several minutes to several days.

Blood buffering capacity

High blood buffering capacity protects the individual from the effect of CO 2 retention and minimizes the increase in blood [H +]. Increased buffer capacity is achieved through the bicarbonatecarbonic acid system and through noncarbonic blood buffers, including Hb, plasma proteins, phosphates and nonprotein thiol groups (Boron and Boulpaep, 2008). Some fossorial mammals are known to have increased blood buffer capacity than their terrestrial relatives ( e.g., Chapman and Bennett, 1975, Lechner, 1976, BarIlan et al., 1985). In the extracellular fluid (ECF), the bicarbonate system is quantitatively the most important mechanism for buffering metabolic acids. Its effectiveness is greatly increased by ventilatory

changes that take place to maintain constant PCO 2, and by renal mechanisms that result in changes in plasma [HCO 3 ]. In other words, higher [HCO 3 ] drives the CO 2 ↔ HCO 3

equilibrium towards CO 2, which can be excreted by the lungs, thus preventing an increase in [H +]. For example, the bicarbonatecarbonic acid system, as assayed by the bicarbonate content of the blood, appears better developed in fossorial valley pocket gophers, Thomomys bottae (Chapman and Bennett, 1975), and in Mediterranean pine voles, Microtus duodecimcostatus (Mathias and Freitas, 1989) than in laboratory rats. Chapman and Bennett (1975) found that the average [HCO 3 ] in the blood of pocket gophers is significantly higher than in rats, (28.1 mmol/kg and 19.8 mmol/kg, respectively). Similarly, average [HCO 3 ] in the Mediterranean pine vole was 32.98 ± 5.32 mmol/kg. The noncarbonic buffering power of several fossorial and semifossorial rodents is also considerably higher than that of nonfossorial species (appendix, table 16). Noncarbonic

Part 2 introduction

- ∆HCO 3 buffering power can be defined as the change in [HCO 3 ] for a unit change in p H,. ∆pH The amount of HCO 3 formed or consumed during respiratory acidbase disorders increases with the noncarbonic buffering power, β (Boron and Boulpaep, 2008). For example, the β value of the semi fossorial fat sand rat, Psammomys obesus , is 69.2 ± 13.7 (BarIlan et al., 1985), much higher than in man (25, Boron and Boulpaep, 2008), or laboratory rat (27.3 ± 6.4, BarIlan et al., 1985). Another manner of presenting noncarbonic buffering capacity is as the change in log

∆logCO PCO 2 with a unit change in pH, i.e., (Chapman and Bennett, 1975, Lechner, 1976, ∆pH Snyder, 1976, Ar et al., 1977, BarIlan et al., 1985). The noncarbonic blood buffer capacity

of most mammal species is around 1.5 log PCO 2/pH (mmHg/pH unit) (appendix, table 16). Fossorial and semifossorial species were shown to have considerably higher buffer capacity; for example, pocket gopher (2.67, Chapman and Bennett, 1975), valley pocket gopher, and mountain pocket gopher, T. umbrinus melanotis (2.2 ± 0.41 and 2.83 ± 0.43 respectively, Lechner, 1976) and the fat sand rat (2.47 ± 0.21, BarIlan et al., 1985).

Ventilatory Regulation of AcidBase Balance

Ventilation also plays a major role in maintaining acidbase balance. Under normoxic

conditions the elimination of CO 2 from the body is primarily through respiration and depends on the gradient of CO 2 between venous blood and alveolar air. If blood PCO 2 rises, more CO 2 + must be excreted by the lungs. A rise in blood PCO 2 results in an increased in blood [H ]; this stimulates the respiratory center in the medulla to increase ventilation which facilitates + CO 2 elimination by the lungs. Therefore, ventilatory regulation refers to changes in [H ] due

to changes in PCO 2 from alterations in ventilation. This change in ventilation can occur rapidly (within minutes), with significant effect on blood [H +].

An increase in FCO 2 in the environment elevates the PCO 2 of both the alveolar air and

arterial blood, thereby decreasing the gradient for CO 2 diffusion (Withers, 1975). When

inspired FCO 2 in the environment is constantly high, as in the case of burrow dwellers,

hyperventilation may be of little use in eliminating CO 2 (Darden, 1972, Soholt et al., 1973,

Boggs et al., 1984). A lower ventilatory response to CO 2 was observed in several fossorial and semifossorial mammals, compared to nonfossorial species. For example, a reduction in the ventilatory response was observed in the completely fossorial valley pocket gopher (Darden, 1972) and Middle East blind mole rat (Arieli and Ar, 1979), and in the semi fossorial Merriam's kangaroo rat (Soholt et al., 1973). 51

Part 2 introduction

Renal Compensation of AcidBase Balance

In acidbase balance, the kidneys are responsible for two major functions: (1) the reabsorption of filtered HCO 3 and (2) excretion of fixed acids. An increase in Pa CO 2 results in increased renal H + secretion and increased bicarbonate reabsorption. In humans and other mammals the renal response kicks in only in cases of continuous acidosis of 6 to 12 hours with a maximal effect reached by 3 to 4 days, when acidosis is chronic (Brandis, 2008). However, Lai et al., (1973) reported that in rats 45% of the increase in plasma HCO 3 occurring during the first hour of exposure to FCO 2 = 0.1 can be attributed to renal reabsorption. BarIlan and Marder (1985) suggested that in rats, mechanisms involving renal carbonic anhydrase are responsible for a significant and rapid change in whole body buffering (a rise in HCO 3 ) that takes place during the initial phase of acute exposure to CO 2 and suggested that this may represent a mechanism of adaptation to burrow high FCO 2 conditions. Renal compensation of acidbase balance was not investigated in the present research.

Research goals

Based on what I learned in the first part of my research, namely that Sundevall's jirds are

able to tolerate FCO 2 as high as 78% in the air they breathe for extended periods of time, and that they can successfully raise a litter under these conditions, my goals were to (1) test

whether Sundevall's jirds voluntarily tolerate high FCO 2 in the air they breathe, and (2) examine the importance of two of the compensatory mechanisms described above in their

ability to cope with high FCO 2. In this light, I hypothesized that Sundevall's jirds voluntarily tolerate high FCO 2 in the air they breathe and tested the prediction that when given a choice between two chambers, one containing high (0.07) and the other atmospheric (0.0004) FCO 2, jirds spend an equal amount of time in both.

Assuming that Sundevall’s jirds are indeed tolerant of high inspired FCO 2 and the resulting hypercapnia, I examined several physiological traits that facilitate coping with such conditions. (1) I examined the blood O 2transport properties of Sundevall's jirds in the

context of its burrow's ostensibly high CO 2 and hypoxic environment. I tested the predictions

that (a) the P50 of Sundevall's jird blood is relatively low and therefore promotes increased

extraction of O 2 from burrow air (b) that the Bohr effect, , is relatively high, and ∆pH facilitates additional O 2 unloading from Hb to tissues. (2) I tested the prediction that Sundevall's jirds have a higher blood buffering capacity than reported for nonburrowing

Part 2 introduction

+ mammals of the same size. I also compared the buffering capacity, [H ], [HCO 3 ] and PCO 2 of the blood of jirds with those of the semifossorial fat sand rat, another Negev Desert rodent of similar body size that may also be exposed to high FCO 2 in their natural environment. (3) I tested the predictions that the ventilatory response of Sundevall's jirds inspiring air with high

FCO 2 is lower than in other mammal species, and that they pay a low or no energetic cost for this ability. I compared the increase in respiration frequency ( fr) of jirds when they breathed air with FCO 2 = 0.07 with that of other mammals. I also compared the activity, resting metabolic rate (RMR) and mean metabolic rate ( MR) of jirds breathing air with FCO 2 = 0.07 to jirds breathing CO 2free air. Also, I tested the prediction that the RMR of Sundevall's jirds does not change when ambient FCO 2 increases sequentially or acutely, up to 0.07.

Methods

2.2 Methods

Animal maintenance

I used Sundevall's jirds from a breeding colony maintained by Prof. Boris Krasnov and Dr. Irina Khokhlova on the Sede Boqer Campus of BGU. The colony was founded in 1997 with progenitors captured in the Negev Desert (Details on maintenance and husbandry of this colony can be found in Khokhlova et al., 2012, Khokhlova et al., 2009). During the course of my experiments, Sundevall’s jirds were housed in standard rat cages (35 × 25 × 15 cm) with wood shavings as bedding, paper towels as nesting material, a pinecone for chewing, and access to millet seeds ad libitum . Also, as a water source, the jirds were fed fresh alfalfa daily. The cages were kept in animal rooms at an air temperature of ~25 °C. Animals used in experiments 1 3 and 5 (below) were kept on a natural light cycle because the animal room had a large window. The animals used in experiment 4 (section 2.2.4) were kept in a windowless animal room on a 12 h:12 h lightdark cycle (lights on at 0800 h). Only female jirds (body mass, mb = 127.92 ± 20.38 g; mean ± SD) were used in the experiments, because I was interested in the physiology of the females, which probably encounter the most extreme

FCO 2 during lactation, when pups are also present in the nest chamber. Care of the animals and the experimental protocols were approved by Ben Gurion University of the Negev (BGU) institutional animal care committee (BGUS1032010) and conformed to all state guidelines and regulations (Israel Ministry of Health IL25062010, Israel Nature and National Parks Protection Authority 2011 / 38248).

The fat sand rat

The fat sand rat, Psammomys obesus Cretzschmar 1828, is a large, diurnal gerbilline rodent, systematically close to Sundevall's jirds (Mendelssohn and YomTov, 1987). Fat sand rats are found in North Africa, ranging from Mauritania to Egypt and Sudan and east across the Arabian Peninsula (Mendelssohn and YomTov, 1999). In Israel they occur in river beds and salinemarsh areas in the Negev Desert, around and south of Beer Sheva, in the Judean Desert and the Dead Sea area, as far north as Jericho (Mendelssohn and YomTov, 1999). The species feeds almost entirely on salty, succulent perennials of the family Chenopodiaceae (Shenbrot, 2004). They live in extensive burrow systems, with 3 5 oval openings, with a vertical diameter of 7 10 cm and a horizontal diameter of 9 13 cm (Shenbrot et al., 2002). The burrows are located near and under the bushes on which they forage (Mendelssohn and YomTov, 1999, Nowak, 1999). The burrow system is a complex of tunnels and separate

Methods

chambers for nesting, sleeping, food storage and toilets. The species is colonial, but each burrow system holds only one adult, except when breeding, when a pair or a family occupies one system (Mendelssohn and YomTov, 1999, Nowak, 1999). I used fat sand rats from the breeding colony in the exhibition hall of HaiBar Nature Reserve near Yotvata in the Arava valley. The progenitors of the colony were captured in the Negev Desert. During the course of the experiment fat sand rats were housed in standard rat cages (35 × 25 × 15 cm) with wood shavings as bedding, paper towels as nesting material, and were provided with a diet of salt bush ( Atriplex halimus and Anabasis articulata ), ad libitum , that was refreshed daily. The cages were kept in animal rooms where air temperature was ~25 °C. Animals were kept on a natural light cycle because the animal room had a large window.

Preparation of gas mixtures

For the preparation of CO 2free air, outside air was pumped through a purge gas generator

(Puregas, Broomfield, CO, USA, model #PCDA112m32C) that removes CO 2 and reduces water vapor to less than 1 ppm. For the preparation of gas mixture containing

fractional concentrations of N 2 and O 2 close to atmospheric and FCO 2 > 0; I mixed dry CO 2

free air and pure bottled CO 2 (Maxima, Ramat Hovav Industrial Zone, Israel) with a twoway gas mixing pump (Digamix 5KA27 Wösthoff, Bochum ,Germany)

For the preparation of other gas mixtures, I mixed desired proportions of bottled CO 2, O 2

and N 2 (Maxima, Ramat Hovav Industrial Zone, Israel) with a three way gas mixing pump (G27, Wösthoff, Bochum, Germany), or used a preordered gas mixture (Maxima, Ramat Hovav Industrial Zone, Israel)

2.2.1 Experiment 1. CO 2 tolerance

To test whether Sundevall's jirds avoided air with high FCO 2 I designed a preference test

which allowed jirds to choose between a chamber containing high CO 2 (FCO 2 = 0.07) and a chamber containing atmospheric air ( FCO 2 = 0.0004 CO 2). I used two identical plastic containers (16 × 16 × 18.5 cm) attached by a transparent plastic tube (21 cm long × 5 cm ID) with a curtain of overlapping plastic strips hung at its midpoint. The curtain impeded the mixing of air between the chambers, but did not hinder the movement of the jirds. I pumped

air with a FCO 2 of 0.07 provided by a 2way gas mixing pump at a rate of 1350ml/min into

one chamber and CO 2free air at ~4000 ml/min into the other. These flow rates were determined by trial and error. At these flow rates the chambers equilibrated to the desired gas

Methods

concentrations within 10 minutes. [CO 2] in each chamber was measured with and infrared gas analyzer (Carbocap® GMT222, Vaisala, Finland. see pg. 13 for instrument specifications) before and after the jird was placed in the system. In the tube connecting the chambers, there were two motion sensors each consisting of a red LED on one side of the passage and photo resistor opposite it. When a jird traversed the tunnel, it blocked the light from the photo resistor and a data logger recorded the individual’s movement and direction. During experiments, the choice chamber was placed in a dark, quiet cabinet. I first introduced the jirds to the choice chamber without the curtain for 10 min or until the individual moved from one chamber to the other, which ever came last. I then inserted the curtain and allowed the animal to explore both chambers until they had crossed between them five times, or for 20 minutes, which ever came last. Thus, the jirds learned that the curtain did not restrict their movement. During this introduction both chambers contained atmospheric air. Nine female Sundevall's jirds were used to determine preference. Each individual was placed in the choice chamber for 4 h during its active phase (after dark outside). To minimize the effects of sidebias, I varied the starting location of the animal and also which chamber contained highCO 2 air. Thus, five individuals were first placed in the righthand chamber and four in the lefthand chamber. Five individuals were first exposed to the side with atmospheric air and four in the chamber with highCO 2 air. For five individuals, the chamber with atmospheric air was on the righthand side and for four individuals it was on the left. I calculated how many minutes each jird spent in each chamber for 3.5 h, beginning 30 min after it was initially placed in one of the chambers, giving the jirds time to explore. A one sample ttest (SPSS, Version 20.0., IBM Corp. Armonk, NY 10504) was used to determine if time spent (in min) in the chamber with atmospheric air differed from the 50% expected by chance.

2.2.2 Experiment 2. Oxygen dissociation curve and Bohr effect

I built the ODCs and calculated the P50 s and Bohr effect values for 11 female Sundevall's jirds using the mixing method (Haab et al., 1960, Snyder, 1976, Scheid and Meyer, 1978). One milliliter of blood was drawn from the suborbital sinus of each with a preheparinized capillary tube (125 l, Clinitubes capillaries D957G70, Radiometer Medical ApS) and collected into a Liheparinized test tube. Blood samples were placed in icewater and analyzed within 90 min of collection.

Methods

The blood sample of each animal was equally divided in two aliquots that were concurrently equilibrated for 15 minutes in two tonometers (type 237, Instrumentation Laboratory, Italy) with humidified gases. In one tonometer I generated oxygensaturated blood and in the other deoxygenated blood. To produce oxygensaturated blood I used a gas

mixture of 0.25 O 2, 0.05 CO 2 and the balance N 2, provided by a 3way gas mixing pump. To produce deoxygenated blood I used a mixture of 0.05 CO 2 and the balance N 2 provided by a 2way gas mixing pump. Flow rate of both gasmixtures was 300 ml/min and equilibration was done at mean jird body temperature, 37.3°C (BenPorat et al., 1976).

To obtain blood with O 2Hb saturations of 25%, 50% or 75% I drew blood from each tonometer in the required proportion into a 100 l syringe (model 710, Hamilton, Reno, Nevada). In order to avoid contamination of the sample with air, I filled the dead space of the syringes with mercury and drew from the saturated blood first and then from the deoxygenated blood. The small ball of mercury that was formed was then used to mix thoroughly the deoxygenated and saturated blood samples by agitating the syringe for 60 seconds. The mixed sample was injected into a blood gas analyzer (Radiometer, ABL80flex,

Copenhagen, Denmark) for measurement of pH, and PO2. I used the nonlinear regression curve fitting function in SigmaPlot ® 11 to fit the experimental data to a Hill plot (Hill, 1910).

Although all blood samples were exposed to a PCO 2 of 5.33kPa (40 mmHg), I found a large variation in blood pH. I used the variation in pH to calculate the Bohr effect, namely the

∆logP50 change in log PO2 (mmHg) with a unit change in pH: . Where pH v is ∆pH

the pH of venous blood (7.25) and pH a is the pH of arterial blood (7.37), P50 v is the PO2 of venous blood at 50% Hb saturation and P50 a is the PO2 of arterial blood at 50% Hb saturation.

P50 alone was also measured for 9 fat sand rats (5 females and 4 males), using the same technique. The P50 values that I measured for jirds and fat sand rats were compared with 0.054 values of predicted P50 calculated from an allometric equation P50(7.4) = 53.34 mb

(SchmidtNielsen and Larimer, 1958).

Methods

2.2.3 Experiment 3. Blood buffer capacity One milliliter of blood was drawn from the suborbital sinus of 12 female Sundevall's jirds and 12 fat sand rats (8 females and 6 males) , by the abovedescribed method. Blood was equilibrated in tonometers to each of four humidified gas mixtures, simulating either venous

or arterial blood under normal FCO 2 (0.05 or 5.33kPa) or under high FCO 2 (0.1 or 10.66 kPa) (table 5). The equilibration with the four gas mixtures was always done in the following sequence: first, an aliquot of blood was equilibrated with a mixture simulating mixed venous blood (mixture A in table 5) and then with a mixture simulating mixed venous blood with

high FCO 2 (mixture B). Then, a second aliquot of blood (kept in icewater until used) was equilibrated, first with a mixture simulating arterial blood (mixture C) and then with a mixture simulating arterial blood with high FCO 2 (Mixture D). Gas mixtures A, B and C were produced with a 3way gas mixing pump and gas mixture D was a preordered mixture (Maxima, Ramat Hovav Industrial Zone, Israel). Equilibration was done at a temperature of 37.3 °C and a gas flow rate of 300 ml/min.

Table 5. The composition of gas mixtures with which blood of Sundevall's jirds, Meriones crassus , and fat sand rats, Psammomys obesus , was equilibrated in order to simulate venous

or arterial blood, under normal FCO 2 (respectively gas mixtures A and C) and high FCO 2 (respectively gas mixtures B and D).

Gas Composition of gas Condition mixture mixtures [%] O2 CO 2 N2

Venous condition at normal FCO 2 A 5 5 90

Venous condition at high FCO 2 B 5 10 85

Arterial condition at normal FCO 2 C 15 5 80

Arterial condition at high FCO 2 D 15 10 75

After 15 minutes of equilibration, equilibrated blood samples were injected into the blood + + ++ gas analyzer for measurement of pH, PO2, PCO 2, % Hct, [HCO 3 ], [Na ], [K ], [Ca ] and [Cl ]. The average barometric pressure on the days of the measurement was 95.79 ± 0.35 kPa

Methods

(718.49 ± 2.7 mmHg) (Jacob Blaustein Institutes for Desert Research Desert Meteorology Weather Station, http://bidr.bgu.ac.il/BIDR/research/phys/meteorology/default.asp).

Noncarbonic blood buffer values were calculated as the change in log PCO 2 (mmHg)

∆logCO with a unit change in pH, , and as the change in [HCO 3 ] (mmol/l) with a unit change ∆pH

- ∆HCO in pH, 3 , for venous and arterial blood. ∆pH

2.2.4 Experiment 4. Behavioral response and physiological costs in jirds inspiring air

with high FCO 2

To assess the behavioral and physiological costs to jirds acutely exposed to high FCO 2, I measured their metabolic rates by indirect calorimetry. Resting MR (RMR) was used to quantify basic energy requirements. Because RMR does not include activity, I also calculated mean MR ( MR), to estimate the average metabolic costs of both physiological and behavioral responses to inspiration of air with high FCO 2. While the jirds were in the chambers, I recorded each individual’s behavior with a video camera and measured respiration frequency

(fr) (breathes/min) and time spent sleeping, grooming, and being active but not grooming. I used eight Sundevall’s jirds in a repeatedmeasures design. Each jird was measured in

CO 2free air and in high FCO 2 (FCO 2 = 0.07). Four jirds were exposed to high CO 2

conditions first, and four were exposed to the CO 2free air first. Each measurement lasted 2 h, and 24 h passed between measurements under different conditions, namely, if an individual was exposed to high CO 2 conditions on day 1, it was exposed to CO 2free air on day 3. Before and after each measurement of MR, I weighed the jirds. I measured MR for 2 h while the jird’s behavior was video recorded for later analysis of behavior and fr. I quantified behavior of each jird for 1 h starting 30 min into the video. Behavior was coded as inactive (sleeping, resting, sitting still), grooming, or other activity. I recorded grooming behavior because excessive grooming is a classic stress response in rodents (Spruijt et al., 1992, van Erp et al., 1994). I quantified fr by counting the number of breaths a jird took during 30 seconds while resting, and multiplying this number by 2 to obtain breaths per minute. fr was counted three times in each CO 2 treatment, and then averaged within CO 2 treatment.

O2 consumption ( O2) and CO 2 production ( CO 2) were measured in an 8channel open flow respirometry system, which allows measurement of up to seven animals at a time, allocating 1 channel for a baseline measurement (see Marom et al. 2006 for a full description of the system). Measurements were made between 0930 h and 1800 h during the animal’s

Methods

inactive phase; I measured one individual at a time and four in a day. Animals were placed in individual, 1.9 L plastic chambers (Lock and Lock, Hana Cobi, Korea). I reduced the volume of the chambers to 1.612 L by stacking eight 25 ml sealed, glass collection jars in each. The chambers were placed in a temperaturecontrolled cabinet (Refritherm5, Struers, Ballerup, Denmark) set at 30 °C, which is in the jirds' thermoneutral zone (BenPorat et al., 1976). Each chamber was fitted with a 7 mm hardware cloth floor above a pan of paraffin oil to trap excreta. Following Bakken and Rowe (2011), no special precautions were taken to make wall and air temperature equal.

For measurements at FCO 2 = 0 incurrent air was dry and CO 2free. For measurements at

FCO 2 = 0.07 I mixed dry CO2free air and pure CO 2 produced using a twoway gas mixing pump. In both cases, air was then passed through a manifold of valves and then into each chamber at an average flow rate of 400 ml/min. Excurrent air passed through an 8channel gas multiplexer (RM Gas Flow Multiplexer, Sable Systems, Las Vegas, NV, USA) that was programmed to sequentially select channels for sampling. FO2 and FCO 2, in the excurrent gas stream were measured by the O 2 and CO 2 analyzer (FoxBoxC field gas Analysis System, Sable systems International, Las Vegas, NV, ISA). Data were collected with data acquisition software (Trace DAQ pro 2.1.2.0, Measurement Computing Corporation, MA, USA).

FO2 and FCO 2 were measured for six cycles for each jird. Each cycle started with a 5 min measurement of the baseline channel followed by a 15 min measurement of air from the chamber of an individual. FO2 and FCO 2 data were recorded every 30 seconds. Incurrent FO2

and FCO 2 were read from the reference channel. O2 and CO 2 were calculated by equations 11.7 and 11.8 from Lighton (2008) 2. RMR was defined as the lowest consecutive 3 min of O2. MR was calculated from the average O2 over all the cycles less the first cycle. O2 was converted to MR using 20.08 J/ml O 2 (SchmidtNielsen 1997).

Statistical analysis: Linear mixed effect models were used to test for differences among

individuals breathing CO 2free air or air with high FCO 2 and tested for differences in RMR,

MR, fr and in proportion of time spent sleeping, grooming, and being active but not grooming

2 O2O2O2CO 2CO 2 CO2CO2CO2O2O2 O2 , CO 2

is the average flow rate of the gas passing through the metabolic chambers. O2 and CO2 are the ′ ′ fractional concentrations of O2 and CO 2, respectively, in incurrent air. O2 and CO2 are the fractional concentrations of O2 and CO 2, respectively, in excurrent air.

Methods

(i.e., , other behavior). The fixed effect was treatment (CO 2free or high FCO 2). Mean mb,

calculated as the sum of mb at the beginning and end of the MR measurement divided by 2, was initially considered as a covariate in all models, but preliminary analyses indicated that it was not significant, so it was not included in any further analyses. Because I used a repeated measures design, I included "individual" as a random effect. The proportions of time spent sleeping, grooming, or being active, but not grooming, were arcsine transformed to normalize data distribution. I could not count fr for one jird breathing high FCO 2 because it never stopped moving. All analyses were done in R (R development core team, 2011), with the packages nlme (Pinheiro et al., 2011).

2.2.5 Experiment 5. The effect of sequential and acute changes in inspired FCO 2 on the energy expenditure of jirds.

To measure the effects of high inspired FCO 2 on the jird's MR I used indirect calorimetry as described above to measure O2 and CO 2 of 9 female Sundevall's jirds ( mb = 127.92 ± 20.38 g). Measurements were made under three conditions: (1) during sequential step

changes in inspired FCO 2; (2) during acute changes in inspired FCO 2; and (3) with no added

CO 2 (FCO 2 = 0). Postabsorptive female jirds, that have been deprived of food for 5 h were placed in individual 0.85 L plastic chambers (Lock&Lock, Hana Cobi, Korea) outfitted with a 7 mm hardware cloth floor above paraffin oil to trap excreta. Following Bakken and Rowe (2011), no special precautions were taken to make wall and air temperature equal. Three chambers at a time were placed together in a temperature controlled cabinet (Thermo Scientific, model Precision 815, Cleveland, OH) and were maintained at 30 °C, within the jird's thermoneutral zone (BenPorat et al., 1976). Gas mixture was pumped into the chambers at an average rate of 360 ml/min using a gas controller and monitor (G245, Qubit Systems Inc., Kingston ON, Canada). Excurrent air from the chambers passed through a gas multiplexer (G244, Qubit Systems Inc.) that was programmed to sequentially select

channels for sampling. One channel was allocated to baseline measurement. FO2 and FCO 2, in the excurrent gas stream were monitored with an O 2 analyzer (S108, Qubit Systems Inc .) and a CO 2 analyzer (S158, Qubit Systems Inc.), respectively. All data were collected with data acquisition software (C950 MC Gas Exchange Software, Qubit Systems Inc.).

Measurements of sequential step changes in FCO 2 began at FCO 2 = 0 (dry, CO 2 free, room air) and gradually increased to 0.01, 0.02, 0.04, 0.07 and then returned to 0.

Measurements of acute changes in inspired FCO 2 began at 0, increased to 0.07, and returned

to 0. Animals were exposed to each FCO 2 for a 135 min block of time. The measurements

Methods

made with no added CO 2 were used as individual controls to separate the effects of time of day and hours spent in the metabolic chamber from the effect of FCO 2. For each of the abovedescribed treatments, the same 9 animals were measured in groups of three, on three consecutive days. Animals were measured in the same order, so that, for example, the same three animals that were measured on day 1 were measured again on days 5 and 8. On days 1 3 animals were measured during sequential step changes in FCO 2 for 13 and a half hours, starting at 0530 h, on days 5 7 animals were measured during acute changes in FCO 2 for 6 hours and 40 minutes, starting at 0745 h and on days 8 10 animals were measured with no added CO 2, for 13 and a half hours starting at 0530 h.

For measurements at FCO 2 = 0 incurrent air was dry, CO 2free. For measurements at

FCO 2 > 0, incurrent air was a gas mixture of dry CO 2free air and pure CO 2 produced by a

twoway gas mixing pump. FO2 in the chamber ranged from 0.21 (at FCO 2 = 0) to 0.1953 (at

FCO 2 = 0.07). At these inspired FO2, according to my own dissociation curve for Sandevall's jirds (see below), the Hb of jirds is fully saturated and therefore they did not experience hypoxia in the experiments. In each 135 min time block, individuals were sampled three times for 10 min. Individuals were sampled sequentially, and each sampling period was separated by a 5 min baseline

measurement in order to correct for drift and quantify FO2 and FCO 2 in incurrent air. Mean values of excurrent FO2 and FCO 2 were calculated for each animal in each time block using the lowest 100 second sequence of oxygen consumption during the 2 nd or 3 rd sampling period.

O2 and CO 2were calculated using equations 11.7 and 11.8 of Lighton (2008) and

converted to MR using 20.08 J/ml O 2 to convert O2 to power in mW (SchmidtNielsen 1997).

MR treatment is defined as a jird's MR during a step change in inspired FCO 2; MR control is defined as MR during control measurements. MR diff is a jird's MR treatment minus its MR control in the corresponding time block.

Statistical analysis: I wished to determine whether RMR of jirds is affected by FCO 2 in inspired air when FCO 2 changes sequentially or acutely from 0 to 0.07 and back to 0. To tease apart the effects of time of day and time spent inside the metabolic chamber from the

effect of FCO 2, I used MR diff as the response variable. I used two linear mixed effect models (one for sequential step changes and one for acute changes) to test the effect of the change in

FCO 2 on MR diff . My categorical fixed factor was FCO 2, with six levels in the first model (0, 0.01, 0.02 0.04, 0.07, 0) and three levels in the second (0, 0.07, 0). The reference level in the

Methods

model was the first block of time during the first two hours of measurement when FCO 2 = 0.

An effect of the change in FCO 2 on jirds MR would be indicated by a significant difference between MR diff of different treatments but not at FCO 2 = 0. To control for individual variation among jirds I included individual as a random factor (Zuur et al., 2009). Body mass may affect MR (Kleiber, 1932); I therefore included it as a

covariate in the models. However, models that included mb were less likely to fit the data as

they had higher Akaike information criteria (AIC) values than models that did not include mb.

Hereafter, I only refer to the models that did not include mb. All analyses were done in R (R development core team, 2011), with the packages nlme (Pinheiro et al., 2011). I used Tukey multiple comparisons adjusted for mixed effects models to test for differences between experimental groups within levels with the glht function in the multcomp R package (Hothorn et al., 2008).

Results

2.3 Results

2.3.1 Experiment 1. CO 2 tolerance

On average, Sundevall’s jirds spent 105.7 ± 86.8 min in the CO 2free chamber and 144.0 ±

72.10 min in the high CO 2 chamber. The time spent in the low CO 2 chamber was not different from that expected by chance (t 8 = 0.02, p = 0.98).

2.3.2 Experiment 2. Oxygen dissociation curve and Bohr Effect The ODC of Sundevall's jirds has the typical mammalian sshape which was fitted to the

.. experimental data according to the Hill equation ( , F2,4 = 1223.1, p = 0.0001) .. (figure 1). Under standard conditions ( i.e., pH =7.37 ± 0.02, [H +] = 42.66 ± 2.62 mmol;

T = 37 °C; PCO 2 = 5.33 kPa (40 mmHg)) P50 = 5.66 ± 0.33 kPa (42.5 ± 2.52 mmHg) (table + + 6). As expected, as [H ] increased blood PCO 2 increased; at pH = 7.25 ± 0.02 ([H ] = 55.32 ± 3.0 mmol), pH = 7.15 ± 0.02 ([H +] = 70.79 ± 3.42 mmol) and pH = 7.02 ± 0.01 ([H +] =

95.50 ± 3.03 mmol) the values of P50 were 6.30 kPa (47.25 ± 2.87 mmHg), 6.73 kPa (50 .5 0 mmHg) and 9.60 kPa (72 mmHg) respectively (figure 20). The Bohr effect value of

Sundevall's jirds is 0.42 (table 6). P50 of nine fat sand rats was calculated with blood pH = 7.17 ± 0.03 ([H +] = 67.78 ± 5.21) (table 6).

* Table 6. Body mass ( mb) and blood gas properties ( P50 , predicted P50(7.4) and Bohr effect) of 11 Sundevall's jirds and 9 fat sand rats.

mb (g) P50 mmHg Predicted Bohr Species Average ± SD Average ± SD P50 (7.4) * effect

P50(7.37) = 42.50 ± 2.52 Sundevall's jird P50(7.25) = 47.25 ± 2.87 Meriones crassus 127.92 ± 20.38 41.04 0.42 N = 11 P50(7.17) = 50.50 P50(7.02) = 72

Fat sand rat

Psammomys obesus 124.53 ± 44.7 P50(7.17) = 47.88 ± 4.72 41.10 N = 9

0.054 * Predicted P50 (7.4) = 53.34 mb , the PO2 (mmHg) at 50% O2 saturation of blood at pH 7.4 at 37 °C (SchmidtNielsen and Larimer, 1958).

Results

Figure 20. The oxyhemoglobin dissociation curve of Sundevall's jirds (full circles, black line)

under "standard" conditions (pH = 7.37, T = 37 °C, PCO 2 = 5.33 kPa = 40 mmHg), and P25 , P50 and P75 of jirds at different blood pH ( T = 37°C, PCO 2 = 5.33 kPa (40 mmHg)), pH = 7.26 (empty circles), pH = 7.15 (full triangles apex down), pH = 7.02 (empty triangles

apex up, P50 and P75 only). Curve was fitted to the experimental data according to the Hill .. equation , F 2,4 = 1223.1, p = 0.0001. ..

2.3.3 Experiment 3. Blood buffer capacity Blood buffer capacities (table 7), pH , PCO 2 and HCO 3 of the blood of Sundevall's jirds and fat sand rats equilibrated by tonometry with gas mixtures (table 5) that simulate venous or arterial blood, under normal FCO 2 and high FCO 2.

Results

Table 7. Blood buffer capacities, pH, PCO 2 and [HCO 3 ] of venous and arterial blood in Sundevall's jirds and fat sand rats which were equilibrated in tonometry in order to simulate venous or arterial blood, under normal FCO 2 and high FCO 2. Values are means ± SD.

+ ∆HCO ∆CO [H] PCO 2 [HCO 3 ] Species N pH ∆ ∆ nmol/l mmHg mmol/l Sundevall's jird Venous 25.02 ± 3.18 1.88 ± 0.64 7.34 ± 0.10 46.64 ± 10.52 32.92 ± 1.83 16.15 ± 3.39 12 Meriones crassus Arterial 25.83 ± 3.75 2.04 ± 0.33 7.29 ± 0.11 52.32 ± 13.00 33.83 ± 1.64 17.49 ± 3.33

Fat sand rat Venous 24.67 ± 4.14 2.35 ± 0.18 7.21 ± 0.06 62.52 ± 8.35 35.83 ± 1.11 13.85 ± 1.68 12 Psammomys obesus Arterial 27.46 ± 2.52 2.37 ± 0.21 7.18 ± 0.07 67.51 ± 9.67 36.42 ± 2.39 13.03 ± 1.55

Results

2.3.4 Experiment 4. Behavioral response and physiological costs in jirds inspiring high

FCO 2 The RMR of Sundevall's jirds at an ambient temperature of 30°C, when breathing dry

CO 2free air was 395.46 ± 53.55 mW (mean ± SD). Respiration rate (r) was 59% higher in jirds inspiring air with high FCO 2 than when inspiring CO 2free air (F 1,6 = 95.04, p < 0.0001, figure 21A, table 8). Sundevall's jirds had an 11.6% higher RMR (F 1,7 = 7.5, p = 0.03, Figure

21B, table 8) and a 21% higher MR (F 1,7 = 9.42, p = 0.02, Figure 21C, table 8) when inspiring air with a high FCO 2 than when inspiring CO 2free air. Jirds did not differ in the proportion of time they spent grooming (F 1,7 = 0.98, p = 0.35, table 8), but slept 35.3% less

(F 1,7 = 22.3, p = 0.002, table 8), and were 174.5% more active overall (F 1,7 = 8.40, p = 0.02,

table 8) when inspiring high FCO 2 air than when inspiring CO 2free air.

Table 8. The effects of acute exposure of Sandevall's jirds to high FCO 2 (0.07) on physiological and behavioral costs. Linear mixed effect models were used to test for differences among individuals breathing CO 2free air or air with high FCO 2 and tested for differences in resting metabolic rate (RMR), mean metabolic rate (MR), respiration rate ( fr) and in proportion of time spent sleeping, grooming, and being active but not grooming (other). Coefficients for the data were obtained with linear mixedeffects model (individual jirds as a random factor). Reference level in the models was measurement at FCO 2 = 0 (Intercept). The fixed effect was treatment ( FCO 2 = 0.07). SE = standard error. DF = degrees of freedom

Value SE DF Fvalue pvalue RMR (Intercept) 395.46 15.95 7 949.53 <.0001 Treatment 656.78 239.50 7 7.52 0.0288 MR (Intercept) 580.25 80.85 7 66.96 0.0001 Treatment 1744.01 568.22 7 9.42 0.0181 fr (Intercept) 77.46 3.54 7 1226.42 <.0001 Treatment 652.27 66.91 6 95.04 1e04 sleeping (Intercept) 1.09 0.15 7 37.59 0.0005 Treatment 5.48 1.17 7 22.34 0.0023 grooming (Intercept) 0.04 0.02 7 10.11 0.0155 Treatment 0.50 0.48 7 0.98 0.3328 other (Intercept) 0.13 0.09 7 7.41 0.0296 Treatment 3.16 1.08 7 8.40 0.0223

Results

Figure 21. Sundevall’s jirds have higher respiration rate (A), resting metabolic rate (B) and mean metabolic rate (C) when inspiring air with high

FCO 2 (0.07) than when inspiring CO 2free air (p > 0.05). N= 8, Error bars depict SD.

Results

2.3.5 Experiment 5. The effect of sequential and acute changes in inspired FCO 2 on energy expenditure in jirds. The minimum MR of 9 Sundevall's jirds at an ambient temperature of 30 °C, when breathing dry CO 2free air for 13.5 hours was 659.73 ± 196.29 mW and their RQ = 0.78 ± 0.1 (mean ± SD). This minimum MR is significantly higher than both RMR and MR (395.46 ± 53.55 mW and 580.25 ± 182.72 mW, respectively ) measured for jirds in the experiment described the previous section (experiment 4, section 2.3.4) (Student's ttest, t = 8.51, p < 0.001 and MannWhitney rank sum test, T = 50, p = 0.03, respectively). In both cases MR was measured in open flow respirometry systems, but in the first experiment the animals were placed in metabolic chambers almost twice the volume of the chambers in the present experiment (1.612 L vs. 0.85 L). Apparently, even though MR in this experiment was calculated using the lowest 100 seconds sequence of O the jirds in the smaller chamber did not attain resting MR. Hence, from here on, I refer to the MR values in this experiment as

MR during sequential step change in FCO 2 (MR S) and MR during acute change in FCO 2

(MR A).

During measurements made in CO 2 free air, MR in control animals gradually decreased with time, from 782.08 mW during the first 2 h of measurements to 603.82 mW during the last 2 h (figure 22). SMR decreased from 933.01 mW in the first time block to 766.98 mW the second time block and remained stable during the following hours of measurements (Figure 22).

Results

Figure 22. Metabolic rate (mW) of nine female Sundevall's jirds during sequential step changes in fractional concentration of CO 2 (FCO 2 full circles) and with no added CO 2 (empty circles). Values are means ± SD.

No difference was found between MR diff in the first time block (intercept) and MR diff in time blocks 4, 5 and 6 ( FCO 2 = 0.04, 0.07 and 0) (table 9). The significance of the differences between MR diff in the first time block and MR diff in time blocks 2 and 3 ( FCO 2 =0.01 and 0.02) were marginal (p = 0.047 and p = 0.05, respectively). Tukey multiple comparisons revealed no difference between any of the treatments (table 10).

Results

Table 9. The effect of sequential step changes in FCO 2 on the response variable MR diff (the difference between a jird's MR in the treatment minus the jird’s MR in the corresponding control). Coefficients for the data were obtained with linear mixedeffects model (individual jirds as a random factor). Reference level in the models was the first period of measurement

at FCO 2 = 0 (Intercept). SE = standard error. DF = degrees of freedom

SE DF tvalue pvalue (Intercept) 0.235 40 1.904 0.064 FCO 2 = 0.01 0.185 40 2.046 0.047 FCO 2 = 0.02 0.185 40 2.016 0.05 FCO 2 = 0.04 0.185 40 0.72 0.475 FCO 2 = 0.07 0.185 40 0.144 0.886 FCO 2 = 0 0.185 40 0.036 0.971

Results

Table 10: Tukey multiple comparisons for the linear mixed effect model based on sequential step

changes in FCO 2. The reference level in the model was the first block of time during the first two hours of measurement when FCO 2 = 0. An effect of the change in FCO 2 on jirds RMR would be indicated by a significant difference between MR diff of different treatments but not at FCO 2 = 0. Tukey multiple comparisons were used to test for differences between experimental groups within levels. SE = standard error.

0* is the last level during the last time block at FCO 2 = 0.

Multiple comparison Estimate SE z value pvalue hypothesis

FCO 2 : 0.01 0 0.379 0.185 2.047 0.316

FCO 2 : 0.02 0 0.373 0.185 2.017 0.333

FCO 2 : 0.04 0 0.133 0.185 0.720 0.980

FCO 2 : 0.07 0 0.027 0.185 0.144 1.000

FCO 2 : 0* 0 0.007 0.185 0.036 1.000

FCO 2 : 0.02 0.01 0.006 0.185 0.030 1.000

FCO 2 : 0.04 0.01 0.246 0.185 1.326 0.770

FCO 2 : 0.07 0.01 0.352 0.185 1.903 0.400

FCO 2 : 0* 0.01 0.372 0.185 2.011 0.336

FCO 2 : 0.04 0.02 0.240 0.185 1.296 0.787

FCO 2 : 0.07 0.02 0.347 0.185 1.873 0.419

FCO 2 : 0* 0.02 0.367 0.185 1.981 0.354

FCO 2 : 0.07 0.04 0.107 0.185 0.576 0.993

FCO 2 : 0* 0.04 0.127 0.185 0.684 0.984

FCO 2 : 0* 0.07 0.020 0.185 0.108 1.000

RMR during step changes in FCO 2 decreased from 709.81 ± 162.04 mW in the first time block ( FCO 2 = 0) to 684.78 ± 169.54 in the second time block (FCO 2 = 0.07) and to 578.5 ±

124.72 mW in the last time block ( FCO 2 = 0) (Figure 23). No differences were found between MR diff during the first time block (intercept) and MR diff in time blocks 2 and 3 (table 11).

Results

Figure 23. Metabolic rate (mW) of nine female Sundevall's jirds during acute changes in fractional concentrations of CO 2 (FCO 2 full circles) and with no added CO 2 (empty circles). Values are means ± SD.

Table 11. The effect of acute changes in FCO 2 on the response variable MR diff (the difference between a jird's MR in the treatment minus the jird’s MR in the corresponding control). Coefficients for the data were obtained with linear mixedeffects model with individual jird as a random factor. Reference level in the models was the first period of measurement at

FCO 2 = 0 (Intercept). SE = standard error. DF = degrees of freedom

SE DF tvalue pvalue (Intercept) 0.218 16 0.984 0.34

FCO 2 = 0.07 0.22 16 0.182 0.858

FCO 2 = 0 0.22 16 0.799 0.436

Discussion

2.4 Discussion

CO 2 tolerance

Sundevall’s jirds voluntarily tolerated FCO 2 = 0.07, and showed no preference for CO 2 free air when given the choice. Most previous studies of voluntary tolerance of high FCO 2 exposed rats and mice to much higher FCO 2 than in the present study (minimum FCO 2 = 0.2), and found that rodents actively avoided such environments. For example, rats and mice withdraw from FCO 2 > 0.20 and spend less time in environments with such high FCO2 when given a choice (Leach et al., 2002, Ziemann et al., 2009). However, an FCO 2 of 0.2 is not biologically realistic and is more than twice as high as the FCO 2 found in my experiment in seminatural burrow of jirds (figure 15, section 1.3.4).

The results for voluntary tolerance of high FCO 2 concur with a study of voluntary high

FCO 2 tolerance in two semifossorial desert rodents, Egyptian spiny mice (Acomys cahirinus ) and Anderson's gerbils ( Gerbillus andersoni allenbyi ) (Prange et al., 1998). When allowed to shuttle freely between a pair of chambers constructed very similarly to the pair used in the present study, one with room air and one with high FCO 2 (0.05 or 0.1) for 15 min, there were no differences between the proportions of the total time or average bout length spent in room air, or FCO 2 = 0.05 by either species, or in FCO 2 = 0.1 by Anderson's gerbil (Egyptian spiny mice were not tested in FCO 2 = 0.1). Like jirds, neither species showed a preference for either environment, suggesting that, at least in the short term, these two species also voluntarily tolerate high FCO 2 within the range found in burrows.

Oxygen dissociation curve and Bohr effect

The blood gas properties ( P50 and Bohr effect) of Sundevall’s jirds were investigated along with those of the fat sand rat. The measured P50 of Sundevall's jirds is similar or slightly higher than that predicted by the allometric equation of SchmidtNielsen and Larimer (1958) (table 6) and the jirds' Bohr effect falls within typical mammalian values (appendix, table 16), suggesting that jirds do not have increased Hb affinity for O 2 or enhanced O 2 unloading at the tissues. The P50 (47.88 ± 4.72 mmHg) of fat sand rats was measured at pH=7.17, and found to be similar to that of Sundevall's jirds at the same pH range (50.5 mmHg, figure 20). Since predicted P50(7.4) of the two rodents is also similar (table 6), I conclude that, like jirds, fat sand rats also do not have increased Hb affinity for O 2. The conclusions I drew in Part 1 of this research, namely, that respiratory gas levels in the burrow of Sundevall’s jirds commonly differ only little from atmospheric air, bring me to

Discussion

conclude also that jirds do not require particular adaptations of increased HbO2 affinity. In the most extreme case, in which I made the walls of an artificial burrow impermeable to gas

exchange, FCO 2 inside the nest chamber, when a mother was present with a brood reached

0.07 (figure 17, section 1.3.4). Such a high level of FCO 2 corresponds to an FO2 = 0.19. An 3 animal inspiring air with FO2 = 0.19 will have an alveolar PO2 (Pa O2) of about 17.28 kPa

(129.6 mmHg) and slightly lower blood PO2, enough to keep the blood of jirds saturated, as can be ascertained from the jird's ODC in figure 20. In freshly drawn blood samples taken from both Sundevall's jirds and fat sand rats there was considerable variation in [H +]. For example, the [H +] in the blood samples of 11 jirds ranged between 39.83 mmol/l (pH = 7.39) and 93.35 mmol/l (pH = 7.02). [H +] in blood samples of 9 fat sand rats ranged between 51.28 mmol/l (pH = 7.29) and 93.32 mmol/l (pH = 7.03). Invariably acid blood from the orbital sinus of mice and rat has often been measured (Gray and Steadman, 1964, Lahiri, 1975, Snyder, 1976, Jürgens et al., 1981). Three explanations for the ubiquitous acidity of rodent blood have been proposed: (1) normal PCO 2 in venous blood of small rodents may be as low as 2.6 kPa (20 mmHg) (Lahiri, 1975), and the reduced pH is a result of experimental conditions, since researchers tend to make measurements on blood equilibrated to a PCO 2 of ~ 5.3 kPa (40 mmHg); (2) the presence of lactate in the blood as a result of anaerobic glycolysis in blood cells which continues in blood that is waiting for measurement, and (Gray and Steadman, 1964, Lahiri, 1975, Snyder, 1976); and (3) the presence of lactate in the blood as a result of anaerobic glycolysis in muscle due to stress caused by struggle (Gärtner et al., 1980).

The PCO 2 values in the blood of both Sundevall's jirds and fat sand rats (table 7) were slightly lower than the conventionally accepted value of 5.3 kPa (40 mmHg) with which they were equilibrated by tonometry. However, a more likely explanation for the variation of acidity of the blood in the present study is the presence of lactate in the blood due to stress caused by struggle, since individual animals demonstrated different levels of struggle when handled. Furthermore, the blood samples of fat sand rats that were taken from the rodent

3 . . .. O 17.28 kPa .

The PCO 2 of jirds arterial blood is 4.51 ± 0.21 kPa (33.83 ± 1.64 mmHg) (table 4), lower than the conventionally accepted value of 5.3 kPa (40 mmHg). The RQ of jirds was found to be 0.78 (section 2.3.5 of this thesis).

Discussion

exhibition of Hai Bar Yotvata, which were not accustomed to being handled by humans, were significantly more acidic than those of the jirds that I raised (student ttest, t = 2.282, p = 0.034, table 7). Moreover, BarIlan et al., (1985) drew blood samples from cannulated fat sand rats (a method which does not cause additional stress to the animals) and reported much higher and more consistent arterial blood pH value of 7.396 ± 0.034 ([H +] = 40.17).

Blood buffer capacity

The blood buffering capacity of the fat sand rats in this study (table 7) is similar to that reported by BarIlan et al., (1985) (appendix, table 16). Sundevall's jirds have a slightly higher blood buffering capacity (appendix, table 16) than nonburrowing mammals such as man ( Homo sapiens 1.55 ± 0.04), dog ( Canis familiaris 1.53) (Lahiri, 1975), rabbit (Oryctolagus cuniculus , 1.56 ± 0.12, BarIlan et al., 1984), and rat (Rattus norvegicus , 1.39, Chapman and Bennett, 1975 or 1.51 ± 0.1, BarIlan et al., 1985). However, their buffering capacity is considerably lower than that of fat sand rats and completely fossorial species such as the valley pocket gopher (2.2 ± 0.4, Lechner, 1976, 2.67, Chapman and Bennett, 1975) and mountain pocket gopher (2.83 ± 0.43) (Lechner, 1976).

Discussion

Table 12. Comparison of blood acidbase variables of rodents. Values are mean ±SD

Pa CO 2 [HCO 3 ] Species pH Reference mmHg mmol/l Fat sand rat 7.39 ± 0.03 30.5 ± 2.9 18.8 ± 2.5 Psammomys obesus BarIlan et al., In vivo Rat 1985 7.45 ± 0.009 33.8 ± 1.2 23.0 ± 1.3 Rattus norvegicus Rat Girard et al., 7.48 ± 0.008 31.5 ± 0.8 24.3 ± 0.2 In vivo Rattus norvegicus 1983 Pocket gopher 28.1 ± 0.89 Thomomys bottae Chapman and In vitro Rat Bennett, 1975 19.8 ± 1.0 Rattus norvegicus Rat BrunPascaud 7.47 ± 0.02 34.5 ± 3.0 25.5 ± 1.5 Rattus norvegicus et al., 1982 Guinea pigs 7.44 ± 0.03 35.7 ± 4.4 24.4 ± 2.8 Cavia porcellus BarIlan and In vivo Rat Marder, 1980 7.45 ± 0.02 35.5 ± 2.6 24.7 ± 1.9 Rattus norvegicus

As mentioned in the previous section, the blood pH of both animal species was relatively low and inconsistent. The bicarbonatecarbonic acid system, as assayed by the bicarbonate content of the blood, does not appear to be especially developed in jirds or fat sand rats as it was found to be in the pocket gopher (Chapman and Bennett, 1975). The blood [HCO 3 ] and

PCO 2 of jirds (table 7) is in the lower range of values reported for other rodent species (table 12). The blood [HCO 3 ] of fat sand rats (table 7) is considerably lower, and Pa CO 2, higher than reported for this species by BarIlan et al., (1985) (table 12), despite the fact that in the present study these variables were obtained at a much lower blood pH, and are therefore expected to be higher and lower, respectively. The differences in the values of blood variables may be contributed to the different methods of sampling. BarIlan et al., measured blood variables in blood that was drawn directly from the arteries of cannulated animals, while in the present study I equilibrated blood with gas mixtures by tonometry, thus skirting

Discussion

the role of the reabsorption of [HCO 3 ] by the kidneys. The difference in blood gas variables between the studies may underscore the important role of the kidneys in maintaining short term changes in acidbase balance in small rodents, as suggested by Lai et al., (1973) and BarIlan and Marder (1985).

Behavioral response and physiological costs in jirds inspiring high FCO 2

In some fossorial and semifossorial mammals the ventilatory response to CO 2 was found to be lower than that of nonfossorial species ( i.e. Bentley et al., 1967, Darden, 1972, Withers, 1975, Boggs et al., 1984) because, for an animal living in an environment where

FCO 2 is high at times, an increase in ventilation would not aid in CO 2 elimination. The ventilatory response of Sundevall's jirds to breathing air with FCO 2 = 0.07 was relatively high, The 59% percent increase in fr was greater than that of the completely fossorial pocket gopher and Middle East blind mole rat , similar to that of the spear nosed bat, Phyllostomus discolor , but lower than that of the laboratory rat (table 12).

In mammals, an increased ventilatory response to CO2 is caused by either an increase in fr, 4 increased tidal volume (VT), or both (examples in table 11). The product of VT and fr is total ventilation 5, given in liters (BTPS 6) per minute (Boron and Boulpaep, 2008), is usually

applied in order to compare between the sensitivity of the ventilatory response to CO 2 between different species ( i.e. Arieli and Ar, 1979, Boggs et al., 1984). Therefore the total

ventilation of jirds breathing air with FCO 2 = 0.07 may be even higher than the 59% increase which I have found.

A lowerthanexpected ventilatory response to CO 2 in burrowing species may also represent an energy conserving strategy for some species (Darden, 1972, Boggs et al., 1984) as in the case of several fossorial and semifossorial mammals which did not change their

MR when breathing air containing FCO 2 from 0.01 to 0.1 (examples in table 15). However, RMR of jirds in the present experiment increased significantly when breathing air containing

FCO 2 = 0.07, suggesting that inspiring air with such high FCO 2 is physiologically demanding

4 The volume of gas inhaled and exhaled at each breath.

5 Often termed minute volume ( Vmin )

6 BTPS: Body Temperature and Pressure Saturated. In respiratory physiology lung volumes and flows are standardized to body temperature (37 °C or 310 K), barometric pressure at sea level and lungs saturated with vapor pressure (47 mmHg or 6.2 kPa).

Discussion

of jirds and, therefore, jirds are not especially adapted to inspire such high levels of CO 2.

Furthermore, in a high FCO 2 environment the jirds showed recognizable signs of discomfort

as they were more active and slept less than jirds in CO 2free air, resulting in a significant increase in their MR. When the movement of the jirds was confined in smaller metabolic chambers their minimum MR was significantly higher than the RMR and MR, which were measured in the larger chambers, making it difficult to draw conclusions from the experiments of step changes in inspired FCO 2. Jird MR was especially high during the first time block of the sequential step change experiment (figure 22), which was the first time that the jirds were placed in the respirometry system. Their high MR was probably the result of stress. In subsequent measurements, MR of the jirds during the first time blocks was considerably lower (control measurements in figure 22 and measurements of acute changes in inspired

FCO 2 in figure 23). The jirds' high MR during the first time block of the sequential step change experiment resulted in a large MR diff . This is probably the reason that I found a

marginally significant difference between MR diff of the first time block and MR diff of time blocks 2 and 3 (when inspired FCO 2 were 0.01 and 0.02, respectively) was detected in the statistical analysis (table 9). Furthermore, Tukey multiple comparisons analysis revealed no difference among any of the treatments (table 10). No changes in MR were found during

acute changes in inspired FCO 2. Thus, I conclude that changes in inspired FCO 2, whether sequential or acute, had negligible effects on the MR of Sundevall's jirds.

Discussion

Table 13. Examples of the mechanisms leading to increased ventilatory response in different

species of mammals breathing air containing high FCO 2. VT = tidal volume, fr = respiration frequency (breaths/min).

Species Mechanism Reference Nine banded armadillo f Boggs et al., 1998 Dasypus novemcincus r Merriam's kangaroo rat f Soholt et al., 1973 Dipodomys merriami r Echidna, V Bentley et al., 1967 Tachyglossus aculeatus T Spiny rat V T Barros et al., 1998 Proechimys yonenagae At FCO 2 = 0.05 Middle East blind mole rat Both f and V Arieli and Ar, 1979 Spalax ehrenbergi r T Rats Both f and V Arieli and Ar, 1979 Rattus norvegicus, Wistar r T Spiny rat Both f and V r T Barros et al., 1998 Proechimys yonenagae At FCO 2 = 0.10 Spiny rat Both f and V r T Barros et al., 1998 Proechimys iheringi At FCO 2 = 0.10

Table 14. The increase in respiration frequency ( fr) (breaths/min) of different mammals as a response to breathing air with high FCO 2 compared to breathing "room" air.

Species FCO 2 (%) increase in fr Reference Pocket gopher 0.07 27.4 Darden, 1972 Thomomys bottae Middle East blind mole rat 0.085 50 Arieli and Ar, 1979 Spalax ehrenbergi Rats 0.085 81.25 Arieli and Ar, 1979 Rattus norvegicus, Wistar Spear nosed bat 0.07 65 Walsh et al., 1996 Phyllostomus discolor Sundevall's jird 0.07 59 Present study Meriones crassus

Discussion

Table 15. Examples of fossorial and semifossorial rodents that did not change their metabolic rate when breathing air containing high FCO 2 and where stated, low FO2.

Species Air composition Reference California ground squirrels FO = 0.173 and FCO = 0.035 Baudinette, 1974 Spermophilus beecheyi 2 2 Middle East blind mole rat FO = 0.055 and FCO = 0.13 Arieli et al., 1977 Spalax ehrenbergi 2 2 Rat FCO = 0.02 or 0.05 Saiki and Mortola, 1996 Rattus norvegicus 2 Spiny rats Proechimys

yonenagae and P. iheringi FCO 2 = 0.05 or 0.1 Barros et al., 1998 and rats Rattus norvegicus Spiny rat FCO 2 = 0.03 or 0.05 Barros et al., 2004 Clyomys bishopi

Summary

SUMMARY I have characterized the means by which a rodent burrow with more than one entrance is ventilated, and offered a new explanation for the underlying mechanism of wind induced ventilation. In addition I have addressed, for the first time, the question of the immediate respiratory environment that semifossorial rodents experience inside their burrows. The

resident animal is usually the main source of CO 2 in a burrow system and, in its immediate

vicinity [CO 2] may rapidly become very different from atmospheric. Notwithstanding, and probably due to technical difficulties, one of the most important limitations of most studies concerning respiratory gas concentrations in burrows is that the location of the animal within the burrow system at the time of sampling was unknown (Roper et al., 2001). I succeeded in overcoming this difficulty, and obtained a realistic depiction of the gaseous milieu inside a burrow by measuring [CO 2] continuously inside nest chambers while (1) simulating the respiration of a resident animal in the chamber, (2) recording the location of the animal with respect to the chamber, allowing me to confirm the connection between the animal’s location and [CO 2] inside the nest chamber, and (3) monitoring live female Sundevall’s jirds and their brood inside the nest chamber.

This is the first study in which [CO 2] in a nest chamber was continuously monitored over relatively long periods of time, and was also related to external environmental conditions and the behavior of the resident animal. Ultimately, I suggest that the main factors which affect the respiratory environment experienced by the animal inside the burrow are: (1) its location within the burrow system in relation to the burrow’s openings and the amount of time it spends in a location, (2) the degree of ventilation by wind induced convection, determined by ambient wind speed and the number and position of burrow entrances, and (3) the degree of gas exchange between the burrow and the atmosphere through the soil, probably determined by soil porosity. Additional factors that may influence gas exchange between the burrow and the atmosphere are the physical variables of the burrow system, such as the number of openings, length of tunnels and the nest chamber depth. Mathematical models predict that the depth of nest chamber may not have much effect on gas composition in it (Withers, 1978,

Wilson and Kilgore, 1978), but Maclean (1981) reported that [CO 2] did increase significantly with increasing depth from 30 cm to 70 cm, in an occupied, artificial burrow of 13lined ground squirrels ( Spermophilus tridecemlineatus ). In their model Wilson and Kilgore (1978) considered gas diffusion to be the sole mechanism for gas exchange and showed that in that case openings of tunnels further than three body length from a mammal have a negligible effect on its respiratory microenvironment. Maclean's findings that gas 82

Summary

composition was unaffected by changing tunnel's length or plugging burrow entrances supports the mathematical model. Although it is often stated that burrowing mammals live in "different" respiratory conditions ( e.g., Baudinette, 1974, Withers, 1975, Barros et al., 2004), my findings suggest that the CO 2 levels, which Sundevall’s jirds commonly experience in their openended burrows do not exceed 2 2.5%. This supports the existing literature, in which most studies of burrow environments (listed in table 3 in Roper et al., 2001) report respiratory gas levels to differ by only 1 2% from ambient. The results of the present study demonstrate that, even at low wind speed at the surface, the random penetration of wind eddies into burrow openings should be sufficient to maintain a large gradient of respiratory gas concentrations over short distances inside the burrow itself, enough to keep [CO 2] relatively low even in deep and remote parts of the burrow. It is noteworthy, however, that maximum values as extreme as

4% and 6% CO 2 have been reported from natural, openended burrows (respectively Baudinette, 1974 and Studier and Proctor, 1971). As stated above, an extreme reduction in soil porosity may result in much higher burrow [CO 2]. In my experiment with live female jirds, when the walls of the burrow were sealed, making it completely impermeable to gas

transfer through the soil, [CO 2] in the nest chamber reached 7 8%. In natural burrow systems extreme reduction in soil porosity might occur after heavy rain due to water that completely or partially clogs pores, (Bornstein et al., 1980), and/or due to soil crusting and sealing (Panini et al., 1997). Nevertheless, further investigation is necessary in order to find

out how exactly soil water content and soil crusts effect burrow [CO 2]. For example, although

soil water content may reduce soil porosity, CO 2 is also extremely soluble in water (Denny, 1993). Consequently the presence of water in the soil nearby the burrow may actually facilitate the removal of [CO 2] from it. Likewise, soil crusting and sealing bring about changes in the pore system of the top layer of the soil, but CO 2 from the nest chamber disperses first into the burrow system, then diffuses through still air, and then into the soil through the entire surface of the burrow, in all directions. Therefore even in the extreme case of complete surface sealing of the top layer of the soil diffusion into the soil matrix from the burrow may still take place, thus the burrow is never completely sealed off from its environment, like in the case of the impermeable burrow used in my experiments. In any

event, it is of note that despite the high [CO 2] in the nest chamber the female and pups in the impermeable burrow lived "normally" for the entire three weeks of measurement, dam and pups were all in good health when trapped after the experiment. Such levels of inspired [CO 2] produce respiratory distress in humans and other terrestrial mammals (Davenport, 1974, 83

Summary

Dejours, 1981, Schaefer, 1982, Halpern et al., 2004). It would be of value to test whether,

when [CO 2] reaches a certain level in the nest chamber, the female responds by leaving the

chamber, allowing [CO 2] to decrease. The physiological relevance of my findings were examined in the second part of this

thesis. Sundevall’s jirds voluntarily tolerated FCO 2 = 0.07, and showed no preference for

CO 2free air when given the choice. The measured blood PCO 2 at 50% saturation ( P50 ) of Sundevall's jirds is similar or even slightly higher than that predicted by the allometric equation of SchmidtNielsen and Larimer (1958) and their Bohr effect value falls within

typical mammalian values, suggesting that jirds do not have increased Hb affinity for O 2 or enhanced O 2 unloading at the tissues. Sundevall's jirds have slightly higher blood buffering capacity than nonburrowing mammals; however, their buffering capacity is considerably lower than that of the semifossorial fat sand rat and other completely fossorial species. When breathing air containing high FCO 2 jirds showed recognizable signs of discomfort, they were

more active and slept less than jirds in CO 2free air and as a result increased their MR. In

addition, the RMR of jirds increased significantly when inspiring air containing FCO 2 = 0.07.

Although Sundevall's jirds are apparently tolerant of high CO 2 in the air they breathe I found no specific physiological traits to suggest that jirds are especially physiologically adapted to a

high CO 2 environments. Burrows are an outcome of natural selection and can be viewed as part of the burrow builder's "extended organism", while their configuration can be considered as part of its "extended physiology"(Turner 2000). I began this research with this concept in mind, although I only later did I read Turner's book. I proposed that burrow dwellers are either physiologically adapted to tolerate high FCO 2 in their environment, or alternatively, the architecture of burrows is such that their ventilation prevents buildup of CO 2. I found that the semifossorial Sundevall's jird is not physiologically adapted to breathing air containing

high [CO 2], but rather the buildup of CO 2 in their environment is attenuated by ventilation driven by turbulence in the lower part of atmospheric boundary layer. The mechanism of burrow ventilation by the penetration of eddies may not only be widespread, but may, in fact, be the dominant mode of gas exchange in natural animal burrows. A better understanding of unsteady state gas exchange in burrows would give us a better understanding of how animals adapt to the burrow environment.

References

References 1. Ar A. (1987). Physiological adaptations to underground life in mammals. In: Comparative physiology of environmental adaptations (Vol. 2). Dejours P. (Ed.) Strasbourg. Pp. 208221. 2. Ar A., Arieli R. and Shkolnik A. (1977). Bloodgas properties and function in the fossorial mole rat under normal and hypoxichypercapnic atmospheric conditions. Respiration Physiology 30: 201218. 3. Ar A., Barnea A., YomTov Y.and MerstenKatz C. (2004). Woodpecker cavity aeration: a predictive model. Respiratory Physiology and Neurobiology 144: 237249. 4. Ar A. and Piontkewitz Y. (1992). Nest ventilation explains gas composition in the nestchamber of the European beeeater. Respiration Physiology 87: 407418. 5. Arieli R. (1979). The atmospheric environment of the fossorial mole rat ( Spalax ehrenbergi ): effects of season, soil texture, rain, temperature and activity. Comparative Biochemistry and Physiology, A 63: 569575. 6. Arieli R. and Ar A. (1979). Ventilation of a fossorial mammal ( Spalax ehrenbergi ) in hypoxic and hypercapnic conditions. Journal of Applied Physiology 47: 10111017 7. Arieli R., Ar A. and Shkolnik A. (1977). Metabolic responses of a fossorial rodent (Spalax ehrenbergi ) to simulated burrow conditions. Physiological Zoology 50: 61 75. 8. Bakken G.S. and Rowe M.F. (2011). Is it necessary to paint metabolism chamber walls? Journal of Thermal Biology 37: 100107. 9. Banchero N., Grover R.F. and Will J.A. (1971). Oxygen transport in the Llama ( Lama glama ). Respiration Physiology 13: 102115. 10. BarIlan A. and Marder J. (1980). Acid base status in unanesthetized, unrestrained guinea pigs. European Journal of Physiology 384: 9397. 11. BarIlan A. and Marder J. (1985). The effects of the inhibition of the renal carbonic anhydrase on the blood acidbase status in hypercapnic rats. Comparative Biochemistry and Physiology, A 81: 121126.

12. BarIlan A., Marder J. and Samueloff S. (1984). In vivo and in vitro CO2 titration curves in the rabbit: Adaptation to hypercapnic conditions. Comparative Biochemistry and Physiology, A 78: 217220. 13. BarIlan A., Marder J. and Samueloff S. (1985). Blood acidbase status, whole blood and whole body buffer values in sand rats (Psammomys obesus ) exposed to hypercapnia. Comparative Biochemistry and Physiology, A 81: 571576. 85

References

14. Barros R.C.H., Abe A.S., Carnio E.C. and Branco L.G.S (2004). Regulation of breathing and body temperature of a burrowing rodent during hypoxichypercapnia. Comparative Biochemistry and Physiology, A 138: 97104. 15. Barros R.C.H., Oliveira E.S., Rocha P.L.B. and Branco L.G.S. (1998). Respiratory and metabolic responses of the spiny rats Proechimys yonenagae and P. iheringi to

CO 2. Respiration Physiology 111: 223231. 16. Bartels H., Schmelzle R. and Ulrich S. (1969). Comparative studies of the respiratory function of mammalian blood. V. Insectivora: shrew, mole and nonhibernating and hibernating hedgehog. Respiration Physiology 7: 278286. 17. Baudinette R.V. (1974). Physiological correlates of burrow gas conditions in the California ground squirrel. Comparative Biochemistry and Physiology, A 48: 733 743. 18. Belnap J. and Troxler T. (2006). Soil lichen and moss cover and species richness can be highly dynamic: the effects of invasion by the annual exotic grass Bromus tectorum and the effects of climate on biological soil crusts. Applied Soil Ecology 32:63–76. 19. BenPorat G., Adler J.H. and Samueloff S. (1976). On thermoregulation in Psammomys obesus , Meriones crassus and Meriones tristrami . Israel Medical Association Journal 12: 897901. 20. Bentley P.J., Herreid II C.F. and SchmidtNielsen K. (1967). Respiration of a monotreme, the echidna, Tachyglossus aculeatus . American Journal of Physiology 212: 957961. 21. Boggs D.F., Kilgore D.L.J. and Birchard G.F. (1984). Respiratory physiology of burrowing mammals and birds. Comparative Biochemistry and Physiology, A 77:17. 22. Boggs D.F., Frappell P.B. and Kilgore D.K. Jr (1998). Ventilatory, cardiovascular and metabolic responses to hypoxia and hypercapnia in the armadillo. Respiration Physiology 113: 101109. 23. Bohr C., Hasselbalch K. and Krogh A. (1904). Ueber einen in biologischer beziehung wichtigen einfluss, den die kohlensäurespannung des blutes auf dessen sauerstoffbindung übt. Skandinavisches Archiv fur Physiologie 16: 402–412. 24. Bornstein J., Hedstrom W.E. and Scott F.R. (1980). Oxygen diffusion rate relationships under three soil conditions. LSA Experiment Station Technical Bulletin 98: 112.

References

25. Boron W.F. and Boulpaep E.L. (2008). Medical Physiology. 2 nd edition. Elsevier Health Sciences. 26. Brandis K. (2008). Acidbase physiology. http://www.anaesthesiamcq.com 27. BrunPascaud M., Gaudebout C., Blayo M.C. and Pocidalo J.J. (1982). Arterial blood gases and acidbase status in awake rats. Respiration Physiology 46: 45–57. 28. Carmi G. and Berliner P. (2009). The effect of soil crust on the generation of runoff on small plots in an arid environment. Catena 74: 3742. 29. Chapman R.C. and Bennett A.F. (1975). Physiological correlates of burrowing in rodents. Comparative Biochemistry and Physiology, A 51: 599603. 30. Clausen G. and Ersland A. (1968). The respiratory properties of the blood of the hibernating hedgehog Erinaceus europaeus L . Respiration Physiology 5: 221233.

31. Clausen G. and Ersland A. (1971) Blood O 2 and acid base changes in the beaver during submersion. Respiration Physiology 11: 104112. 32. Dalgliesh W. A. and Boyd D. W. (1962). CBD28. Wind on Buildings. Canadian Building Digest. 33. Damiani R. Modesto S.,Yates A. and Neveling J. (2003). Earliest evidence of cynodont burrowing. Proceeding of the Royal Society 270: 17471751. 34. Darden T.R. (1972). Respiratory adaptations of a fossorial mammal, the Pocket gopher ( Thomomys bottae ). Journal of Compartive Physiology 78: 121137. 35. Davenport H.W. (1974). The ABC of acidbase chemistry: the elements of physiological bloodgas chemistry for medical students and physicians 6th edition, revised. Chicago: The University of Chicago Press. The University of Chicago Press, Chicago & London. 36. Degen A., Khokhlova I.S. and Kam M. (2011). Milk production of the dam limits the growth rate of Sundevall’s jird ( Meriones crassus) pups. Mammalian Biology 76: 285–289. 37. Degen A.A. (1997). Ecophysiology of small desert mammals. Springer Verlag Berlin Heidelberg. 38. Degen A.A., Khokhlova I.S., Kam M. and Nagy K.A. (1997). Body size, granivory and seasonal dietary shifts in desert gerbilline rodents. Functional Ecology 11: 5359. 39. Dejours P. (1981). Principles of comparative respiratory physiology. 2 nd edition. Elsevier /NorthHolland Biomedical Press, Amsterdam, New York, London. 40. Denny M. (1993). Air and water: the biology and physics of life’s media. Princeton University Press. 87

References

41. Douglas W.H.J., Schaefer K.E., Messier A.A. and Pasquale S.M. (1979). Proliferation

of pneumocyte II cells in prolonged exposure to 1% CO 2. Undersea Biomedical Research, Submarine Supplement. S135142. 42. Ebensperger L.A. and Bozinovic F. (2000). Energetics and burrowing behaviour in the semifossorial degu Octodon degus (Rodentia: Octodontidae). Journal of Zoology 252: 179186. 43. Forman C.W. (1954). A comparative study of the oxygen dissociation of mammalian hemoglobin. Journal of Cellular and Comparative Physiology 44: 421429. 44. Ganot Y., Dragila M.I. and Weisbrod N. (2012). Impact of thermal convection on air circulation in a mammalian burrow under arid conditions. Journal of Arid Environments 84:5162. 45. Gärtner K., Büttner D., Döhler K., Friedel R., Lindena J., and Trautschold I. (1980) Stress response of rats to handling and experimental procedures. Laboratory Animals 14: 267274. 46. Giardina B. Brix O. Colosimo A. Petruzzell R. Cerroni L. and Condo S.G. (1990). Interaction of hemoglobin with chloride and 2,3bisphosphoglycerate: a comparative approach. European Journal of Biochemistry 194: 6165. 47. Girard P., BrunPascaud M. and Pocidalo J.J. (1983). Acidbase status of awake rats after cannulation of aorta and vena cava. Kidney International 24: 795799. 48. Goyal S. P. and Ghosh P. K. (1993). Burrow structure of two gerbil species of Thar Desert, India. Acta Theriologica 38: 453456. 49. Gray L.H., and J.M. Steadman (1964). Determination of the oxyhaemoglobin dissociation curves for mouse and rat blood. Journal of Physiology 175: 161171. 50. Greene R.A. and Reynard C. (1932). The influence of two burrowing rodents, Dipodomys spectabilis spectabilis (kangaroo rat) and Neotoma albigula albigula (pack rat) on desert soils in Arizona. Ecology 13: 7380. 51. Haab P.E., J. Piper, and H. Rahn (1960). Simple method for rapid determination of an

O2 dissociation curve of the blood. Journal of Applied Physiology 15: 11481148. 52. Hall F.G. (1966). Minimal utilizable oxygen and the oxygen dissociation curve of blood of rodents. Journal of Applied Physiology 21: 375378. 53. Hall F.G., Dill D.B. and Barron D.S.G. (1936). Comparative physiology at high altitude. Journal of Cellular and Comparative Physiology 8:301313.

References

54. Halpern P., Raskin Y., Sorkine P. and Organezov A. (2004). Exposure to extremely high concentrations of carbon dioxide: a clinical description of a mass casualty incident. Annals of Emergency Medicine 43: 196199. 55. Hamburger H.J. (1918). Anionenwanderungen im serum und blut unter dem

einfluss von CO 2, saure and alkali. Biochemische Zeitschrift 86: 309327 56. Harrison D.L. and Bates P.J.J. (1991). P. 354. The Mammals of Arabia. 2 nd edition. Sevenoaks: Harrison Zoological Museum. . 57. Hill A.V. (1910). The possible effects of the aggregation of the molecules of haemoglobin on its dissociation curves. Journal of Physiology 40: 47. 58. Hill R.W. Gordon A.W. and Anderson M. (2012). Animal physiology. 3 rd edition. Sinauer Associates Inc. Publishers. 59. Hillel D. (1998). Environmental soil physics. Academic Press, USA. 60. Hinze A. Pillay N. And Grab S. (2006). The burrow system of the African ice rat Otomys sloggetti robertsi . Mammalian Biology 71: 356 – 365. 61. Hölldobler B. and Wilson E.O. (1990). The Ants. Belknap Press. Cambridge, MA. 62. Hoogland J. L. (1995). The blacktailed prairie dog: social life of a burrowing mammal. The University of Chicago Press. Chicago, IL. 63. Hothorn T., Bretz F. and Westfall P. (2008) Simultaneous inference in general parametric models. Biometrical Journal 50: 346–363. 64. Jammalamadaka S. and Sarma Y. (1988). A correlation coefficient for angular variables. Statistical Theory and Data Analysis 2. North Holland: New York, NY. 65. Jammalamadaka S. Rao and SenGupta A. (2001). Topics in Circular Statistics, Section 8.2, World Scientific Press, Singapore. 66. Johansen K., Lykkeboe G. Webber R.E. and Maloiy G.M.O. (1976). Blood respiratory properties in the naked mole rat, Heterocephalus glaber , a mammal of low body temperature. Respiration Physiology 28: 303314. 67. Jürgens K.D., Bartels H., and R. Bartels (1981). Blood oxygen transport and organ weights of small bats and small non flying mammals. Respiration Physiology 45: 243 260. 68. Kam M., CohenGross S., Khokhlova I.S., Degen A.A. and Geffen E. (2003). Average daily metabolic rate, reproduction and energy allocation during lactation in the Sundevall Jird Meriones crassus . Functional Ecology 17: 496503. 69. Kennerly T.E., Jr (1964). Microenvironmental conditions of the pocket gopher burrow. Texas Journal of Science. 16: 395441. 89

References

70. Khokhlova I.S., Fielden L.J., Degen A.A. and Krasnov B.R. (2012). Ectoparasite fitness in auxiliary hosts: phylogenetic distance from a principal host matters. Journal of Evolutionary Biology 25: 20052013. 71. Khokhlova I.S., Serobyan V., Krasnov B.R. and Degen A.A. (2009). Is the feeding and reproductive performance of the flea, Xenopsylla ramesis, affected by the gender of its rodent host, Meriones crassus ? The Journal of Experimental Biology 212: 1429 1435. 72. Kinlaw A. (1999). A review of burrowing by semifossorial vertebrates in arid environments. Journal of Arid Environment 41: 127145. 73. Kleiber M. (1932). Body size and metabolism. Hilgardia 6: 315351. 74. Kleineidam C., Ernst R. and Roces F. (2001). Windinduced ventilation of the giant nests of the leafcutting ant Atta vollenweideri . Naturwissenschaften 88: 301305. 75. Krasnov B., Shenbrot G., Khokhlova I., Degen A.A. and Rogovin K.A. (1996a). On the biology of Sundevall’s jird ( Meriones crassus Sundevall , 1842) (Rodentia: Gerbillinae) in the Negev Highlands, Israel. Mammalia 60: 375391. 76. Krasnov B., Shenbrot G., Khokhlova I. and Ivanitskaya E. (1996b). Spatial patterns of rodent communities in the Ramon erosion cirque, Negev Highlands, Israel. Journal of Arid Environments 32: 319327.

77. Kuhnen G. (1986). O 2 and CO 2 concentrations in burrows of euthermic and hibernating golden hamsters. Comparative Biochemistry and Physiology, A 84: 517 522. 78. Lahiri S. (1975). Blood oxygen affinity and alveolar ventilation in relation to body weight in mammals. American Journal of Physiology 229: 529536. 79. Lai Y.L., Martin E.D., Attebery B.A. and Brown E.B. Jr. (1973). Mechanisms of extracellular pH adjustments in hypercapnia. Respiration Physiology 19: 107114. 80. Leach M.C., Bowell V.A., Allan T.F. and Morton D.B. (2002). Aversion to gaseous euthanasia agents in rats and mice. Comparative medicine 52: 249257. 81. Lechner A.J. (1976). Respiratory adaptations in burrowing pocket gophers from sea level and high altitude. Journal of Applied Physiology 41: 168173. 82. Lighton J.R.B (2008). Measuring metabolic rates, a manual for scientists. Oxford University Press. 83. Lovegrove B.G. and Jarvis J.U.M. (1986). Coevolution between molerats (Bathyergidae ) and a geophyte, Micranthus (Iridaceae). Cimbabasia A 8: 7985.

References

84. Maclean G.S. (1981). Factors influencing the composition of respiratory gases in mammal burrows. Comparative Biochemistry and Physiology, A 69: 373380. 85. Marder J. and BarIlan A. (1975) . Acidbase metabolism in the albino rat during hypercapnia. Physiological Zoology 48: 282–289. 86. Marom S., Korine C., Wojciechowski M.S., Tracy C.R. and Pinshow B. (2006). Energy metabolism and evaporative water loss in the European freetailed bat and Hemprich’s longeared bat (Microchiroptera): species sympatric in the Negev desert. Physiological and Biochemical Zoology 79: 944–56. 87. Mathias M.L. and Freitas J.P. (1989). Oxygen transport and blood buffering capacity in the Mediterranean pine vole. Miscellania Zoologica (Barcelona) 13: 177180. 88. Mendelssohn H. and Yom Tov Y. (1999). Mammalia of Israel. The Israel Academy of Sciences and Humanities. Jerusalem. 89. Mendelssohn H. and YomTov Y. (1987). Plants and animals of the land of Israel, volume 7: Mammals. Alon A. (Ed.) Ministry of Defense Publishing House, Tel Aviv, Pp. 112114 (in Hebrew). 90. Messier A.A. and Schaefer K.E. (1973) The Bohr effect in chronic hypercapnia. Naval Submarine Medical Research Laboratory Report Number 732. 91. Morin J. and Van Winkel J. (1996). The effect of raindrop impact and sheet erosion on infiltration rate and crust formation. Soil Science Society of America journal 60: 12231227. 92. Nikol’skii A.A. and Khutorskoi M.D. (2001). Thermal characteristics of mammalian burrows in summer (using a burrow of the steppe marmot as an example). Doklady Biological Sciences 378: 240243. 93. Nikol’skii A.A. and Savchenko (2002). Effect of atmosphere temperature near the soil surface on air convection in mammalian burrows (as exemplified by of steppe marmot burrow). Doklady Biological Sciences 384: 242245. 94. NOAA (2012). Trends in atmospheric carbon dioxide. http://www.esrl.noaa.gov/gmd/ccgg/trends/global.html. Updated 2012. 95. Nowak R. (1999). Walker's Mammals of the World. Baltimore: John Hopkins University Press. 96. Olszewski J.L. and Skoczeń S. (1965). The airing of burrows of the mole Talpa europaea Linnaeus, 1758. Acta Theriologica, 10: 181–193.

References

97. Panini T., Torri D., Pellegrini S., Pagliai M. and Salvador Sanchis M.P. (1997). A theoretical approach to soil porosity and sealing development using simulated rainstorms. Catena 31: 199–218. 98. Parer J.T. and Metcalfe J. (1967). Respiratory studies of monotremes. II. Blood of the echidna ( Tachyglossus aculeatus ). Respiration Physiology 21: 307316. 99. Peinke D.M. and Brown C.R. (2005). Burrow utilization by springhares ( Pedetes capensis ) in the Eastern Cape, South Africa. African Zoology 40: 3744. 100. Pinheiro J., Bates D., DebRoy S., Sarkar D. and R.D.C. Team. 2011. nlme: Linear and Nonlinear Mixed Effects Models. R package version 3.1102. 101. Prange H.D., Reilly W.M. and Pinshow B. (1998). Voluntary tolerance of high levels

of CO 2 by desert rodents. The FASEB Journal 12: A419. 102. Prange H.D., Shoemaker Jr. J.L., Westen E.A., Horstkotte D.G. and Pinshow B.

(2001) Physiological consequences of O2dependent chloride binding to hemoglobin. Journal of Applied Physiology 91: 3338. 103. R developemnet core team (2011). R: A language and environment for statistical computing. R Foundation for Statistical Computing, Vienna, Austria. 104. Randall J.A., Rogovin K.A. and Shier D.M. (2000). Antipredator behavior of a social desert rodent: footdrumming and alarm calling in the great gerbil, Rhombomys opiums . Behavioral Ecology and Sociobiology 48: 110118. 105. Reichman O. and Smith S.C. (1990). Burrows and burrowing behavior by mammals. Current Mammalogy 2: 197244. 106. Riggs A. (1960). The nature and significance of the Bohr effect in mammalian hemoglobins. The Journal of General Physiology 43: 737752. 107. Roper T.J. and Kemenes I. (1997). Effect of blocking of entrances on the internal environment of badger Meles meles setts. Journal of Applied Ecology 34: 13111319 108. Roper T.J and Moore J.A.H. (2003). Ventilation of badger Meles meles setts. Mammalian Biology 68:277283. 109. Roper T.J., Bennett N.C., Conradt L. and Molteno A.J. (2001). Environmental conditions in burrows of two species of African molerat, Georhychus capensis and Cryptomys damarensis . Journal of Zoology 254: 101107.

110. Saiki C. and Mortola J. (1996). Effect of CO 2 on the metabolic and ventilatory responses to ambient temperature in conscious adult and newborn rats. Journal of Physiology 491: 261269.

References

111. Schaefer K.E. (1982). Effects of increased ambient CO 2 levels on human and animal health. Experientia 38: 11631168. 112. Scheid P. and Meyer M. (1978). Mixing technique for study of oxygen hemoglobin equilibrium: a critical evaluation. Journal of Applied Physiology 45: 818822. 113. SchmidtNielsen K. (2002). Animal physiology adaptation and environment. 5 th edition. Cambridge University Press. 114. SchmidtNielsen K. and Larimer J.L. (1958). Oxygen dissociation curves of mammalian blood in relation to body size. American Journal of Physiology 195: 424 428. 115. SchmidtNielsen K. and Pennycuick P. (1961). Capillary Density in Mammals in Relation to Body Size and Oxygen Consumption. American Journal of Physiology 200: 746750. 116. SchmidtNielsen, K. 1997. Animal Physiology. Cambridge University Press, Cambridge. 117. Sede Boqer Desert Meteorology Weather Station. http://bidr.bgu.ac.il/BIDR/research/phys/meteorology/index.html#top. 118. Shafran A., Ronen Z., Gross A., Weisbrod N. and Adar E. (2005). Effects of Surfactants Originating from Reuse of Grey water on Capillary Rise in the Soil. Water Science and Technology 52: 157166. 119. Shenbrot G. (2004). Habitat selection in a seasonally variable environment: test of the isodar theory with the fat sand rat, Psammomys obesus , in the Negev Desert Israel. Oikos 106: 359 365. 120. Shenbrot G., Krasnov B., Khokhlovaw I., Demidovaz T. and Fielden L. (2002). Habitatdependent differences in architecture and microclimate of the burrows of Sundevall’s jird ( Meriones crassus ) (Rodentia: Gerbillinae) in the Negev Desert, Israel. Journal of Arid Environments 51: 265279. 121. SiggaardAndersen O. (2005). Acid base balance. Encyclopedia of respiratory medicine. Laurent G.J. and Shapiro S.D (Eds.) Elsevier Ltd. Pp. 510. 122. Snyder G.K. (1976). Respiratory characteristics of whole blood and selected aspects of circulatory physiology in the common shortnosed fruit bat Cynopterus hrachyotis. Respiration Physiology 28: 239247. 123. Soholt L.F. (1974). Environmental conditions in an artificial burrow occupied by Merriam's Kangaroo rat, Dipodomys merriami. Journal of Mammalogy 55: 859864.

References

124. Soholt L.F., Yousef M.K. and Dill D.B. (1973). Responses of Merriam's kangaroo rats, Dipodomys merriami , to various levels of carbon dioxide concentration. Comparative Biochemistry and Physiology, A 45: 455–462. 125. Spruijt B., van Hooff J.and Gispen W. (1992). Ethology and neurobiology of grooming behavior. Physiological Reviews 72: 825. 126. Studier E.H. and Baca R.P. (1968). Atmospheric conditions in artificial rodent burrows. Southwestern Naturalist 13: 401410. 127. Studier E.H. and Proctor J.W. (1971). Respiratory gases in burrows of Spermophilus Tridecemlineatus . Journal of Mammalogy 52: 631633. 128. Stull R.B. (1988). An introduction to boundary layer meteorology. Kluwer Academic Publishers. Dordrecht, the Netherlands 129. Turner J. S. (2001). On the Mound of Macrotermes michaelseni as an organ of respiratory gas exchange. Physiological and Biochemical Zoology 74: 798822 130. Turner S.J. (2000). The extended organism; the physiology of animalbuilt structures. Harvard University Press. Cambridge, MI. 131. van Aardt W.J., Bronner G., Buffenstein R. (2007). Hemoglobinoxygenaffinity and acidbase properties of blood from the fossorial molerat, Cryptomys hottentotus pretoriae . Comparative Biochemistry and Physiology, A 147: 50–56. 132. van Erp A.M.M., Kruk M.R., Meelis W. and WillekensBramer D.l.C. (1994). Effect of environmental stressors on time course, variability and form of selfgrooming in the rat: handling, social contact, defeat, novelty, restraint and fur moistening. Behavioural Brain Research 65: 4755. 133. Villarreal D., Clark K.L., Branch L.C., Hierro J.L. and Machicote M. (2008). Alteration of ecosystem structure by a burrowing herbivore, the plains vizcacha (Lagostomus maximus ). Journal of Mammalogy 89: 700711. 134. Vogel S., Ellington C.P. and Kilgore D.L. (1973). Windinduced ventilation of the burrow of the prairiedog, Cynomys ludovicianus . Journal of Comparative Physiology 85: 114. 135. Vorhies C.T. and Taylor W.P. (1922). Life history of the kangaroo rat Dipodomys spectabilis spectabilis Merriam. U.S. Department of Agriculture Bulletin 1091: 140. 136. Walsberg G.E. (2000). Small mammals in hot deserts: some generalizations revisited. BioScience 50: 109 – 120.

References

137. Walsh J.P., Boggs D.F. and Kilgore D.L. Jr. (1996). Ventilatory and metabolic

responses of a bat, Phyllostomus discolor , to hypoxia and CO 2: implications for the allometry of respiratory control. Journal of Comparative Physiology, B 166: 351358. 138. Wei DB., Wei L., Zhang JM. and Yu HY. (2006). Bloodgas properties of plateau zokor ( Myospalax baileyi ) Comparative Biochemistry and Physiology, A 145: 372 375. 139. Westen E.A. and Prange H.D. (2003). Reexamination of the mechanisms underlying the arteriovenous chloride shift. Physiological and Biochemical Zoology, 76: 603614 140. White F.N., Bartholomew G.A. and Kinney J.L. (1978).Physiological and Ecological Correlates of Tunnel Nesting in the European BeeEater, Merops apiaster. Physiological Zoology 51: 140154. 141. Whitford W.G. and Kay F.R. (1999). Biopedturbation by mammals in deserts: a review. Journal of Arid Environments 41: 203230. 142. WhittingtonJones G.M. (2007). The role of aardvarks ( Orycteropus afer ) as ecosystem engineers in arid and semiarid landscapes of South Africa. M.Sc. thesis, Rhodes University, Grahamstown, South Africa. 143. Wilson K.J. and Kilgore D.L. (1978). The effects of location and design on the diffusion of respiratory gases in mammals burrows. Journal of Theoretical Biology 71: 73101. 144. Withers P.C. (1975). A comparison of respiratory adaptations of a semifossorial and a surfacedwelling Australian rodent. Journal of Comparative Physiology 98: 193 203. 145. Withers P.C. (1978). Models of diffusionmediated gas exchange in animal burrows. The American Naturalist 112: 11011112. 146. Ziemann A.E., Allen J.E., Dahdaleh N.S., Drebot I.I., Coryell M.W., Wunsch A.M., Lynch C.M., Faraci F.M., Howard M.A. 3rd, Welsh M.J. and Wemmie J.A. (2009). The amygdala is a chemosensor that detects carbon dioxide and acidosis to elicit fear behavior. Cell 139: 10121021. 147. ZuckerMilwerger D. (2013). Gas exchange in rodent burrows: the relative contributions of different mechanisms and the underlying physical processes of mass flow. M.Sc. thesis, The Albert Katz International School for Desert Studies, Ben Gurion University of the Negev.

References

148. Zuur A.F. Ieno E.N., Walker N.J., Saveliev A.A. and Smith G.M. (2009). Mixed effects models and extensions in ecology with R. Springer Science + Business Media, LLC.

Appendix

Table 16: the blood gas properties of small mammal species. 0.054 Predicted P50(7.4) = 53.34 mb , the PO2 (mmHg) at 50% O 2 saturation of blood at pH 7.4 at 37 °C (SchmidtNielsen and Larimer, 1958)

Bohr buffer Predicted Source P50 effect capacity P50 Common name Scientific name mb (g) mmHg mmHg (kPa) ∆log ∆log CO ∆[HCO (kPa) ∆pH ∆pH ∆pH Sundevall's Meriones jirds crassus

Eastern Tamias striatus 70 2.04 Maclean, 1981 chipmunk

16.0 38.5 European mole Talpa europaea 140 0.47 Bartels et al., 1969 (2.13) (5.13)

Heterocephalus 23.3 41.5 Naked mole rat 105 0.43 Johansen et al., 1976 glaber (3.11) (5.53)

Middle East Spalax 29.5 37.9 blind mole rat 196 0.53 1.32 Ar et al., 1977 ehrenbergi (3.93) (5.05)

California Spermophilus 26.0 35.6 598 Baudinette, 1974 ground squirrel beecheyi (3.47) (4.75)

Appendix

Bohr buffer Predicted Source P50 effect capacity P50 Common name Scientific name mb (g) mmHg mmHg (kPa) ∆log ∆log CO ∆[HCO (kPa) ∆pH ∆pH ∆pH Blacktailed Cynomys 22.0 34.1 Hall, 1966 1280 prairie dog ludovicianus (2.93) (4.55)

Whitefooted Peromyscus 33.2 20 45.4 (6.05) mouse leucopus (4.43)

Peromyscus 36.7 Golden mouse 23 45.0 (6.00) nuttalli (4.89)

Reitlirodontomys 32.2 Harvest mouse 12 46.6 (6.22) humulus (4.29)

Pitymys 34.0 Forman, 1954 Pine mouse 30 44.4 (5.92) pinetorum (4.53)

Microitus 33.8 Meadow mouse 39 43.8 (5.84) pennsylvanicus (4.51)

Sigmodon 33.0 Cotton rat 120 41.2 (5.49) hispidus (4.40)

Mesocricetus 22.0 41.9 Golden hamster 88 auratus (2.93) (5.58)

Appendix

Bohr buffer Predicted Source P50 effect capacity P50 Common name Scientific name mb (g) mmHg mmHg (kPa) ∆log ∆log CO ∆[HCO (kPa) ∆pH ∆pH ∆pH 34.2 45.6 House mouse Mus musculus 18 (4.56) (6.08)

34.2 44.5 White mouse Mus musculus 29 (4.56) (5.93)

Rattus 35.5 38.0 Norway rat 540 norvegicus (4.73) (5.06)

Rattus 32.4 38.7 Forman, 1954 White rat 387 norvegicus (4.32) (5.15)

Sciurus 26.2 38.5 Gray squirrel 425 carolinensus (3.49) (5.13)

Glaucomys 30.1 42.6 Flying squirrel 65 volans (4.01) (5.68)

Shortbeaked Tachyglossus 21.3 33.7 1680 0.49 Parer and Metcalfe, 1967 echidna aculeatus (2.84) (4.49)

Egyptian fruit Rousettus 29.5 ± 0.5 Giardina et al., 1990 bat aegyptiacus (3.93)

Appendix

Bohr buffer Predicted Source P50 effect capacity P50 Common name Scientific name mb (g) mmHg mmHg (kPa) ∆log ∆log CO ∆[HCO (kPa) ∆pH ∆pH ∆pH

Peromyscus 50.1 45.3 Deer mouse 21 maniculatus (6.68) (6.03)

49.2 45.1 White mouse Mus musculus 22.5 (6.56) (6.01)

Sigmodon 39.5 42.3 Cotton rat 72.5 hispidus (5.26) (5.64) SchmidtNielsen and Mesocricetus 28.6 41.1 Larimer, 1958 Golden hamster 127 auratus (3.81) (5.47) Data extracted from figure 2

Rattus 32.4 39.4 White rat 272 norvegicus (4.32) (5.25)

34.0 36.8 Guinea pig Cavia porcellus 950 (4.53) (4.91)

Didelphis 43.6 36.8 Opossum 950 virginiana (5.81) (4.91)

100

Appendix

Bohr buffer Predicted Source P50 effect capacity P50 Common name Scientific name mb (g) mmHg mmHg (kPa) ∆log ∆log CO ∆[HCO (kPa) ∆pH ∆pH ∆pH Oryctolagus 30.8 35.2 Rabbit 2200 cuniculus (4.11) (4.69)

Felis silvestris 39.4 35.2 Cat 2250 catus (5.25) (4.69)

Canis lupus 31.3 32.3 Dog 10750 familiaris (4.17) (4.31)

27.3 31.1 Sheep Ovis aries 21800 (3.64) (4.15) SchmidtNielsen and Capra aegagrus 29.4 30.5 Larimer, 1958 Goat 31240 hircus (3.91) (4.07) Data extracted from figure 2

24.9 28.8 Man Homo sapiens 92500 (3.31) (3.84)

Sus scrofa 32.7 28.0 Pig 153000 domesticus (4.36) (3.73)

Bos primiigenius 25.4 26.3 Cow 490000 taurus (3.38) (3.51)

101

Appendix

Bohr buffer Predicted Source P50 effect capacity P50 Common name Scientific name mb (g) mmHg mmHg (kPa) ∆log ∆log CO ∆[HCO (kPa) ∆pH ∆pH ∆pH Equus ferus 24.3 26.1 Horse 580000 caballus (3.24) (3.47)

Thomomys Pocket gopher 2.67 bottae Chapman and Bennett, 1975 Rattus Laboratory Rats 1.39 norvegicus

Ashgrey Pseudomys 56.6 43.1 18.2 ± 0.5 0.00129* mouse albocinereus (7.55) (5.75)

Pseudomys 66.5 40.5 Heath mouse 56.7 ± 2.4 0.00075* shortidgei (8.87) (5.40) Withers, 1975 Egyptian fruit Rousettus 30.8 ± 0.4 40.8 0.55 ± *Bohr effect is expressed as 146 ± 7.5 bat aegyptiacus (4.11) (5.43) 0.004 ∆ ∆CO 2 Phyllostomus Pale spear 28.6 ± 0.5 43.4 discolor 45.2 ± 1.3 0.55 ± 0.04 nosed bat (3.81) (5.97) (omnivorous)

Black mastiff Molossus ater 31.5 ± 0.6 43.8 38.2 ± 1.4 0.55 ± 0.04 bat (insectivorous) (4.20) (5.84)

102

Appendix

Bohr buffer Predicted Source P50 effect capacity P50 Common name Scientific name mb (g) mmHg mmHg (kPa) ∆log ∆log CO ∆[HCO (kPa) ∆pH ∆pH ∆pH Mouse eared 33.3 ± 1.9 45.3 Myotis myotis 20.6 ± 0.9 0.55 ± 0.04 bat (4.44) (6.04)

Common Pipistrellus 36.6 ± 1.2 49.0 4.85 ± 0.2 0.47 ± 0.03 pipistrelle pipistrellus (4.88) (6.53)

Greater white Crocidura 33.8 ± 1.1 47.4 Jürgens et al.,1981 8.95 ± 0.5 0.66 ± 0.08 toothed shrew russula (4.51) (6.32)

35.2 ± 0.9 50.8 Etruscan shrew Suncus etruscus 2.45 ± 0.1 0.61 ± 0.05 (4.69) (6.78)

Apodemus 33.5 ± 1.3 44.9 Wood mouse 24.8 ± 2.1 0.60 ± 0.05 sylvaticus (4.47) (5.98)

Clethrionomys 29.8 ± 1.4 33.2 ± 0.9 44.4 0.60 ± 0.05 Bank vole glareolus (4.43) (5.92)

Psammomys Fat sand rat 2.32 ± 0.35 68.8 ± 18.8 obesus BarIlan et al., 1985 Rattus Rat 1.51 ± 0.01 27.3 ± 6.4 norvegicus

103

תקציר בעלי חיים ממגוון רחב של טקסונים, מנמלים ועד שנבובים, חיים במחילות מתחת לקרקע. מחילה מספקת הגנה מפני טורפים ומפני תנאי מזג אוויר קיצוניים, וגם מקום בטוח לגדל בו צאצאים ולאגור מזון. בספרות מקובל שמבנה המחילה בשילוב החדירות הנמוכה כביכול של קרקעות למעבר גזים , ונשימתו של בעל החיים עשויים לגרום לכך

שבמחילות החלק היחסי של פחמן דו חמצני (FCO 2 ) ושל חמצן (FCO 2 ) יהיו, בהתאמה,

מעל ומתחת לאלה שבאטמוספרה. אצל יונקים, שאיפה מתמשכת של אוויר עם FCO 2 גבוה עלולה לגרום למצב של חמצת נשימתית ( respiratory acidosis ). לפיכך, לבעלי חיים

המבלים את רוב חייהם במחילות יש תכונות פיזיולוגיות המאפשרות להם לשאת FCO 2

גבוה באוויר אותו הם נושמים, או לחילופין, מחילותיהם בנויות כך שהצטברות CO 2 נמנעת על ידי אוורור נאות. בעבודה זו חקרתי היבטים אקולוגים ופיזיולוגים בחייו של מכרסם מתחפר- למחצה ( semifossorial ), מריון המדבר, Meriones crassu s Sundevall, 1842 . חלקה הראשון של העבודה הינו מחקר של הסביבה הנשימתית במחילה של מריון, כפי שהיא נחוות על ידי בעל החיים בסביבתו המיידית, והגורמים המשפיעים עליה. ראשית, אפיינתי את המנגנונים השונים לאוורור של מערכת מחי לות פתוחה והצעתי הסבר חדש לאופן הפעולה של מנגנון האוורור על ידי רוח. מצאתי שאוורור המחילה נעשה על ידי מספר מנגנונים שחשיבותם משתנה אפילו במהלך שעות היממה כתלות בתנאי הסביבה, ובעיקר כתלות במהירות הרוח. נראה שבמהירות רוח גבוהה מ -2 \מ ש המחילה מאווררת היטב, זא ת על ידי פרצי אוויר ( eddies ) הנוצרים כתוצאה מזרימה טורבולנטית בשכבת הגבול באטמוספרה וחודרים באופן לא סדיר לעומקים שונים בתעלה הראשית של המחילה. על אף שפרצי האוויר אינם חודרים באופן ישיר אל אזורים במחילה המתפצלים מהתעלה הראשית (כמו למשל תא הרביה), אני מציעה שגם באזורים אלה יש אוורור נאות בזכות מפל ריכוזים גדול וכמעט קבוע הנשמר בינם לבין התעלה הראשית. מפל הריכוזים מאפשר

דיפוזיה של CO 2 אל התעלה הראשית שם הוא מסולק על ידי פרץ האוויר הבא. גם במהירות רוח נמוכה מ -2 \מ ש נמדדו פרצי אוויר בתעלה הראשית. בנוסף, במהירות רוח נמוכה דיפוזיה ( diffusion) של גז באוויר העומד במחילה ודרך הקרקע עשויה להיות מנגנון חשוב לאוורור המחילה. במקרה זה, למידת הנקבוביות של הקרקע תרומה מכרעת לאוורור של חללים ללא מוצא ה נמצאים בעומק המחילה רחוק מן הפתחים.

בהמשך בדקתי איך ריכוז הפחמן הדו חמצני [CO 2] במקום מסוים בתוך המחילה מושפע ממיקומו של בעל החיים. נראה שהמיקום המדויק של חרטום החיה (המקור

המידי של CO 2 ) ביחס לחיישן המודד משפיע על הריכוז הנמדד ועל קצב העלייה\ ירידה של

[CO 2] . מרבית המחקרים העוסקים בסביבה הנשימתית במחילות של יונקים אינם

כוללים מידע לגבי המיקום של בעל החיים בתוך המחילה בזמן דגימת הגזים בה. זאת, קרוב לודאי, בגלל הקשיים הטכניים הכרוכים בחדירה לא הרסנית לעומק המחילה ( Roper et al., 2001 ). תוצאות המחקר הנוכחי, מצביעות על כך שמידע לגבי מיקום בעל החיים הינו חיוני מאחר וריכוז הגזים הנשימ תיים בקרבת בעל החיים עשוי להיות שונה בתכלית מזה שבשאר חלקי המחילה.

לבסוף, בכדי לגלות מהן רמות הCO 2- איתן מתמודדים מריונים במחילותיהם באופן רגיל, הכנסתי מריונות בהריון לתוך מחילות מלאכותיות, ולאחר ההמלטה מדדתי את

[CO 2] והטמפרטורה במספר מקומות לאורך המחילה ואת מהירות הרוח והטמפרטורה מחוץ למחילה במשך שלושה שבועות בהם הנקבות הניקו את הגורים. תוצאותיי מצביעות

על כך שרמת הCO 2- במחילות של מריונים אינה עולה על 2-2.5% . אחוזים אלה אינם נחשבים מסוכנים מבחינה פיזיולוגית עבור מכרסמים. באחת מהמחילות המלאכותיות, שהייתה ל -א חדירה למעבר גזים דרך דופנותיה, ובה חילוף הגזים עם הסביבה נעשה אך

ורק דרך שתי הכניסות, נמדד [CO 2 ] גבוהה מ7%- . יש לציין, שנראה שהנקבה והגורים במחילה זו ניהלו אורח חיים תקין לאורך שלושת שבועות המדידה, ושחמשת הגורים היו במצב טוב כשנלכדו בסוף תקופת המדידה. בחלקו השני של המחקר הראיתי שכאשר ניתנת למריונים בחירה בין שתי סביבות

מחיה, האחת עם CO 2 גבוה (7% ) והשנייה עם אוויר ללא CO 2 (0% ) הם אינם מעדיפים סביבה אחת על פני האחרת. בעקבות ממצאים אלה, המשכתי ובדקתי מספר תכונות

פיזיולוגיות שעשויות לתרום לעמידות של מריונים בסביבה נשימתית עם FCO 2 גבוה ו\ או

FO2 נמוך. ראשית בדקתי את יכולת נשיאת החמצן בדם של מריונים; מצאתי שלחץ

החלקי של חמצן בדם שבו מחצית מההמוגלובין רווי ( P50 ) דומה, או קצת גבוה יותר מהמצופה מבעל חיים בעל גודל גוף של מריון לפי המשוואה האלומטרית (Schmidt Nielsen and Larimer, 1958 ) ושאפקט בוהר ( Bohr effect ) של מריונים דומה לזה של יונקים אחרים. לפיכך, נראה שלהמוגלובין של מריונים אין משיכה מוגברת או יכולת פריקה מוגברת של חמצן ברקמות. המשכתי ובדקתי את קיבול הבאפר ( buffering + capacity ), ריכוז יוני המימן [ H] , ריכוז הבי קרבונט [ HCO 3] והלחץ החלקי של פחמן דו

חמצני (PCO 2 ) בדם של מריונים והשוואתי אותם עם אלה של הפסמון Psammomys obesus , מכרסם מדברי בעל גודל גוף דומה לזה של מריון וקרוב לו מבחינה טקסונומית, שגם הוא עלול להיחשף לריכוז גבוה של בסביבתו הטבעית במחילה. מצאתי שלמריונ ים יש קיבול באפר דומה לזה של יונקים שאינם חיים במחילות, אך הוא נמוך באופן משמעותי מזה של הפסמון ומזה של מינים מתחפרים אחרים. לפיכך, הסקתי שלמריונים אין קיבולת באפר מוגברת בדם.

במספר מינים של יונקים מתחפרים ( fossorial ) ומתחפרים למחצה ( semifossorial )

נמדדה תגובה נשימתית מופחתת כתוצאה משאיפת אוויר עם FCO 2 גבוה. בכדי לבדוק

את התגובה הנשימתית ( respiratory response ) של מריונים השואפים אוויר עם FCO 2 גבוה מדדתי את העלייה בתדירות הנשימה ( respiration frequency ) של מריונים כתוצאה

מנשימת אוויר עם CO 2 = 7% לעומ ת אוויר ללא CO 2 . בנוסף, השוואתי את מידת הפעילות, הקצב המטבולי במנוחה ( resting metabolic rate RMR ) והקצב המטבולי

הממוצע ( mean metabolic rate ) של מריונים הנושמים אוויר ללא CO 2 לאלה של

מריונים הנושמים אוויר עם CO 2 = 7% . העלייה בתדירות הנשימה של המריונים אינ ה

מצביעה על תגובה נשימתית מופחתת בשאיפת אוויר עם FCO 2 גבוה. בנוסף, כאשר שהו

המריונים בסביבה עם FCO 2 גבוה הראו המריונים סימנים של חוסר מנוחה, הם היו

פעילים יותר וישנו פחות ממריונים בסביבה עם אוויר ללא CO 2 וכתוצאה מכך הקצב המטבולי הממוצע שלהם עלה. גם ה - RMR של המריונים היה גבוה באופן מובהק כששהו

בסביבה עם FCO 2 גבוה לעומת ה - RMR שלהם בסביבה עם אוויר ללא CO 2 . ממצאים

אלה מצביעים על כך שבכדי לשהות בסביבה עם CO 2 על מריונים לשלם מחיר אנרגטי. לבסוף, מדדתי את השינוי בקצב המטבולי של מריונים בזמן שינויים קטנים

והדרגתיים בCO 2- (מCO 2=0 -ל CO 2=1%, 2%, 4%, 7% ושוב ל- CO 2=0 ) ובזמן

שינויים גדולים וחדים ב - CO 2 (מCO 2=0 -ל CO 2=7% ושוב ל- CO 2=0 ) באוויר אותו הם

נושמים. לשינויים בFCO 2- לא הייתה השפעה על הקצב המטבולי של המריונים, גם כאשר השינויים היו קטנים והדרגתיים וגם כאש ר היו גדולים וחדים.

על אף שמריונים הינם עמידים מאוד בפני FCO 2 גבוה באוויר אותו הם נושמים לאורך זמן, לא מצאתי תכונות מיוחדות המצביעות על הסתגלות פיזיולוגית

( physiological adaptation ) ייחודית לסביבת מחיה עם CO 2 גבוה. מכך אני מסיקה

שבסביבתם הטבעית במחילות מרי ונים אינם נאלצים להתמודד עם FCO 2 גבוה, זאת משום שמבנה המחילות שלהם, והאוורור שלהן על ידי חדירת פרצי אוויר, מונע הצטברות

של CO 2 . .

מילות מפתח: מריו מדבר, יונק מתחפר, מחילה , איוורור מחילות, אפקט בוהר, P50 , קיבולת בו פר של ד, קצב מטבולי

הצהרת תלמיד ת המחקר עם הגשת עבודת הדוקטור לשיפוט

אני , ענבל בריקנר- בראון, החתו מה מטה מצהירה בזאת:

X חיברתי את חיבורי בעצמי, להוציא עזרת ההדרכה שקיבלתי מאת מנחה/ים.

X החומר המדעי הנכלל בעבודה זו הינו פרי מחקרי מתקופת היותי תלמיד/ת מחקר.

__ בעבודה נכלל חומר מחקרי שהוא פרי שיתוף עם אחרים, למעט עזרה טכנית הנהוגה בעבודה ניסיונית. לפי כך מצורפת בזאת הצהרה על תרומתי ותרומת שותפי למחקר, שאושרה על ידם ומוגשת בהסכמתם.

תאריך : 3.7.2013 שם התלמידה: ענבל בריקנר- בראון

חתימה ______

העבודה נעשתה בהדרכת פרופ' ברי פינשאו ופרופ' פדרו ברלינר

המחלקה לאקולוגיה מדברית ע"ש מיטרני והמחלקה לחקלאות באזורים צחיחים ע"ש ווילר, המכונים לחקר המדבר ע"ש יעקב בלאושטיין

יחסי הגומלין בין פיזיולוגיה של נשימה של מכרסם במחילה לבין משטר הא וורור של המחילה: פרק בפיזיולוגיה האקולוגית של מריון המדבר ( Meriones crassus)

מחקר לשם מילוי חלקי של הדרישות לקבלת תואר "דוקטור לפילוסופיה"

מאת

ענבל בריקנר- בראון

הוגש לסינאט אוניברסיטת בן גוריון בנגב

אישור המנחים:

______

אישור דיקן בית הספר ללימודי מחקר מתקדמים ע"ש קרייטמן

______

כ"ה תמוז תשע"ג 3 יולי, 2013

באר שבע

יחסי הגומלין בין פיזיולוגיה של נשימה של מכרסם במחילה לבין משטר האוורור של המחילה: פרק בפיזיולוגיה האקולוגית של מריון המדבר ( Meriones crassus)

מחקר לשם מילוי חלקי של הדרישות לקבלת תואר "דוקטור לפילוסופיה"

מאת

ענבל בריקנר- בראון

הוגש לסינאט אוניברסיטת בן גוריון בנגב

כ"ה תמוז תשע"ג 3 יולי, 2013

באר שבע