Great Basin Naturalist Memoirs

Volume 7 of Article 4

8-1-1983

Morphological structure and function in the desert heteromyid rodents

Joyce C. Nikolai Department of Biology, University of Utah, Salt Lake City 84112

Dennis M. Bramble Department of Biology, University of Utah, Salt Lake City 84112

Follow this and additional works at: https://scholarsarchive.byu.edu/gbnm

Recommended Citation Nikolai, Joyce C. and Bramble, Dennis M. (1983) "Morphological structure and function in the desert heteromyid rodents," Great Basin Naturalist Memoirs: Vol. 7 , Article 4. Available at: https://scholarsarchive.byu.edu/gbnm/vol7/iss1/4

This Article is brought to you for free and open access by the Western North American Naturalist Publications at BYU ScholarsArchive. It has been accepted for inclusion in Great Basin Naturalist Memoirs by an authorized editor of BYU ScholarsArchive. For more information, please contact [email protected], [email protected]. MORPHOLOGICAL STRUCTURE AND FUNCTION IN DESERT HETEROMYID RODENTS'

Joyce C. Nikolai- and Dennis M. Bramhle-

.\bstract.— The functional morphology of desert heteromyid rodents (, Dipodomys, Microdipodops) is reviewed with considerable new information provided. Specific attention is given to the interaction of anatomical structure and the behavioral and ecological patterns of these rodents. Inflation of the auditory bullae, although apparently related to improved hearing, is also shown to directly impact the structure and function of the feeding apparatus in desert heteromyids. The mechanics of high speed pouch- ing behavior in Dipodomys deserti as well as the rates of digging activities of various heteromyids are described using data from slow motion films. The biomechanical consequences of cheek pouch loading for body size and locomotor behavior yield theoretical predictive models concerning interspecific differences in foraging behavior, dietary prefer- ence, and microhabitat selection. Stnictura! modifications of the forelimb associated with use of external cheek pouches reduce the mechanical competence of these limbs for shock absorption during fast quadrupedal running. The relative size of various front and hind limb segments are correlated with quadrupedal, tripodal, and bipedal gaits in heteromyid rodents. The in- terdependence of body balance, gait, and speed are examined in Dipodomys merriami. Factors possibly contributing to the origin of in rodents are reviewed and discussed.

The heteromyid rodents of (1932) comparative study of vertebral archi- offer, potentially, a superb opportunity to ex- tecture in saltatorial rodents, and Wood's amine the importance of morphological de- (1935) important survey of the fossil and Re- sign as a determinant of behavioral and eco- cent . Herman's (1979) recent logical patterns under natural conditions. multivariate statistical analysis of hind limb This follows from the considerable range of bone and muscle morphology in bipedal ro- morphologies found within this circum- dents constitutes a very significant extension scribed group as well as the impressive beyond the older comparative anatomical breadth of habitats that its living representa- works. tives presently occupy. Perhaps more signifi- Recent functional morphologic studies cant is the fact that heteromyids have recent- have been more analytical and experimental, ly become the focus of numerous but also of more limited scope. Thus, Pink- investigations aimed at gaining a better un- ham (1976) has investigated the gaits and me- derstanding of their biology on multiple lev- chanics of quadrupedal and bipedal running els (e.g., physiology, behavior, ecology, com- in Liomys and Dipodomys by combining high mimity level interactions). Such studies have speed cinematography with force platform begun to offer the kinds of information recordings. Using similar techniques, but also against which carefully framed hypotheses of including cineradiography, Biewener et al. a hmctional-morphologic nature might be (1981) have studied the mechanical behavior critically appraised. of the major hindlimb tendons in Earlier morphologic studies on hetero- . Additional information on the ankle me- myids are primarily descriptive but also con- chanics of Dipodomys and the physiological tain comments on form-function relationships properties of its associated mu.sculature has that are, of necessity, relatively superficial been presented by Williamson and Frederick and speculative. Excellent works of this type (1977). Kaup (1975) has also commented on are Howell's (1932) monograph on the myol- the biomechanical and evolutionary signifi- ogy and osteology of Dipodotnys, Hatt's cance of hind limb anatomy in heteromyids.

'From the .symposium "Biology of Desert -s," presented at the annual meeting of the .\nierican Society of Mammalogists, hosted by Brigham Young University, 20-24 June 1982, at Snowbird, Utah. 'Department of Biology, University of Utah, Salt Lake City. Utah 84112.

44 1983 Biology of Desert Rodents 45

Of special note are the investigations of the Websters (see below) on the auditory appa- ratus of heteromyids. Not only have their studies provided a large body of comparative morphologic data, but they also offer one of the few examples wherein specific hypoth- eses concerning the adaptive value of a major morphological complex have been tested. In the present paper we briefly review the existing data on the functional morphology of heteromyid rodents and point out significant gaps in our knowledge. Considerable atten- tion is given the forelimbs and cheek pouches, two structures that have received httle attention in the past but whose struc- tural organization may place important con- straints on the behavior of these . Here and elsewhere we have tried to show how functional morphologic analyses can lead to predictive, testable models con- cerning the natural behavior and ecology of heteromyid rodents. Limitations of both time and materials have forced us to restrict the present dis- cussion to "desert heteromyids" of the genera Perognathiis, Dipodoniys, and Microdipodops.

This is done with full knowledge that a better understanding of the fimctional anatomy and behavior of the modern heteromyines (Liomys, ) would imdoubtedly broaden our appreciation of form-fimction and evolutionary patterns within pe- rognathine and dipodomyine heteromyids and might well alter some of our conclusions.

Skull and Neck Morphology Fig. 1. Influence of auditory specialization on the feeding apparatus of desert heteromyids. A generahzed desert rodent, Neotoma lepida (A) is compared to Pe- (B), Dipodomys merriami (C), and Skull rognathtts forinosus Microdipodops megacephalus (D). Inflation of the audi- tory bufla (stippled) reduces the area of origin of the The most striking cranial feature of desert temporalis musculature (hatched) and also restricts gape heteromyids and that which has received the by crowding the mandible from behind. The specialized most attention is the enlarged middle ear everted angle (ea) of the mandible reduces the impact of chambers or auditory bullae. The auditory bullar inflation in heteromyids. Maximum gape between cheek teeth (but not incisors) in Perognathiis (35°) is bullae are moderately inflated in Pe- the about equal to that in Neotoma (36°), but extreme rognathiis, but grossly so in both Dipodoniys middle ear hypertrophy has severely reduced gape in and Microdipodops (Fig. 1). Relative to over- Dipodomys and Microdipodops. All skulls drawn to same all head size, the middle ear chambers length. achieve their greatest volume in the latter (Webster 1961). Detailed comparative The innovative studies of the Websters and morphologic data on the auditory region of their collaborators have gradually revealed heteromyids have been provided by Webster the functional and probable adaptive signifi- and Webster (1975, 1977, 1980). cance of the modified ears of desert hetero- 46 Great Basin Naturalist Memoirs No. 7 myids. Auditory specializations in these ro- presumed that mortality was due chiefly to dents, and in certain Old World desert spe- higher rates of upon kangaroo rats cies (Lay 1972), improve the detection of rel- whose ability to detect low frequency sound atively low frequency sound, especially in the had been reduced. Laboratory recordings in- 1-3 KHz range. Selective sensitivity to these dicate that the predatory strikes of both frequencies has been established on the basis and produce significant sound in of physiological (Ruppert and Moushegian the 1-3 KHz range (Webster 1962). In sum, 1970, Vernon et al. 1971, Webster and Stro- the available data strongly imply that the ther 1972, Webster and Webster 1972) and specialized auditory apparatus of desert het- experiments (Webster and Web- behavioral eromyids is indeed adaptive, and that it may suite of structural features ap- ster 1972). A confer its greatest advantage on individuals pear to be responsible for increased sensi- foraging under conditions of dim tivity to low frequency sound by lowering illumination. impedance and, hence, increasing the trans- There has been little functional analysis of mission of such sound from the external to in- the feeding mechanism of heteromyid ro- ner ear. Among these features are: a rela- (1) dents. Most studies have been concerned with tively large, compliant tympanic membrane; dental morphology as it relates to systematics a small, low-mass, high-leverage ossicular (2) and the identification of fossil materials chain; and an enlarged middle ear cham- (3) (Lindsay 1972, Shotwell 1967, Wood 1935). ber. The latter feature is apparently a com- The cheek teeth of modern desert hetero- pensatory adjustment that reduces middle ear myids are relatively simple and lophodont. In damping of the large ear drum (Legouix et al. Dipodomys the cheek teeth are hypsodont 1954, Webster 1962, Wisner et al. 1954). Ex- and rootless. Enamel is confined to the ante perimental reduction of middle ear volume in rior and posterior faces, a condition paral- kangaroo rats significantly reduces sensitivity leled in the Geomyidae (Wood 1937). The to low frequency sound (Webster 1961, Web- mandible tends to be small (relative to head ster and Webster 1972, 1980). Certain struc- size) in all living heteromyids, but is marked- tural modifications of the inner ear (Webster ly so in Dipodomys and Microdipodops. The 1961, Webster and Stack 1968) and related smallness of the mandible causes it to be areas of the brain (Webster et al. 1968) may rather severely underslung. This position, to- also reflect selective sensitivity to low fre- gether with the dorsal location of the eyes quency sound in Dipodomijs. (Howell 1932), assures that the movements of How the specialized ears of kangaroo rats the hands during feeding and pouching be- might contribute to individual fitness has also havior are kept well below eye level. been examined. Captive kangaroo rats (D. Inflation of the middle ear has impinged merrkimi) were tested to see how effectively and by what means they avoided the attacks directly on the masticatory apparatus by (1) of owls and rattlesnakes (Webster 1962, forcing a reduction of the temporalis mus- Webster and Webster 1971). Animals with culature and (2) crowding the mandible from unimpaired hearing were usually able to behind, thereby placing serious limitations on avoid capture by these predators even in to- gape. The latter problem has been partially tal darkness or when blinded. Those with im- circumvented by the development of an paired hearing (i.e., artificially reduced everted angular process. Reorientation of the middle ear volumes) could also avoid attack, angular process delays its contact with the but only when there was sufficient light to undersurface of the bulla as the jaw is opened see the movements of the predator. Blind (depressed). This permits a wider gape than kangaroo rats with impaired hearing could would otherwise be possible. Even so, middle not escape predation. A comparison of mor- ear inflation restricts gape in all desert heter- tality rates between Dipodomys with normal omyids, but most especially in Dipodomys and impaired hearing under field conditions and Microdipodops (Fig. 1). A restriction in suggests higher mortality among impaired gape will limit the size of resistant food items animals, especially during dark of the moon that an can effectively gnaw. Exactly intervals (Webster and Webster 1971). It is how restructuring of the posterior region of 1983 Biology of Desert Rodents 47

the lower jaw has influenced other biome- Reorganization of the cervical region ap-

chanically important descriptors of mastica- pears to accomplish two functions. First, it tory fimction (e.g., force, rate, and direction helps to foreshorten the anterior trunk, which of mandibular movements; organization of tends to keep the distribution of body mass

adductor musculature) is unknown at this rearward. This eases the problem of counter- time. balancing the body over the hind limbs when Another prominent feature of the cranium in the bipedal pose. Secondly, modiflcations

of desert heteromyids is the strongly pro- of the neck increase its mechanical strength jecting, tubular nasal region. TTie nasal pas- and stability while the animal is involved in

sage is occupied by closely spaced turbinate bipedal hopping. Hatt (1932) argued that bones. The length of the nasal passage as well neck specialization was required to reduce

as the diameter of its individual air channels bobbing of the head. This idea has been ac- are presumably important to the water con- cepted by many subsequent workers, but has serving, counter-current heat exchanger pos- never been experimentally verified. Hatt sessed by Dipodcnnijs (Jackson and Schmidt- himself offered no functional analysis in sup- Nielsen 1964, Schmidt-Nielsen et al. 1970). port of his model. The relative development of the nasal region in desert heteromyids as regards their ability Cheek Pouches to detect subsurface accumulations of (Frye and Rosenzweig 1980, Reichman 1979) Structure, Use and Significance merits examination. External, fur-lined cheek pouches are a The interior of the skull of desert hetero- unique, derived feature of geomyoid rodents. myids exhibits at least one obvious special- They are not present at birth, but rapidly de- ization. Well-formed bony partitions project velop during the early postnatal period from medially from the otic capsules into the infoldings of the facial skin (Lackey 1967). In space between the cerebral and cerebellar the adult, each pouch opens externally via a lobes of the brain. They are pronounced in long slitlike aperture. Internally, the pouch Dipodomys and Microdipodops, but more continues rearward to an expanded base that modestly developed in Perognathus. These rests over the shoulder blades. Geomyids can partitions, which tend to compartmentalize voluntarily evert the pouches for cleaning the brain within the cranium, appear to be (Vaughan 1966) and perhaps in some cases to true tentorial ossifications. As such they can- help empty their contents. In both geomyids not be directly related to inflation of the and heteromyids superficial facial muscula- middle ear. Whether or not such structures ture is used to control the tension in the skin have any fimctional connection with the rap- guarding the entrance to the pouch and a id accelerations of tlie head and brain expe- special "pouch muscle," derived (in part) rienced by bipedal heteromyids invites from the trapezius complex, returns the investigation. everted pouch to its normal position (Chias- son 1954, Hill 1937, Howell 1932). of cheek Neck Two fairly obvious advantages pouches are (1) reducing the time required to Hatt (1932) described vertebral modifica- gather food on the surface, hence reducing

tions that seem to be associated with bipedal exposure to predators and, (2) reducing the saltation in rodents. Among them are: (1) ex- locomotor energy expended in foraging, by treme shortening and compaction of the cer- allowing an animal to collect and store a = with fewer trips. The vical region, (2) pronounced dorsiflexion ( given amount of food where hyperextension) of the neck, and (3) partial or latter may be especially important complete fusion of the anterior (excluding food resources tend to be widely scattered atlas) neck vertebrae. These specializations (Reichman and Oberstein 1977). Another pos- are common to both Old World (, sible advantage of external cheek pouches to ) and New World (Heteromyidae) desert heteromyids is that of water con- bipeds, but are most pronounced in Jaculus, servation. Unlike internal cheek pouches (in- Dipus, and Dipodarnys (Hatt 1932). dependently evolved in many mammalian 48 Great Basin Naturalist Memoirs No. 7

f 26 f 34

Fig. 2. Seed pouching in Dipodomys deserti. Tracings of representative frames of slow motion film (200 fps) of D. (U-serti illustrate one complete pouching cycle. Millet seed and the kangaroo were placed on a glass surface and filmed from below u.sing a mirror. See text for details. groups; Murray 1975), the fur-lined pouches only for the time required by Pewgnathus of heteromyids effectively isolate dry food and Dipodomys to husk relatively large seeds materials from the moist mucous membranes (Rosenzweig and Sterner 1970). Some insight of the oral chamber. This prevents ab.sorption into the much more rapid process of pouch- of water by the food—water that would be ing has been gained recently from high speed lo.st to the environment when the food was films made of an adult female D. deserti col- later cached in the ground. Given the critical lecting unhusked millet and sunflower seed problems of water balance faced by desert from a gla.ss plate. heteromyids (see MacMillen, this volume), High speed pouching in D. deserti is highly the savings potentially attributable to the u.se stereotyped. Figure 2 illustrates selected of external cheek pouches may be significant. stages in a typical pouching cycle. Both fore- Previous workers have noted tlie speed limbs move in synchrony and each limb with which desert heteromyids are able to serves only the ipsilateral pouch. At the in- collect and pouch seeds. Nonetheless, the itiation of the cycle the limbs are thrust for- speed of food handling has been quantified ward and downward toward the seeds as the 1983 Biology of Desert Rodents 49 hands are simultaneously pronated and parable data are not yet available for Pe- opened (Fig. 2: F1-F18). The hands are next rognathus and Microdipodops. closed on the seeds and then retracted to- Our films also hint at the mechanism by ward the mouth (Fig. 2: F18-F26). During which kangaroo rats distinguish between the retraction stage the hands are supinated edible and inedible items during high speed so that the palms face directly upward by the pouching. In no instance were unacceptable time the hands are below the pouch openings items recognized and rejected while in the hands. films (Fig. 2: F26). The forearms are next elevated The suggest that pouch items are quickly tested for such that the fingers penetrate into the ex- suitability before pouching by being pinched between the pursed lips or, treme anterior end of the openings (Fig. 2: in the case of large items (sunflower seeds), F34). In the final stage of the pouching ma- between the lower incisors and the lips. neuver, the food is released and the hands are Pinching of the food appears to be another pulled downward away from the mouth consequence of the coupling of mandibular ready to commence the next cycle (Fig. 2: motion to forelimb movement. Those items F41). judged imacceptable by the "pinch test" are The cine records reveal two additional as- then retrieved from the front end of the pects of the pouching mechanism. First, the pouch and thrown backward beneath the reduced first digit ( = thumb) is used in semi- animal. opposable fashion. This small digit is held be- side the large palmar tubercle and, hence, opposes the remaining fingers (II-V) when Mechanical Constraints grasping food items. Second, the pouching

cycle of the forelimbs is attended by synchro- Pouch size is important in that it estab- nized mandibular movements. Each time the lishes the maximum quantity of food material hands are drawn toward the pouches, the that a heteromyid can transport. The rela-

mandible is pulled rearward. Opening of the tionship between pouch size and body size mouth at this time appears to allow the might therefore influence the foraging tactics hands to enter the pouches while the pouch of desert heteromyid rodents. To examine entrances are themselves kept tightly closed this issue, Morton et al. (1980) have recently to prevent the exit of seeds already within measured mean pouch volume in 13 popu- them. The backward movement of the lower lations representing 11 of hetero- jaw appears to induce tension in the lips myids and one species of geomyid {Tho- which, during pouching, are pursed behind momys bottae). Volume was determined by the incisor teeth. The tension causes the lips filling the pouches of dead animals to near to draw inward away from the lateral walls capacity with material of uniform size and of the pouch, thereby creating small gaps at density (unhusked millet seed; 0.71 g cm -3) the extreme front end of the pouch into and then converting the weight of the con- which the hands are thrust. As the mandible tents to volume. These authors predicted that moves forward (= jaw closing), the pouch pouch volume (Vp) should scale as body mass openings are again closed and the hands (Mb) raised to the first power (Mb^o) using withdrawn. the standard allometric expression (y = ax^).

Seed pouching in D. deserti is rapid, with a They argued that if this relationship existed, mean pouching rate for millet seed of 9.01 larger heteromyids could collect and trans- cycles per second. Some cycles are executed port more food relative to actual metabolic in less than 90 milliseconds. Depending on need (a Mb" '5) than small heteromyids. How- how many seeds are grasped in each hand, ever, their prediction for the scaling of pouch pouching rates range between approximately volume to body size was realized (Vp a 20 and 60 millet seeds sec'. Though the con- }^\)iM3^ only when the sample was limited to ditions under which these values were ob- small heteromyids (< 30 g) and the much tained are admittedly artificial, they indicate larger pocket (116 g). They found no the potential speed and efficiency of the statistically significant relationship between pouching mechanism of Dipodomys under fa- pouch volume and body size within the genus vorable circumstances. Unfortunately, com- Dipodomys. To explain this finding, Morton 50 Great Basin Naturalist Memoirs No. 7

mass was produced only when Thoniomys was included, an inclusion that seems unwar- ranted in view of its systematic position, body plan, locomotor mechanics, and forag- ing behavior. Finally, biomechanical con- straints may prohibit the maintenance of geo- metric similarity between pouch volume and body size in heteromyid rodents. All desert heteromyids use some form of saltation, quadrupedal (Perognathus) or bi- pedal {Dipodomys, Microdipodops), when moving fast. Balance and stability are bio- B. mechanical problems that may increase with

speed, especially if the gait involves rapid changes of direction. Several distinctive structural modifications of the neck in biped- al heteromyids appear to relate to the special problem of head stability (see earlier). The mass of the head will be a critical determi- nant of any stabilization mechanism. More- over, the head must enter into any consid- Fig. 3. A, Stylized illustration of anatomical relation- eration of body balance (particularly in ship of external cheek pouch to head skeleton in Di- it is the largest poikmnjs. Principal anchorage of pouch (and contents) to bipeds) since among and

skeleton is to rostmm and mandible at points indicated heaviest structures forward of the point of by arrows. B, Simplified diagram of mechanics of head limb support. stability under two locomotor conditions. In .smooth bi- Anatomically, the cheek pouches are an- pedal hopping, acceleration of the head relative to chored to the head skeleton (Fig. 3A). At rest, pouch contents (Bh) is slight and has a largely horizontal trajectory. The opposing inertial reaction force (Fi) of much of the load provided by the pouch con- the pouch load is also small and passes close to the fid- tents rests upon the back. However, during cnnn (dot) at the cranio-cervical joint. Accordingly, the forward acceleration of the body, the load force has a short moment arm (m) about the head-neck will tend to shift backward due to inertial joint. The resultant destabilizing torque (= Fi*m)

(clockwise) is likewise small and is opposed by a lag. An appreciable fraction of this inertial counterclockwise tortjue supplied by the neck muscula- force will act on the head. If the mouth is ture (Mn). A much larger and more vertically oriented held closed, most of the inertial force acting inertial reaction force (Fi') results from the rapid, steep on the mandible will be relayed to the ros- trajectory of the .head (Pe) as is occasionally seen during predator escape. This force has a large moment arm (m') trum of the skull through the masseter about the hilcnmi and therefore generates a much great- muscle. The inertial load from each cheek er destabilizing torque on the head. See text. pouch can, for mechanical purposes, be re- garded as concentrated at a single point well et al. (1980) suggest that either relatively out on the rostrum (Fig. 3B). large body size, a preferred diet of high ca- The same figure illustrates the functional loric seeds, and/or bipedalism may have re- consequences of cheek pouch load under two leased kangaroo rats from normal allometric locomotor conditions. In slow, smooth biped- constraints. al hopping, accelerational forces are small Several factors indicate that pouch volume and the resultant inertial force is nearly hori- might not increase as the first power of body zontal. The line of action of this force passes ma.ss. First, such a relationship implies the close to the cervicocranial joint, thereby maintenance of geometric similarity, a pat- yielding only a modest destabilizing torque tern rarely encountered within a phylogenet- on the cranium. In the second case, that of ic .series encompassing an appreciable range escape from a predator, the animal accel- of body size (Gould 1966). Second, in the erates very rapidly in a more vertical trajec- analysis of Morton et al. an isometric rela- tory, similar to that recorded for Dipodomys tionship between pouch volume and body merriami when avoiding the strike of snakes 1983 Biology of Desert Rodents 51

en 52 Great Basin Naturalist Memoirs No. 7

°! 1983 Biology of Desert Rodents 53

filled. pouches only partially When gathering (Quinn 1983). Assuming that predation is a very light items, the same animal could con- major factor in habitat selection, it is pos- ceivably fill its pouches to their volumetric sible, therefore, that the relative effect of limit without ever reaching the load limit. pouch load could influence their choice of The proposed connection between cheek microhabitat. Small species might be ex- pouch loading and predation again rests with pected to forage preferentially in areas of simple biophysical considerations. Suppose a close cover if exposure to predation and the is sifting the for seeds. If distance to the nearest protection increases with the suddenly attacked, it will attempt to leap up "openness" of the habitat. Within and away to avoid capture. The rate with both Perognathus and Dipodomys the largest which the rodent accelerates away from the species should be able to successfully operate in the more open habitats since they attacker is determined by the simple New- are less handicapped by pouch load. tonian relationship, a = F/m, where a is ac- Strictly speaking, the model predicts dif- celeration, m is the mass of the animal and F ferences in microhabitat availability, not is the propulsive force applied to the ground. their actual use. On biomechanical grounds If F is the maximmn force the rodent is ca- large heteromyids are not necessarily ex- pable of generating, it follows that maximum cluded from areas of relatively close cover, acceleration, maximimi take-off velocity and but small species should be excluded from maximum distance covered by the leap (hold- open areas. The range of microhabitats po- ing take-off angle constant and ignoring aero- tentially exploited by heteromyids, with re- dynamic drag) will decline in direct propor- gard to the mechanics of predator escape, tion to cheek pouch load. Hence, the ought therefore to expand with increasing acceleration of a loaded heteromyid is given body size. This raises an interesting question by the expression: a = F/(Mb -I- Mp), where with respect to Microdipodops, which is bi- Mb is the mass of the body and Mp is the ad- pedal but also in the size range of Pe- ditional mass added the contents. by pouch rognathus. At present there are insufficient If heteromyids load their cheek pouches in data to compare the foraging tactics of Mi- constant proportion to head (as mass argued crodipodops with that of the smallest (but above), the potential consequences for pred- substantially larger) kangaroo rats (e.g., D. ator escape are easily ascertained. Among merriami, ordii). Theoretically, pouch loading desert heteromyids the relative loss of accel- should place a kangaroo at greater risk eration due to maximal cheek pouch loading in open habitats than even the smallest should scale as Mbo^3_i g^ as head to body Dipodomys. mass. Small species will therefore be more Unfortunately, we do not yet know how adversely affected than large ones. All other acceleration potential actually scales with things being equal, small heteromyids should body size in heteromyids nor how diis poten- be at greater risk from predation trans- when tial compares with the speed of attack by porting a full load in the pouches. If, for ex- natural predators. These and other con- ample, a 110 g Dipodoniys and a 10 g Pe- founding factors might conceivably alter ex- rognathus were both carrying pouch loads pectations of habitat restriction drawn from equaling percent of 50 head mass, the max- simple biomechanical considerations. Still, imum rate of acceleration of the kangaroo rat the present model based on mechanical con- would be lowered by 4.6 percent and the straints offers straightforward predictions pocket mouse would suffer a 6.8 percent loss that are subject to testing. of function. The ability of heteromyids to accelerate sharply is imdoubtedly a key element in their FoRELIMBS defense against predators. Although vertical leaps and erratic changes of direction have The forelimbs of desert heteromyids have been recorded for D. merriami (Webster several functional demands placed upon

1962), during predator escape there is some them. Foremost among these are food han- evidence that D. microps nm directly to a dling and digging. They are also involved in burrow or bush when suddenly startled body support and propulsion in Perognathus 54 Great Basin Naturalist Memoirs No. 7

at all speeds, but only at comparatively slow even though they are naturally confined to rates of travel in Dipodomys and sandy . Microdipodops. Digging Methods

There are three main methods of digging Digging Activities utilized by heteromyid rodents. Very loose soils, as dry, fine sand are often Nearly all small desert live be- such moved low ground at least part of the day, where by pulling small piles of soil between the ani- front simul- soil acts as a buffer against temperature ex- mals' hind feet using both limbs tremes and desiccation. Below 30 cm of taneously. These motions appear to be very similar those of the forelimbs sandy soil, soil temperature remains relative- to during high ly constant throughout the day, despite fluc- speed pouching of seeds. When a sufficient tuations of 20 C or more at the soil surface pile of soil has accumulated under the body, (Kenagy 1973, Larcher 1980). Uniform tem- the hind limbs are used to kick the sand fur- peratures throughout the year, however, are ther back. This method of digging is used by not achieved except at much greater soil D. merriami and D. deserti during surface for- depths. Most desert heteromyids dig elabo- aging and in the initiation of new tunnels. rate multibranched burrow systems (Ander- Soils of intermediate hardness are loosened son and Allred 1964, Culbertson 1946, Quinn by scratch digging techniques that employ 1983, Vorhies and Taylor 1922) where they the front limbs in an alternating pattern. This spend the day. Typically, they emerge above digging method is employed on the surface groimd to forage only after sunset. when burying seeds and foraging as well as Dipodomys burrows tend to have multiple underground when constructing or maintain- entrances, which are sometimes plugged dur- ing tunnel systems (Eisenberg 1963, 1975, ing the day (Hawbecker 1940, Tappe 1941, Nikolai and Bramble pers. obsers). Soil loos- Vorhies and Taylor 1922). They have a max- ened in this way may be moved with the hind imum depth of 30-75 cm (Anderson and feet by kicking or, when underground, the Allred 1964, Culbertson 1946, Vorhies and animal may turn around and push the soil Taylor 1922). The burrows of Perognathus with its forelimbs and chest (Eisenberg 1963). tend to be less branched. Generally they have This latter method of transporting soil is in- only one or two entrances and are rather variably used to move soil up a tunnel ramp deep, with nest chambers 85-193 cm below preparatory to plugging the entrance (Eisen- the surface (Eisenberg 1963, Kenagy 1973). berg 1963, Nikolai and Bramble pers. obsers).

Little is known about (Mi- The soil is then usually patted into place with crodipodops) burrow systems, except that rapid alternating hand movements. Slow mo- those of M. pallidus and M. megacephalus are tion films show the frequency of such move- short and simple (Eisenberg 1963, O'Farrell ments to be approximately 11.6 cycles per and Blaustein 1974), a fact that may mini- second in Dipodomys merriami and 5.5 cps in mize the energetic cost of torpor (Kenagy Perognathus formosus when working in damp 1973). sand (Table 1). The relative digging abilities of hetero- The third method of digging has been ob- myids has been given very little consid- -served only in Perognathus on extremely re- eration. There is some evidence, however, sistant soils. Here the animal uses its incisors that species may partition the land available on the basis of soil composition and particle size (Hardy 1945, Hoover 1973). Some spe- Table 1. Digging rates for lieteroiiuids in damp packed sand. See text for details. cies appear to be restricted to soft friable soils, but others are able to use harder, rocky BW Digging Patting soils. This would suggest that some species Species (g) (stroke/s) (stroke/s) may be luiable to dig in hard soils. Deynes P. hngimembris 10.0 4.00 P. fDnnosiis 21.1 (1954), however, found that P. merriami gil- vus and P. penniciUatus eremictts were able D. iiicnidini 42.0 7.58 11.6 to dig burrows in heavy clay-loam hard pan. I), denerti 115.0 5.27 8.23 1983 Biology of Desert Rodents 55

to chew through cenientUke soils (Deynes In desert heteromyids each of the four

1954). This digging behavior is similar to that main digits (II- V) bears long, curved but thin seen in the closely related Geomyidae (Hill claws. The claws are used extensively in bur-

1937). It is conceivable that mechanical re- rowing but also appear to serve as winnow- ing rakes to snag seeds strictions on gape (Fig. 1) preclude this type as the hands sift of digging in Dipodomys and Microdipodops. through fine sediment. The reduced first digit has a naillike covering. As suggested above, this finger seems to be semiopposable in Limb Morphology Dipodomys; from its similar structure, the same function may be expected in Micro- The forelimbs of Dipodomys and Micro- dipodops and Perognathus. dipodops, like other bipedal rodents, are short The scapula and humerus of the bipedal compared to the hind limbs (Herman 1979, heteromyids resemble, in several ways, those Howell 1932). Much of the reason for this of highly fossorial mammals. The humerus is seems to stem from the strong negative relatively short, stout, and wide across the allometry of the hand relative to body size in distal epicondyles (Howell 1932). The ratio of bipeds as compared to the slight positive al- epicondylar width to humeral length is about lometry in quadrupeds. The humerus is abso- .30-.33 and .33, respectively, in Dipodomys lutely shorter in bipeds at all sizes, but body and Microdipodops as compared to .34-.36 in its length increases with size at the body the pocket gopher, Thomomys. The relation- same rate as in quadrupedal species (Table 4). ship is .23-. 29 in Perognathus, a value similar The tiny hands of Dipodomys and Micro- to that of generalized quadrupedal rodents probably reflect specialization for dipodops (.24-.28) and also close to the figures re- high speed seed handling and pouching. Rap- ported for Heteromys (.21) and Liomys (.26) id food handling will, in turn, reduce the (Wood 1935). The wide epicondyles of biped- time animal forage an must beyond the safety al heteromyids and other digging mammals of its burrow. The very high rates at which are associated with powerful extensor and D. deserti pouches seed have already been flexor muscles of the wrist and hand as well mentioned. The small hands of the bipedal as highly developed pronators and supinators heteromyids may improve manual dexterity of the forearm (Hildebrand 1982). The scap- better fit by providing a between hand and ula of all heteromyids and geomyids has a small food items. Reduction of the hands may distinct postscapular fossa (Hill 1937, Howell also facilitate high velocity movements of the 1932). A similar fossa has been independently foreamis by reducing the moment of inertia evolved in several groups of highly fossorial of the distal limb segments. A reduction of mammals (e.g., armadillos, , etc.) and mass will higher rates of os- permit cyclic is fimctionally associated with an enlarged cillation without an increase in muscular teres major muscle (Hildebrand 1982). This force (i.e., energy expenditure). structural feature is but one of several that The absolutely higher rates of limb os- hint that the common ancestor of the Hetero- cillation in Dipodomys than in Perognathus myidae and Geomyidae was more fossorial while digging (Table 1) is somewhat surpris- than the more generalized living geomyoids ing. Normally, maximum limb frequency {Heteromys, Liomys) would indicate. would be expected to scale negatively on In both heteromyids and geomyids the body mass, as does maximum stride frequen- ability to pronate and supinate the forearm is cy while running in quadrupedal mammals especially well developed. This ability seems (Heglund et al. 1974). The reduced limb mass likely to be associated with the maneuvers re- associated with the relatively smaller hands quired (see earlier) to effectively place mate- of Dipodomys is, however, unlikely to pro- rials in the cheek pouches. What deserves vide a complete explanation for its more rap- special notice are the structural special- id limb movements as compared to pocket izations that permit such forearm mobility. mice. We suspect that the faster forelimb Unlike most mammals in which pronation movements of Dipodomys may also be the and supination involve chiefly long axis rota- product of historical selection for higher tion of the radius, the movement is accom- rates of food gleening and pouching. plished in geomyoid rodents primarily by 56 Great Basin Naturalist Memoirs No. 7 long axis rotation of the ulna and radius as a also suggested that bipedal locomotion is unit. Such exceptional motion of the ulna more energetically efficient than quad- correlates with an extremely "loose" elbow rupedalism (Dawson 1976, Dawson and Tay- joint in which the ulna is free to deflect in- lor 1973) or that bipedal locomotion is more ward and outward. A special check ligament effective in predator avoidance (Eisenberg connects the lateral epicondyle of the hu- 1975). merus to the lateral crest of the ulna. It pre- Based on energetics studies, Dawson (1976) vents excessive medial deflection of the fore- has argued that small bipedal mammals, in- arm on the humerus and serves to stabilize cluding heteromyids, are able to move faster the otheiAvise weak elbow joint when the using less energy than their quadrupedal hand is flexed in the supine position (as in counterparts. This suggests that quadrupedal pouching and some digging maneuvers). The heteromyids (such as Perognathus, Heteromys, strengthen elbow ligament cannot, however, and Liornys) might have foraging strategies movements the hand is pronated (palm when that do not utilize high-speed running. Sever- down), as would be the case in quadrupedal al ecological studies have been undertaken to locomotion. This raises the possibility that test this idea (Reichman 1981, Thompson structural limitations may make bipedalism 1980, 1982a, 1982b). obligatory for larger desert heteromyids A high-speed bipedal gait would be most when ninning at higher speeds. The presence effectively used by a small desert rodent to of a specialized elbow mechanism in hetero- avoid predators (avian, mammalian) while myids and primitive geomyids (Thomomijs), moving from its burrow to a protected forag- together with its absence in selected repre- ing site or from one site to another. Bipedal sentatives of other rodent families (i.e., Mu- hoppers may also be more adept than quad- ridae, Sciuridae, Cricetidae, Zapodidae, rupeds at avoiding predators in open habitat Chinchillidae), indicates that the mechanism due to the greater maneuverability and accel- is derived for rodents but primitive for the . eration offered by the bipedal hop. Further, if bipedal hopping is energetically less costly than quadrupedal nmning, then maximimi Hind Limbs and Locomotion speed may be greater in bipedal hoppers than in similarly sized quadrupeds. These The locomotor repertory of heteromyids considerations seem to imply that quad- can be divided into two major classes: biped- rupedal heteromyids might be forced to con- al and quadrupedal. Both bipedal and quad- fine their foraging to one or two shrubs close rupedal heteromyids use quadrupedal gaits, to their burrow entrance, but bipedal hetero- such as the walk, half-bound and full-bound myids may be free to forage imder several during slow progression. These gaits are very more widely scattered shrubs. Thompson similar in their footfall pattern to the gaits (1980, 1982a, 1982b) has compared the forag- used by other quadrupedal rodents (Gam- ing behaviors of D. deserti, D. merriami, and baryan 1974). At higher speeds, however, P. longimembris (see Table 2), and his results some heteromyids {Dipodoinys and Micro- seem to support this hypothesis. dipochps) employ the bipedal hop, a gait that The energetic cost of locomotion in kan- does not use the front limbs for support. garoos increases linearly with speed during Since much of the thnist of this gait is associ- quadnipedal (pentapedal) movement, but re- ated with dorso- ventral oscillation of the ver- mains constant or may even decrease slightly tebral column, the bipedal hop is allied with during bipedal hopping (Dawson and Taylor the gallop as an asymmetrical gait (Badoux 1973). This contrasts sharply with the pattern 1965, Hatt 1932, Howell 1965). of quadrupedal mammals wherein the ener- A question that has plagued researchers for getic cost continues to increase linearly at all some time is: Why did bipedalism evolve in speeds (Taylor et al. 1970). It has been sug- some heteromyids and not in others? Bipedal gested that elastic storage of strain energy in locomotion seems certainly to permit greater the muscles and tendons of the hind limb and specialization of the forelimbs for digging back of may be responsible for and food handling. Several researchers have some of their "energy savings" (Alexander 1983 Biology of Desert Rodents 57

Table 2. Comparison of foraging behaviors of bipedal and quadrupedal heteronivids 58 Great Basin Naturalist Memoirs No. 7

Quadrupedal LH •- LF - RF H- RH h-

Tripodal

LH H H 1 LF -

RF I I RH»- Bipedal LH LF - RF - RH K

I sec,

Fig. 7. Gait diagrams for D. merriami. Three representative gait diagrams depict the footfall patterns of the quad- rupedal bound (top), tripodal half bound (center), and bipedal hop (bottom) at 60 cm /sec. During quadrupedal and tripodal locomotion, front limb support ends at touch-down for the hind limbs. Note that, although speed is the

same for all three gaits, stride frequency is highest for the bipedal hop and stride length is greatest for the tripodal half bound.

transition speeds = 3.7 (Body Weight) i"'*. Pe- speed, but well below that of a similarly

detes capensis, which is morphologically very sized Perognathns. similar to Dipodomys (Berman 1979), has a The ratio of forelimb length to hind limb ratio of 35.4 and a quadrupedal-bipedal tran- length has been used extensively as an index sition of 4 km/hr (Thompson et al. 1980), of bipedality (Berman 1979, Howell 1932). It

which is slightly lower than that predicted by is clear that the hind limbs are elongated in the kangaroo rat equation. Kangaroos have bipedal hoppers and that most of the length-

ratios of 43.7-47.7 (Gambaryan 1974) and ening occurs in the distal segments (i.e., tibia quadnipedal-bipedal transition speeds of 6.5 and foot elements) (Berman 1979, Emerson

km/hr (Dawson and Taylor 1978). This is ms., Howell 1932). However, whether or not slightly above the value predicted by the the forelimbs are shortened relative to body

Dipodomys equation (Fig. 8). of the size is the subject of some controversy (Gam- genus Allactaga have front to hind limb ra- baryan 1974, Howell 1932). Most osteological

tios of 27.5 (Gambaryan 1974) and therefore measures of animal size in common use (i.e., should have very low quadnipedal-bipedal basicranial length, thoracolumbar length, transition .speeds. The front limb to hind limb body length) are greatly modified in kan- ratio of Perognathus (55.7-57.7) indicate that garoo rats, and therefore of dubious value it would, in theory, be able to hop bipedally when making comparisons with more gener- only at very high speeds. The limb propor- alized rodent species. An exception to this is tions of Microdipodops (51.2) predict a rela- the length of the basioccipital bone of the tively high (juadnipedal-bipedal transition skull, which appears to be relatively immodi- 1983 Biology of Desert Rodents 59

E 8-

"O CD / CD / Q- /

O -E ^ M M if)a O n

10 log BW (g)

Fig. 8. Gait transition speeds. The trot-gallop transition speed for quadrupeds (dashed line) is 5.5 Mb-^ (Heglund et al. 1974). The quadrupedal-bipedal transition speed for kangaroo rats (solid line) is 3.7 Mb-^'^'^ (Nikolai ms.). The measured quadrupedal-bipedal transition speeds for bipedal hoppers are: (a) Red kangaroo (18 Kg), [K], 6.5 Kin/hr (Dawson and Taylor 1973); (b) sp. (3 Kg), [P], 4.0 Km/hr (Thompson et al. 1980); (c) D. deserti (.104 Kg), [D], 2.5 Km/hr (Tliompson et al. 1980); (d) D. merriami (.0426 Kg), [M], 2.2 Km/hr (Nikolai ms.); (e) D. merriami (.032

Kg), [M], 2.0 Km/hr (Thompson et al. 1980). See text for details.

Bed in kangaroo rats, as well as in many ume and muscular strength as a cross section- other mammals (L. Radinsky, pers. comm.). al area (Alexander 1968), the ability of the In desert heteromyids basioccipital length leg muscles to absorb the shock of impact scales isometrically with body mass. We have during bipedal hopping should decrease with therefore used this as a standard against increasing body size if geometric similarity is which limb segment lengths are compared maintained. Increasing the length of the leg,

(Table 4). however, will increase the contact time and The hind limbs of Dipodomys and Micro- thus the time course over which impact dipodops are greatly elongated, and all three shock can be absorbed. The ratio of calcaneal limb segments show strong positive allometry length to total foot length also increases with .514 with respect to body size (Table 4). The foot body mass in bipedal hoppers [= .09 + and tibia are much longer in bipedal rodents Mb, R2 = .72, P < .02 for 7 species of Di- than in similarly sized quadrupedal rodents, podomys and Microdipodops], thus improving ankle and there is a lesser difference in femur size the mechanical advantage of the large

(see also Herman 1979). It is interesting that extensor muscles. as body size increases the relative length of The vertebral column, pelvic girdle, and the hind limb increases in bipedal hetero- hind limb musculature of bipedal hetero- myids {Dipodomys and Microdipodops). This myids are considerably modified compared to may have important biomechanical con- quadrupedal species (Berman 1979, Hatt sequences. Since body mass scales as a vol- 1932, Howell 1932). In general, the long 60 Great Basin Naturalist Memoirs No. 7

Table 4. Comparison of the allometric equations of limb segment length of bipedal heteromyids and quadrupedal rodents using basioccipital length as a standard unit of relative body size. See text for details. [Bipedal heteromyids:

M. megacephaliis (1), D. merriami (2), D. ordii (3), D. microps (3), D. spectabilis (1); Quadrupedal rodents: P. long- sp. sp. Eutamicis sp. imembris (1), P. fomiosus (3), ?. parvus (2), Tlwmomys sp. (2), Pewmyscus (1), Neotoma (2), (1),

Citelhis sp. (1)]. Number of animals in parentheses. 1983 Biology of Desert Rodents 61

Sr^^^^^e::^ ^r:^^^^ 62 Great Basin Naturalist Memoirs No. 7 diurnal rodent but could easily spell the dif- treadmill show that the forelimbs are in- ference between capture and escape to a soli- volved in body support and shock absorption tary, nocturnal animal in open terrain. in every stride (Bramble and Nikolai pers. Morphological specialization for leaping obs.). Further evidence of quadrupedal rather among vertebrate animals invariably results than bipedal bounding is found in the struc- in elongation of the hind limbs relative to the ture of the limbs. The forelimb-hindlimb forelimbs (Herman 1979, Howell 1965). As a length ratio of Zapus is similar to that of bi- will consequence, the forelimbs necessarily pedal heteromyids (i.e., Microdipodops; Ber- stresses as they act incur higher mechanical man 1979), but the forelimb is constructed as shock absorbers to break the fall of the differently. Forelimb length relative to body longer, faster bounds generated by the rear size is comparable to that seen in generalized legs (Gambaryan 1974). Stress on the fore- quadrupedal rodents. The hand is large limbs will be amplified if a high-speed rather than reduced, and there are no obvious bounding gait incorporates abrupt changes of specializations favoring pronation and supi- direction, since the front limbs brake the for- nation of the forearm. We tentatively con- ward momentum of the body as the turn is clude that true bipedal locomotion probably executed. Quadnipedal mammals with such does not exist in the Zapodidae and that leap- gaits (e.g., cursorial lagomorphs, ungulates) ing specializations of this group have little to exhibit extreme modification of the elbow to do with the evolutionary pathways leading to increase its resistance to injury. Long axis ro- bipedal saltation in modern desert rodents. tation of the radius and ulna is severely cur- The development of bipedal saltation as tailed or eliminated (Hildebrand 1982, How- seen in modern heteromyids cannot have ell 1965). This effectively precludes been associated with the occupation of desert pronation and supination of the hands, mak- environments as we know them today. True ing them nearly useless in feeding. In all bi- of North America appear to be of pedal rodents, by contrast, the hands are used fairly recent origin (i.e., later Pleistocene; extensively in feeding and digging. There has Van Devender 1977), whereas heteromyids presumably been strong selection for the rap- exhibiting structural modification for bipedal id, efficient use of the forelimbs in order to saltation date from at least the later Miocene reduce foraging time. It is in this context that (Voorhies 1975, Wood 1935). Voorhies' tlie evolutionary significance of bipedalism (1975) recent suggestion that bipedalism may becomes clearer. This locomotor strategy have first arisen among primitive dipodo- seems to offer the only viable means of com- myines living in sandy, floodplain habitats bining, in a single animal, limb special- deserves special consideration. Such re- izations and functions which are otherwise stricted environments within incompatible. may occur otherwise typical savannah habitat and may The notion that bipedalism in rodents is di- be relatively arid during periods of low rain- rectly linked to the occupation of open, arid fall. More importantly, sandy floodplains are habitats is seriously contested only by the modem Zapodidae. These rodents favor mes- frequently characterized by widelv .scattered ic, well-vegetated environments both in the vegetation and thus qualify as "open" habi- New and Old worlds. Their possible relation- tat. If bipedal heteromyids arose under con- ship to the Dipodidae (Eisenberg 1981, Fokin ditions such as these, their distinctive lo- 1978) as well as the presumed bipedal habits comotor specializations may have of 7Aipus would seem to make the locomotor considerably predated other specific adapta- behavior of the.se rodents of special value in tions (mostly physiological) to desert life. unraveling the history of bipedalism in desert rodents. However, although Zapus is unques- Literature Cited tionably capable of rapid, prodigious leaps, Ai.nKiuH, P., S. C. F. Osteh, D. B. Wake. there seems to be no solid evidence that it is J. Gould, and 1979. Size and shape in ontogeny and plivlogeny. really capable of sustained, bipedal saltation. Paleobiology 6:296-317. Slow motion films of Zaptis princeps running Alexanoeh, R. McN. 1968. .\niinal nieclianies. I'niv. of and leaping solid on ground as well as on a Washington Press. Seattle. 1983 Biology of Desert Rodents 63

Alexander, R. M, and A. Vernon. 1975. The mechan- Gambaryan, p. p. 1974. How mammals rim. Halsted ics of hopping by kangaroos (Macropodidae). Press, J. New York. 367 pp. Zool., London 177:265-303. CiouLD, S. J. 1966. Allometry and size in ontogeny and Anderson, A. O., and D. M. Allred. 1964. Kangaroo phylogeny. Biol. Rev. 41:587-640. rat burrows at the site. Nevada test Great Basin Hardy, R. 1945. The influence of types of soil upon the Nat. 24:93-101. local distribution of some mammals in south- .'Vrmitage, K. B. 1981. Sociality as a life history tactic of western Utah. Ecol. Monogr. 15(1):71-108. the . ground Oecologia 48:36-49. Hatt, R. T. 1932. The vertebral column of ricochetal ro- Badoux, D. M. 1965. Some notes on the fimctionai anat- dents. Bull. Amcr. Mus. Nat. Hist. 63 (Article omy of Macroptis giganteus Zimm. with general 6):599-745. remarks on the mechanics of bipedal leaping. R^wbecker, a. G. 1940. The burrowing and feeding Acta Anat. 62:418-433. habits of Dipodomys ventistus. J. . Barash, D. p. 1973. The social biology of the Olympic 1940:,388-396.

marmot. Anim. Behav. Monographs 6(3): 171-245. Hayden, p., and J. J. Gamblno. 1966. Growth and devel- Bartholomew, G. A., and G. H. Gary. 1954. Locomo- opment of the , Perognathus tion in mice. pocket J. Mammal. .35(3):386-392. longimembris. Growth 30:187-197. Bartholomew, G. A., and H. H. Gaswell. 1951. Loco- Heglund, N. G., C. R. Taylor, and T. A. McMaho.n. motion in kangaroo rats and its adaptive signifi- 1974. Scaling stride frequency and gait to animal cance. .32:155-169. J. Mammal. size: mice to horses. Science 186:1112-1113. Ber.man, S. L. 1979. Convergent evolution in the hind Hildebrand, M. 1982. Analysis of vertebrate structure. limb of bipedal rodents. Unpublished dis- 2d ed. John Wiley and Sons, New York. 654 pp. sertation. Univ. of Pittsburgh, Pittsburgh, Hill, J. E. 1937. The morphology of the pocket gopher Pennsylvania. mammalian genus Tliomomys. Univ. California Biewener, a., R. McN. Alexander, and N. G. Publ. Zool. 42(2):81-172. Heglund. 1981. Elastic energy storage in the Hoover, K. D. 1973. Some ecological factors influencing hopping of kangaroo rats (Dipodomi/s specfabilis). the distribution of two species of pocket mice Zool., 195:.369-383. J. London (genus Perognathus). Unpublished dissertation. BuTTERwoRTH, B. B. 1961. A comparative study of New Mexico State Univ., Las Cnices. growth and development of the kangaroo rats Howell, A. B. 1923. Abnormal hairy growths upon the Dipodomijs deserti Stephens and Dipodoniys mer- tails of the 4:56-58. heteromyidae. J. Mammal. riami Mearns. Growth 25:127-1.39. 1932. The saltatorial rodent Dipodomys: the Ghew, R. M., and B. B. Butterworth. 1959. Growth fimctionai and comparative anatomy of its mus- and development of Merriam's kangaroo rat, cular and osseous systems. Proc. Amer. Acad. Dipodomijs nierriami. Growth 23:75-95. Arts and Sci. 67:376-536. Ghiasson, R. B. 1954. The phylogenetic significance of 1965. Reprint. Speed in animals. Hahier Pub- rodent cheek pouches. Mammal. 35(3):425-427. J. lishing Co., New York. 270 pp. GsuTL B. 1979. Patterns of adaptation and variation in Jackson, D. C., and K. Schmidt-Nielsen. 1964. the Great Basin kangaroo rat (Dipodomys vii- Gountercurrent heat exchange in the respiratory crops). Univ. Galifornia Publ. Zool. Vol. IH, 75 passages. Proc. Natl. Acad. Sci. USA pp. 51:1192-1197. Gulbertson, a. E. 1946. Observations on the natural Kaup, U. L. 1975. Investigation on the musculature sys- history of the Fresno kangaroo rat. Mammal. J. tem of the hind extremity in bipedal rodents (in 27(3): 189-203. German). Zool. Anz. Jena 194:416-436. Dawson, T. 1976. Energetic cost of locomotion in J. Kenagy, G. 1972. Saltbush leaves: excision of hypersa- Australian hopping mice. Nature 259:305-307. J. line tissue by a kangaroo rat. Science Dawson, T. and G. R. Taylor. 1973. Energetic cost ' J., 178:1094-1096. of locomotion in kangaroos. Nature 246:313-314. 1973. Daily and seasonal patterns of activity and Deynes, a. 1954. Habitat restriction and the digging energetics in a heteromyid rodent community. ability of certain pocket mice. Mammal. J. Ecology 54:1201-1219. 35:453. A. 1967. Growth and development of Di- Lackey, J. Eisenberg, J. E. 1963. The behavior of heteromyid ro- podomys stepbensi. Mammal. 48:624-632. dents. Univ. Galifornia Publ. Zool. 69:1-100. J. Larcher, W. 1980. Physiological plant ecolog)'. 2d Ed. 1975. The behavior patterns of desert rodents. Springer-Verlag Publishers, Berlin. Pages 189-224 in I. Prakash and P. K. Ghosh, eds. Lay, D. M. 1972. The anatomy, physiology, fimctionai Rodents in desert environments. Monographae and evolution of the specialized Biologicae. W. Junk, The Hague, The significance, gerbilline rodents. Morphol. Netherlands. hearing organs of J. 1981. Mammalian Radiations. Univ. of Ghicago 138:41-120. Legouix, P., F. Petter, and A. Wisner. 1954. Etude Press, Ghicago. 610 pp. J. de L'Auditien chez des Mammiferes a Bulles Emerson, S. B. Unpublished manuscript. 42 pp. Tympanique Hypertrophis. Mammalia 18(3):262- FoKiN, L M. 1978. The locomotion and morphology of the locomotory organs in jerboas (in Russian). 271. 1972. Small mammal fossils from the Academy of Sciences of USSR, Zoology Institute. Lindsay, E. H. Barstow Formation, California. Univ. California Leningrad. 1 19 pp. Frye, R. Publ. Geol. Sci. 93:1-104. J., AND M. L. RosENZwEiG. 1980. Glump size Proske, and D. Warren. 1978. Mea- selection: a field test with two species of Di- Morgan, D. L., U. stiffness and mechanism of podornys. Oecologia 47:323-327. surements of muscle 64 Great Basin Naturalist Memoirs No. 7

elastic storage of energy in hopping kangaroos. J. Thompson, S. D., R. E. MacMillen, E. M. Burke, and Physiol. 282:253-261. C. R. Taylor. 1980. The energetic cost of biped- Morton, S. R., D. S. Hinds, and R. E. MacMillen. al hopping in small mammals. Nature 1980. Cheek pouch capacity in heteroinyid ro- 287:223-224. dents. Oecologia 46:143-146. Van Devender, T. R. 1977. Holocene woodlands in the Murray, P. 1975. The role of cheek pouches in cerco- southwestern deserts. Science 198:189-192. pithecine monkey adaptive strategy. Pages Vaughan, T. 1966. Food-handling and grooming behav- 151-194 in R. H. Vuttle, ed. Primate fimctional iors in the plains pocket gopher. Mammal. morphology and evolution. Mouton, The Hague, J. 47:1.32-1.33. Tlie Netherlands. Vernon, P. Herman, and E. Peterson. 1971. Coch- manuscript. 43 J., Nikolai, J. C. Unpublished pp. lear potentials in the kangaroo rat. Dipodomijs. a. R. 1974. O'Farrell, M. J., and Rlaustein. Micro- Physiol. Zool. ,35:248-255. dipodops inegacephalus. Pages 1-3 in Mamma- lian Species No. 46. Voorhies, M. R. 1975. A new genus and species of fossil rat its PiNKHAM, C. F. A. 1976. A biomechanical analysis of kangaroo and burrow. J. Mammal. leaping in the kangaroo rat and the Mexican 56(1): 160- 176. . Proc. Utah Acad. Sci., Arts, VoRHiES, C. T., AND W. P. Taylor. 1922. Life history of

Lett. 5.3(2): 1-15. the kangaroo rat Dipodomijs spectabilis spec- B. of spe- QuiNN, J. 1983. The ecology two kangaroo rat tabilis Merriam. U.S. Dept. Agr. Bull. No. 1091. cies (genus Dipodomijs) in southwestern Utah. 40 pp. Unpublished thesis. Univ. of Utah, Salt Lake Webster, D. B. 1961. The ear apparatus of the kangaroo City. rat, 108:12.3-148. Dipodomijs. Amer. J. Anat. Reichman, O. 1979. Desert granivore foraging and its J. 1962. A function of the enlarged middle ear cav- impact on seed densities and distributions. Ecolo- ities of the kangaroo rat, Dipodomijs. Physiol. gy 60(6): 1085-1092. Zool. .35:248-255. 1981. Factors influencing foraging in desert ro- Webster, D. B., R. F. Ackermann, and G. C. Longa. dents. In Alan C. Kamil and Theodore D. Sar- 1968. Central auditory system of the kangaroo gent, eds., Foraging behavior. Garland Pub- rat, Dipodomijs merriami. Neurol. bshing. New York. J. Comp. O. 1.33:477-494. Reichman, J., and O. Oberstein. 1977. Selection of seed distribution types by Dipodomijs merriami Webster, D. B., and C. R. Stack. 1968. Comparative and Perognathiis ampins. Ecology 58:636-643. histochemical investigation of the Organ of Corti Reynolds, H. G. 1960. Life history notes on Merriams in the kangaroo rat, gerbil, and guinea pig. J. kangaroo rat in southern Arizona. J. Mammal. Morphol. 126:413-434. 41:48-58. Webster, D. B., and W. F. Strother. 1972. Middle ear Rosenzweig, M. L., and P. W. Sterner. 1970. Popu- morphology and auditory sensitivity of hetero- lation ecology of desert rodent communities: inyid rodents. Amer. Zool. 12:727 (Abstr.). body size and seed-husking as bases for hetero- Webster, D. B., and M. Webster. 1971. Adaptive value inyid coexistence. Ecology 51:217-224. of hearing and vision in kangaroo rat predator Ruppert, a. L., and G. Moushegian. 1970. Neuronal re- avoidance. Brain, Behav. Evol. 4:310-322. sponses of kangaroo rat ventral cochlear nucleus 1972. Kangaroo rat auditory thresholds before to low-frequency tones. Expt. Neurol. 26:84-102. and after middle ear reduction. Brain, Behav. Schmidt-Nielsen, K., F. R. Hainsworth, and D. E. Evol. 5:41-53. MuRRisH. 1970. Counter-current heat exchange 1975. Auditory systems of Heteromyidae: func- in the respiratory passages: effect on water and heat balance. Resp. Physiol. 9:263-276. tional morphology and evolution of the middle ear. Sherman, P. W. 1977. Nepotism and the evolution of J. Morphol. 146:343-376. alarm calls. Science 197:1246-1253. 1977. Auditory systems of the Heteromyidae: diversity. 152:153-170. Shotwell, J. A. 1967. Tertiary geomyoid rodents of cochlear J. Morphol. Oregon. Univ. Oregon. Mus. Nat. Hist. Bull. 1980. Morphological adaptations of the ear in 9:1-51. the rodent family Heteromyidae. .^mer. Zool. Tappe, D. T. 1941. Natural history of the Tulare kan- 20:247-254. garoo rat. Mammal. 22:117-148. J. Williamson, R. G., Jr., and E. C. Frederick. 1977. A Thompson, S. D. 1980. Microhabitat use, foraging be- functional analysis of ankle extension in the rico- havior, energetics and community stnictvne of chetal rodent (Dipodomijs merriami). Zbl. Vet. heteroinyid rodents. Unpublished dissertation. Med. C. Anat. Histol. Embryol. 6:157-166. Univ. of California, Irvine. A., P. WisNER, J. Legouix, and F. Petter. 1954. Etude __ 1982a. Microhabitat utilization and foraging be- Histologique de L'Oreille d'un Rongeur a Rulles havior of bipedal and quadnipedal heteromvid Tympaniques Hypertrophies, Meriones crassus. rodents. Ecology 63(5):1303-13I2. Mammalia 18:.37i-374. 1982b. Stnicture and species composition of Wood, .\. E. 19.35. Evolution and relationsliips of the desert heteromyid rodent species assemblages: ef- heteromyid from fects of a simple habitat manipulation. Ecologv rodents with new forms the 63(5):1313-1.321. Tertiary of western North .\inerica. (Jarnegie Unpublished manuscript. The energetic cost of Mus., Ann. 24:73-262. locomotion in the kangaroo rats Dipodomifs 1937. Parallel radiation among the geomyid ro- dents. 18:171-176. deserti and D. merriami. 35 pp. J. Manunal.