This content has been downloaded from IOPscience. Please scroll down to see the full text.

Download details:

This content was downloaded by: dgoldman

IP Address: 130.207.140.49

This content was downloaded on 02/07/2015 at 13:37

Please note that terms and conditions apply.

Controlled preparation of wet granular media reveals limits to lizard burial ability

View the table of contents for this issue, or go to the journal homepage for more

2015 Phys. Biol. 12 046009

(http://iopscience.iop.org/1478-3975/12/4/046009)

Home Search Collections Journals About Contact us My IOPscience

iopscience.iop.org/page/termshttp://iopscience.iop.org/1478-3975/12/4http://iopscience.iop.org/1478-3975http://iopscience.iop.org/http://iopscience.iop.org/searchhttp://iopscience.iop.org/collectionshttp://iopscience.iop.org/journalshttp://iopscience.iop.org/page/aboutioppublishinghttp://iopscience.iop.org/contacthttp://iopscience.iop.org/myiopscience

Phys. Biol. 12 (2015) 046009 doi:10.1088/1478-3975/12/4/046009

PAPER

Controlled preparation of wet granular media reveals limits to lizardburial ability

Sarah S Sharpe, RobynKuckuk andDaniel I GoldmanSchool of Physics, Georgia Institute of Technology, USA

E-mail: daniel.goldman@physics.gatech.edu

Keywords: biomechanics, granularmedia, locomotion

Supplementarymaterial for this article is available online

AbstractMany animalsmovewithin ground composed of granularmedia (GM); the resistive properties ofsuch substrates can depend onwater content and compaction, but little is known about how suchparameters affect locomotion or the physics of drag and penetration. Using apparatus to controlcompaction ofGM, our recent studies ofmovement in dryGMhave revealed locomotion strategies ofspecialized dry-sand-swimming reptiles. However, these animals represent a small fraction of thediversity and presumed burial strategies of fossorial reptilian fauna.Here we develop a system to createstates of wetGMof varyingmoisture content and compaction in quantities sufficient to study theburial and subsurface locomotion of theOcellated skink (C. ocellatus), a generalist lizard. X-rayimaging revealed that inwet and dryGM the lizard slowly buried ( 30≈ s) propagating awave fromhead to tail, whilemoving in a start-stopmotion. During forwardmovement, the head oscillated, andthe forelimb on the convex side of the body propelled the animal. Although body kinematics and ‘slip’were similar in both substrates, the burial depthwas smaller inwetGM. Penetration and drag forceexperiments on smooth cylinders revealed that wetGMwas 4≈ ×more resistive than dryGM. In total,ourmeasurements indicate that while the rheology of the dry andwetGMdiffer substantially, thelizardʼs burialmotor pattern is conserved across substrates, while its burial depth is largely constrainedby environmental resistance.

1. Introduction

A diversity of animals live on and within granularmedia (GM), materials that can display solid andfluid-like features. DryGM such as that found in sandydeserts, are collections of particles that interactthrough dissipative, repulsive contact forces [1]. Usinglaboratory controlled substrates, recent studies(including those from our group) have investigatedlocomotion on dry GM; these include subsurfacesand-swimming of a lizard and snake [2–4], andabove-ground sidewinding snakes [5], crawling turtles[6], and running lizards [7–9]. Theoretical, computa-tional and robotic models have helped elucidate howanimals interact with the surrounding GM to generatesuccessful forward locomotion [10–15].

Critical to these studies of movement in dry GMhas been the development of apparatus (e.g. air flui-dized bed trackways) that can create laboratory

versions of the environment where parameters, likecompaction and grain size, can be systematically var-ied, resistive forces can be measured, and animal stu-dies can be compared with models that incorporatethe same substrate properties. The combinedapproach of experiment, modeling and substrate con-trol has been particularly successful in developing anunderstanding of how the sandfish lizard Scincus scin-cus swims within dry GM by propagating a wave ofcurvature down its body, thereby propelling itselfthrough a ‘frictional fluid’ (for a review see [16]). Thesandfish is a dry sand specialist having adaptationswhich include a shovel-shaped snout, fringed toes, andventral breathing which enable it tomove effectively indryGM.

Wet GMmake up an even greater proportion [17]of terrestrial substrates, occurring in rainforest soils,coastal regions, and agricultural lands. The addition ofsmall amounts of water to dry GM can have large

OPEN ACCESS

RECEIVED

19December 2014

REVISED

16April 2015

ACCEPTED FOR PUBLICATION

5May 2015

PUBLISHED

25 June 2015

Content from this workmay be used under theterms of theCreativeCommonsAttribution 3.0licence.

Any further distribution ofthis workmustmaintainattribution to theauthor(s) and the title ofthework, journal citationandDOI.

© 2015 IOPPublishing Ltd

http://dx.doi.org/10.1088/1478-3975/12/4/046009mailto:daniel.goldman@physics.gatech.eduhttp://dx.doi.org/10.1088/1478-3975/12/4/046009http://crossmark.crossref.org/dialog/?doi=10.1088/1478-3975/12/4/046009&domain=pdf&date_stamp=2015-06-25http://crossmark.crossref.org/dialog/?doi=10.1088/1478-3975/12/4/046009&domain=pdf&date_stamp=2015-06-25http://creativecommons.org/licenses/by/3.0http://creativecommons.org/licenses/by/3.0http://creativecommons.org/licenses/by/3.0

effects on the rheological properties of these substrates[18–20]. Despite the myriad of animals that walk, run,and burrow within wet terrestrial soils, few studieshave explored these locomotion strategies [21–23] andeven fewer animal studies have systematically variedwater content and/or compaction [24] of the sub-strate. We note that the rheology of such unsaturatedwet substrates differs from the fully saturated sub-strates that have been recently used to challenge otherorganisms, like razor clams [25] and sand lancefish [26].

Soil water content fraction can vary with factorssuch as geographical location, weather conditions,time of day, or depth beneath the surface.Water indu-ces cohesion between sand grains due to the surfacetension of the fluid and capillary effects [19]. Thisincreased cohesion results in an increased angle of sta-bility for a granular pile, such that sandcastle struc-tures and tunnels can be formed [19, 28], and caninfluencemechanical properties such as shear strength[29]. As water content increases, wet GM is usuallygrouped into four states: pendular—in which liquidbridges form between particle contacts; funicular—where the liquid fills the pore-space spanningmultipleparticles but some air-filled voids remain; capillary—where all pores are filled with liquid; and the slurrystate—in which particles are completely immersed inthe liquid and no capillary action occurs at the surface[18, 19]. The top layer of soil found on the majority ofterrestrial surfaces are in the pendular or funicularstate [17].

In a review of the physics of wet GM, Mitarai [19]stressed the importance of ‘careful preparation’ oflaboratory versions of wet GM to obtain reproducibleresults, andmentioned that compacting the wetmediain a different way could generate different mechanicalresponses. GM compaction is characterized by thevolume fraction, ϕ, defined as the ratio of the volumeof dry GM to the volume of the occupied space [1]. Fordry granular substrates in nature, ϕ can range from0.55 to 0.63 [30]. The compaction can be controlled inthe laboratory setting by using an air-fluidized bed[13, 31] in which a combination of air-flow and shak-ing is used to generate differentϕ states. A 4% increasein ϕ can increase resistance forces by 50% [2]. WetGM can exhibit a larger range inϕ than dry GM; lowerϕ states can be achieved in wet GM by the stabilizingeffects of the liquid bridges [19]. However, in wet GMexperiments, compaction is typically not reported orcontrolled.

Currently, there is no well-established techniqueto create repeatable, homogeneous preparations ofwet GM with controllable compaction and wetness.The techniques that have been used previouslyinclude: (1) mixing the liquid and drymedia, followedby slow pouring into a container [32]; (2) vigorousshaking of the liquid-sand preparation to homo-geneously distribute the water [18, 29]; (3) construct-ing thin (2.5–5 mm thick) layers of sand and using a

fine jet to spray a known quantity of water onto eachlayer before another is added [33]; (4) imbibition (awetting process) in which water is allowed to percolatethrough the media [34]; (5) use of a humidifier [35] tointroduce water into the substrate; and (5) drainage (adrying process) [34]. Studies that have attempted tocontrol or alter the compaction do so by using a ram[36] or by tapping the material [18, 35, 37] to achievehigher ϕ states. While these methods can achievehomogeneity for high ϕ states, either these methodsdo not produce low ϕ preparations or, if they do, inour experience these preparations may contain largevoids which vary in size and location from trial to trial.

An example of an animal that successfully [27, 38]contends with GM of different water contents is theOcellated skink (Chalcides ocellatus), a habitat general-ist lizard found in coastal regions and near vegetationwhere moist sand is common (figure 1) [27]. TheOcellated skink has a slender body with relativelyreduced limbs [39] that lack toe fringes and, like thespecialist sand-swimming sandfish lizard, has awedge-shaped head. Ocellated skinks are commonlyfound near vegetation and ‘harder’ soil whereas thesandfish is found farther from vegetation on the dry,aeolian sands of the dunes. Fields observations haveshown that the Ocellated skink spends most of its timenear cover and will bury into granular substrates toseek refuge fromheat and predators [40].We note thatOcellated skinks co-exist with the sandfish lizards inthe sand dunes of North Sinai, Egypt [40]. However,within this environment the Ocellated skink occupiesa different micro-habitat. Thus for these two lizards ofthe family Scincidae that occupy the same regions ofnorthern Africa and the middle east, it is interesting tocontemplate how the generalist Ocellated skinkʼs bur-ial pattern compares to the specialist sandfish skink indry GM and how the Ocellated skinkʼs burial patternchanges with substrate properties.

In this paper we present a new technique whichuses a sieve apparatus to deposit mixed wet GM into acontainer, resulting in an initially loosely packed statewithout large voids; we then utilize additional shakingto achieve a desired compaction. Using this techniquein combination with x-ray imaging allows us to sys-tematically study how substrate properties affect bur-ial mechanics and performance of the Ocellated skink,and also enables characterization of substrate resis-tance. We will show that, despite common ancestryand an overlap of ranges, the Ocellated skink does notswim like the sand-specialist sandfish lizard (the onlyother sand-swimming lizard studied in detail to date[2, 3, 15]) which swims at relatively high forwardspeeds ( 10≈ cm s−1 or 0.9 SVL s−1) using large ampli-tude body undulations while its limbs remain close toits body. Instead we will show that the generalist Ocel-lated skink uses a more complex burial pattern invol-ving body undulation, head oscillation and limb use.The locomotion pattern is similar across substrateconditions, and drag forcemeasurements indicate that

2

Phys. Biol. 12 (2015) 046009 S S Sharpe et al

this pattern constrains the animal to bury to depthslargely limited bymuscular force of body bending.

2.Materials andmethods

2.1.Drymedia preparationLoosely packed (LP) dry GM preparations (ϕ = 0.58 ±0.01) were created using an air-fluidized bed with anarea of 22.9 ×40.6 cm2 and filled with spherical glassparticles to a height of 10 cm ([2, 3]). The particleswere similar in size (diameter = 0.27 ± 0.04 mm) andweight (density = 2.5 g cm−3) to sand found in nature.An evenly distributed air flow was forced upwardthrough the grains; below a critical flow rate grainsremained stationary, and above this flow rate thegrains moved relative to each other (which is referredto as fluidization). To generate the LP states, themedium was fluidized followed by a slow decrease inflow rate to zero. Air flow was off during forcemeasurements and animal burial trials.

2.2.Wetmedia preparation2.2.1.WetmediamixtureA high-speed hand mixer (Hamilton Beach/Proctor-Silex, Southern Pines, NC) was used to combine 14 kgof dry spherical glass particles (diameter = 0.27 ±0.04 mm, density = 2.5 g cm−3) with 0.14, 0.42 and0.7 kg of water, such that W 0.01,= 0.3, and 0.05water content preparations were achieved, whereW isdefined as the ratio of the mass of the water to mass ofthe dry GM. The media were mixed for at least 3 min

to distribute water evenly throughout the preparation.Media preparation and experimentation took place ina room with a humidity level of ≈41% and tempera-ture of ≈25 °C. The wet mediums were used inexperiments for only two hours after preparation tolimit changes in water content due to evaporation. Wecharacterized water loss due to evaporation by prepar-ing W 0.01, 0.03, 0.05= mixtures and weighing thesubstrates periodically. The substrates were stirredonce every hour. These experiments revealed that Wdecreased by 0.0014 ± 0.0003 after two hours for allstartingW preparations.

2.2.2. Apparatus developmentA ‘sieve’ apparatus was constructed to deposit mediahomogenously into a testing container (figure 2, see SIVideo 2 available at http://stacks.iop.org/pb/12/046009/mmedia). The sieve apparatus consisted of twopolycarbonate boxes (30.5 ×30.5 ×19.5 cm3) that werestacked vertically, a holding plate, and a vertical shaker.The top box was separated from the bottom containerby a stainless steel woven mesh grid (wirediameter = 1.2 mm, opening area of 3 ×3mm2). Theprepared wet media was poured into the top containerof the sieve apparatus. Cohesion between grains pre-vented the media from falling through the mesh whenthe apparatus was stationary. The stacked boxes fit intoa holding plate and were secured using straps whichprevented relative motion between the boxes and plateduring shaking (figure 2(A)). Sinusoidal vertical vibra-tions, with a frequency of 60 Hz and average amplitude

Figure 1.TheOcellated skink (Chalcides ocellatus) is a habitat generalist andmoves on andwithin bothwet and drymedia. (A) TheOcellated skink in nature (obtained fromHerpetofauna of Europe, http://hylawerkgroep.be/jeroen/index.php?id=38) (B) side view oftheOcellated skink head and forelimb. (C)Map showingwhere theOcellated skink has been found (blue regions) which includenorthernAfrica, southern Europe, andwestern Asia (modified from [27]).

3

Phys. Biol. 12 (2015) 046009 S S Sharpe et al

http://stacks.iop.org/pb/12/046009/mmediahttp://stacks.iop.org/pb/12/046009/mmediahttp://hylawerkgroep.be/jeroen/index.php?id=38

of 1.75mm, were applied to the holding plate via theelectromagnetic shaker (VTS, Aurora, OH). Theseparameters were selected due to our observations thatmedia readily fell through the mesh when these vibra-tionparameterswere used (figure 2(B), and see SIVideo2 stacks.iop.org/pb/12/046009/mmedia). The shakerwas controlled using custom software (LabVIEW,National Instruments, Austin, TX) and a power ampli-fier (Model # 7550, Techron, Elkhart, IN). Vibrationswere monitored using an accelerometer (PCB

Piezotronics, Depew, NY) that was attached to theholding plate. The software monitored the shaking andmaintained the desired shaking parameters using a PIDcontroller. Changing the amplitude and/or frequency ofthe vibrations changed the efficacy of inducing mediadeposition. Analysis of how these parameters affect thefinal medium compaction is beyond the scope of thepresent study.

Shaking ceased after the sieved material in thelower container exceeded a height of 7 cm. A slit and

Figure 2.Wetmedia sieve apparatus. (A) Schematic of shaking apparatuswith thewetmedia loaded into top container prior toshaking. (B) Agglomerates of wet grains fall through themesh due to vertical vibrations and accumulate in the lower container.(C) Leveled preparation used for experiments, with thefinal volume of 30.5 ×30.5 ×6.9 cm3. Comparison showing (D) x-ray of awetpreparation (W 0.01= ) that wasmixed then dropped onto a hard surface to clear large voids and (E) a preparation (ϕ=0.575,W 0.01= ) inwhich the sieve techniquewas used. ((F) and (G)) Pixel intensity (on a 0–255 grayscale) across a randomly chosenhorizontal line (dashed yellow line) is shown below each corresponding x-ray image ((D) and (E), respectively). Higher intensities(appearingmorewhite) indicate areas containing voids and darker areas (appearingmore black) have a higher concentration ofmedia.

4

Phys. Biol. 12 (2015) 046009 S S Sharpe et al

http://stacks.iop.org/pb/12/046009/mmedia

groove located at 7.06 ± 0.16 cm from the base in thelower box allowed a thin aluminum plate to slide intothe container separating the GM into two sections.The top media above the thin plate was removed leav-ing a volume of wet media (30.5 ×30.5 ×6.9 cm3) witha level surface which was used for experiments(figure 2(C)). The volume fraction,ϕ, was determinedby weighing the wet media. Due to the leveling techni-que, the volume of the space occupied remained con-stant across trials. The volume of the dry GM was

determined from the relationVGM=M

W(1 )wet

GMρ +, where

GMρ is the density of the GM, Mwet is the mass of thewetmedia andW is the water content. Due to the errorin the volume and W measurements, we estimate asystematic error in theϕ calculation of ± 0.007.

2.2.3. Technique characterization and capabilitiesHomogeneity—we evaluated the homogeneity of asieved wet media preparation (W = 0.01, ϕ = 0.575)using x-ray imaging (source: OEC-9000, RadiologicalImaging, Hamburg, PA, USA; detector: Varian flatpanel 25250 V Csl, Palo Alto, CA, USA) (figures 2(D)and (E)). The x-ray was set to an energy of 20 mA and60 kV. Variations within the substrate were detectedby examining the pixel intensity (on a 0–255 grayscale)where higher intensities (appearing more white)indicate areas containing voids and darker areas(appearingmore black) have a higher concentration ofmedia. The sieved preparation was compared to wetmedia (W = 0.01) that was mixed then compacted by‘hand’ by dropping the media container from a heightbetween 5 and 8 cmmultiple times onto a hard surfaceto remove large voids. Fluctuations in the local densityhand prepared states were larger ( 50≈ grayscales,figures 2(D) and (F)) compared to fluctuations in thesieve preparation ( 10≈ grayscales, figures 2(E)and (G)).

Compaction—it took at least 70 s of shaking todeposit 7 cm (in height) of media in the bottom con-tainer. The average time for all wet media ( 14≈ kg) tofall through the mesh increased with wetness content(188.5 ± 62.6 s for W 0.01= , 287.9 ± 84.8 s forW 0.03= , and 316.4 ± 119.7 s forW= 0.05). The totalshaking time was varied between 70 and 2250 s toachieve varying compactions.

2.3. ForcemeasurementsTo measure penetration resistance and drag force inwet and dry GM, a robotic arm (CRS robotics,Burlington, Ontario, Canada) with 6 degrees of free-dom was used to drive a stainless steel cylindrical rod(diameter = 1.6 cm and length = 3.81 cm) through themedia. A support rod (diameter = 0.63 cm) and nut(height = 0.55 cm, corner-to-corner width = 1.25 cm)were used to attach the cylindrical intruder to a forcesensor (ATI industrial, Apex, NC, USA), accurate to0.06 N, located on the robot end effector. The cylinderwas submerged 3.2 cm into the GM (measured from

its center axis to the surface of the substrate), and thecylinder was dragged 12.7 cm with its long axisoriented perpendicular to the direction of motion(figure 3(A)). Speed during penetration and drag was1 cm s−1 (for intruders of this size force is insensitive tospeed in dry GM below 50≈ cm s−1 [2, 3] andpreliminary studies indicate the same force insensitiv-ity in wet media). The total projected area of thecylinder, submerged support rod and nut during dragwas 8.4 cm2. The instantaneous force periodicallyfluctuated due to material fracturing that occurredduring movement within the substrate (see results,figure 3(B)); consequently, the average drag force wascalculated during steady state movement from thestart of one oscillation to the start of another oscilla-tion. Although we did not investigate the fluctuationsin the plowed wet GM, it would be interesting tocompare their dynamics to the force fluctuationsobserved in compact dryGMdescribed in [31].

To quantify the average drag force as a function ofdepth on an intruder with the same projected area(6.1 cm2), the drag force on the support rod and nutwas subtracted from the drag force of the cylinder +support rod and nut. Drag force as a function of depthwas quantified in dry media ( 0.58ϕ ≈ ) and inW 0.03= wetmedia (53.2 55.7ϕ⩽ ⩽ ).

2.4.Ocellated skink experimentsOcellated skinks, Chalcides ocellatus, (figure 1) werepurchased from commercial vendors (East Bay Vivar-ium, Berkeley, CA, USA and Ocean Pro Aquatics,Chino Hills, CA, USA). The four animals used in thisstudy had an average snout-vent length (SVL) of 10.9± 1.3 cm, snout-tail length of 18.6 ± 3.2 cm and massof 21.5 ± 6.8 g. The ocellated skinkʼs body height is 1.3± 0.2 cm andwidth is 1.8 ± 0.3 cm (N= 4 animals). Allanimals were housed individually in large containers(21 ×43 ×28 cm3) filled with moist sand to a depth of15 cm and were provided with a water dish and mossfor hiding. Ocellated skinks were given sixmealwormscoated in a supplemental calcium powder twice a weekand allowed to eat ad libitum. The holding room wasmaintained on a 12 h:12 h light:dark cycle. All experi-mental procedures were conducted in accordancewiththe Georgia Institute of Technology IACUC protocolnumbers (A08012, A11066) and Radiation Safetyprotocol (X-272).

We compared the burial strategy of the Ocellatedskink on LP (ϕ = 0.56 ± 0.01) wet GM (W 0.03= ) tothe strategy used on LP dry GM (ϕ = 0.58 ± 0.01). LPwet media was used because Ocellated skinks did notbury as readily into closely packed (CP) wet media.Ocellated skinks were placed on top of the media pre-paration inside of a 14.5 cm diameter hollow plasticcylinder resting on the surface of the substrate whichinduced burial at a desired location. For enhancedcontrast, a minimum of 11 lead markers (≈ 1 mm2

each, mass ≈ 0.01 g) were placed on the skinkʼs dorsal

5

Phys. Biol. 12 (2015) 046009 S S Sharpe et al

midline at 0.1 SVL increments and one marker wasplaced on each limb. Above-surface visible light videowas recorded at 63 frames per second (fps) using ahigh speed camera (AOS Technologies AG X-PRI,

Baden Daettwil, Switzerland) to characterize generalfeatures of above surface movement such as time toburial and limb use. Top-view subsurface kinematicswere recorded using an x-ray system (OEC 9000,Radiological Imaging Systems, Hamburg, PA, USA)coupled to a high speed camera (Fastcam 1024 PCI,Photron, San Diego, CA, USA) which recorded at60 fps. The different above surface and subsurfacevideo frame rates were set due to camera constraints.For dry GM preparations, the x-ray system was set to85 kV at 20 mA, and for wet preparations, to 73 kVand 20 mA. Ocellated skink burial time was deter-mined from above surface video as the time it tookfor the animal to submerge such that the head to ventwas covered by the substrate. When the anterior por-tion of the animal was buried prior to the start of theabove surface recording, then the time to burial wascalculated as the time it takes for the above surfacefraction of the SVL to submerge divided by the pro-portion of the SVL above surface at the beginning ofthe recording.

To determine angle of entry and final depth ofburial in wet and dry GM, biplanar x-ray videos wereacquired. TheOEC-9000 x-ray systemwas again usedto record top view images. An additional x-ray system(source: Spellman XRB502 Monoblock, Hauppauge,NY, USA; flat panel detector: Varian 25250 V Csl,Palo Alto, CA) was oriented 90° with respect to thetop view x-ray such that side view images of the sub-surface locomotion were simultaneously obtained.The flat panel detector captured images at 30 fps. Thewidth of the wet GM container was decreased to

17≈ cm by using foam inserts. Decreasing the widthwas necessary to enhance contrast between the Ocel-lated skink and surrounding media during side viewimaging. Opaque lead markers were placed on thedorsal midline, and along the animalʼs side closer tothe ventral surface to enable visualization of the Ocel-lated skink in both top and side view images. 3Dreconstructions were made using custom software(Matlab, Mathworks, Natick, MA, USA). Depthof burial was assessed by measuring the distancefrom the substrate surface to the ventral midpoint(≈ 0.5–0.6 SVL) along the body at the animals finalresting position.

3. Results

3.1. Penetration and drag forces in awet substrateInstantaneous force during drag (figure 3(B)) andpenetration (figure 3(C)) revealed a large increase(≈ 4 times greater) in force between dry and wet GMpreparations. During penetration, force increasedwith depth (figure 3(C)) for both dry and wetsubstrates. During withdrawal, there was a slightnegative force in all substrates. We attribute thisnegative force to the weight of material above the rodand the shearing forces that occur between the grains.

Figure 3.Rod drag and penetration force reaction inGMwithvaryingwater contents. (A) Forcewas obtained by dragging astainless steel cylinder (diameter = 1.6 cm, length= 3.81 cm)and a support rod (diameter = 0.64 cm) at a speed of 1 cm s−1

and at a depth of 3.18 cmmeasured from the center of thecylinder to the surface. (B) Representative instantaneous dragforce as function of distance at varyingwater contents (orangeisW=0 (i.e. dry), red is W 0.01= , green is W 0.03= andblue is W 0.05= ). Compactions in these representative trialsrange betweenϕ=0.55–0.59. (C) Representative instanta-neous vertical force on the cylinder during penetration as afunction of depth. Zero depth occurs when the cylindricalintruderʼs center is level with the top of theGM(i.e. 1/2diameter has penetrated into themedia).

6

Phys. Biol. 12 (2015) 046009 S S Sharpe et al

Force fluctuations occurred in wet substrates duringpenetration and drag which corresponded to materialfracturing visible at the surface. These large forcefluctuations resemble shear band formation andfracturing during drag in CP dry media [31]. Themechanism by which these bands occur in wet mediahas not been investigated and is beyond the scope ofthis work. In agreement with previous studies [2, 31],these large force fluctuations were not present duringdrag in LP drymedia.

Average drag in W 0.01= wet media (ϕ = 0.58)was 4≈ times larger than the average drag force in drymedia (0W,ϕ= 0.58; figure 4). The average force con-tinued to increase as W increased from 0.01 to 0.05(ANCOVA, F P(5, 59) 32.0,= < 0.001), but by asmaller amount ( 1≈ N per 0.01 increase inW). Resis-tance forces increased with compaction inW 0.01, 0.03, 0.05= preparations (figure 4(B),ANCOVA, F P(5, 59) 7.5,= < 0.01). Force changeswere large with small changes in ϕ; for example, forceincreased by 50% in W 0.03= media from approxi-mately 8.0 N atϕ=0.53 to 11.8 N atϕ=0.58.

3.2.Ocellated skink burial inwet and drymediaAfter being placed on the W 0.03= wet media,Ocellated skinks took between 1 and 30 min beforeinitiating burial. Lightly squeezing or gently tappingthe animals’ tails occasionally induced burial morequickly. Examples of X-ray and above surface imagesduring Ocellated skink burial are shown in figure 5.Ocellated skinks rarely buried into substrate prepara-tions with 0.58ϕ > and so LP preparations of wetGM (ϕ ⩽ 0.56) were used in all experimental trials.Initiation of burial took less time in dry GM (between1 and 15 min).

After initiation, Ocellated skink burial was slowrelative to sandfish burial (see SI Video 1 available athttp://stacks.iop.org/pb/12/046009/mmedia). Burialtime was moderately smaller in dry media (22.9 ±13.4 s) compared to wet media (29.4 ± 9 s, ANOVA,F P(7, 36) 4.5, 0.05= < ) and differed slightlyamong animals tested (ANOVA, F (7, 36) 3.0= ,N = 4 animals). This slow burial was in part a con-sequence of the start-stop locomotion strategy that theOcellated skink used during burial (figure 6).Figures 6(C) and (D) show the movement of the 5thmarker (at the 0.4 SVL location) on the Ocellatedskink for a representative trial in wet and dry media,respectively, where the displacement is a measure ofthe total distance along the path length from thebeginning of the trial. The plateaus (pink regions)indicate pauses while the green regions correspond toprogression. Ocellated skinks moved for an average of0.6 s in both wet and dry media before pausing (N = 4animals, n = 5–7 trials each); histograms of the move-ment and stopping times are shown in figures 6(E)–(H). The average subsurface speed of the 0.4 SVL loca-tion during movement was higher in dry media (0.21

± 0.05 SVL s−1; 2.4 ± 0.5 cm s−1) compared to in wetmedia (0.11 ± 0.03 SVL s−1; 1.1 ± 0.2 cm s−1, T-test,P 0.01< ).

Most Ocellated skinks placed their hindlimbsnear their sides prior to limb submergence (see SIVideo 1 available at http://stacks.iop.org/pb/12/046009/mmedia). This occurred when 0.8 ± 0.2 ofthe SVL was submerged in the media for both wetand dry substrates and across all animals tested(ANOVA, P 0.05> ). In both wet and dry media,Ocellated skinks had a curved body posture duringburial resembling a serpentine curve with 1.5≈waves along the body and the subsurface slip waslow (i.e. all body locations followed a similartrajectory, figures 6(A) and (B), see SI Video 3available at http://stacks.iop.org/pb/12/046009/mmedia). While the hindlimbs were placed by theanimalʼs sides before burial, the skink continued touse its forelimbs during subsurface locomotion.Protraction and retraction of the forelimb on theconvex side of the body was observed during for-ward progression and the forelimb on the concaveside was held near the body (see SI Video 3 availableat http://stacks.iop.org/pb/12/046009/mmedia).

The head rotated from side to side (oscillated)during forward locomotion at ≈2–4 Hz (see SIVideo 3 available at http://stacks.iop.org/pb/12/046009/mmedia). Figure 7(A) shows the trajectoryof markers along the body. The side to side motion isclearly visible in the trajectory of the snout marker(blue trajectory), while other markers follow alongits mean path. There was a lower amplitude devia-tion in the 2nd and 3rd marker as well, indicatingthat the pivot point was between the 0.2 and 0.3 SVLlocations. The angle of excursion about the pivotpoint was estimated by low pass time-filtering, theposition of the snout (1st) marker, using a first orderButterworth filter and 15 Hz cut-off frequency, andcomparing the vector between the filtered 1st mar-ker position and the 4th marker position with thevector between the actual 1st marker position and4th marker position (figure 7(B)). The maximumangle of excursion in dry media (7.6° ± 2.3°, n = 10trials, N = 3 animals) was higher than the maximumangle in wet media (4.5° ± 1.3°, n = 9 trials, N = 3animals; ANOVA, F P(2, 18) 10.3, 0.01= < ,figure 7(C)) and did not statistically vary acrossanimals.

There was a substantial difference in the burrow-ing angle between wet and dry substrates (see SIVideo 4 available at http://stacks.iop.org/pb/12/046009/mmedia). Ocellated skinks in dry media bur-rowed at an average angle relative to the horizontalsurface of 31.6° ± 8.6°, while in wet media the anglewas smaller 8.0° ± 3.8° (figure 8(C)). The averagedepth of burial in dry media (5.0 ± 1.5 cm) was overthree times as large as the average depth of burial intowetmedia (1.6 ± 0.5 cm;figure 8(B)).

7

Phys. Biol. 12 (2015) 046009 S S Sharpe et al

http://stacks.iop.org/pb/12/046009/mmediahttp://stacks.iop.org/pb/12/046009/mmediahttp://stacks.iop.org/pb/12/046009/mmediahttp://stacks.iop.org/pb/12/046009/mmediahttp://stacks.iop.org/pb/12/046009/mmediahttp://stacks.iop.org/pb/12/046009/mmediahttp://stacks.iop.org/pb/12/046009/mmediahttp://stacks.iop.org/pb/12/046009/mmediahttp://stacks.iop.org/pb/12/046009/mmediahttp://stacks.iop.org/pb/12/046009/mmedia

4.Discussion

4.1. Preparation of states of wetGMThe new sieving method created large homogenouspreparations of wet GM (i.e. no large voids) withvariable compactions. Low ϕ states were achieved bystopping the vibrations soon after media deposition,while high ϕ states were achieved by prolongedshaking. Depending on shaking time, compactionvaried between ϕ = 0.53–0.60. As expected due to thestabilizing effect of liquid bridges, the minimum ϕobtained in wet media was lower than theminimumϕachieved in dry media (ϕ = 0.58) using the air-fluidization technique. We expect that lower compac-tions (ϕ < 0.53) could be achieved by shaking the topcontainer only; in our system, both the top and bottomcontainers were shaken simultaneously during thedeposition process. Therefore, the media that wasdeposited into the lower container near the onset ofshaking may be at a slightly higher compaction.However, we were unable to distinguish a compactiongradient within the material using x-ray imaging.Applying vertical vibrations to the top container alone

would help to remove the compaction gradient, if itexists, and create an overall lower ϕ. The highestcompaction achieved (≈0.6) occurred after 40 min ofshaking. For all wet media preparations, the compac-tion of the media increased logarithmically withshaking time. Fiscina et al [37] found that ϕ as large as0.62 could be achieved in water–GM mixtures. Sincewe did not observe an asymptote in compaction withincreased shaking time, we expect that larger compac-tion preparations are possible using our setup forshaking time >40min. However, for animal experi-mentation this might be impractical since the mediapreparation timewould be inconvenient.

4.2. Resistance forcesWe found that penetration and drag resistance forcesincreased with wetness (for example, drag forcesincreased 400% from dry to W 0.01= media) andwith compaction (drag forces increased 50% betweenLP and CP in W 0.03= media). These resultsemphasize the importance of controlling and report-ing both W and ϕ when performing animal androbotic locomotion studies in wet GM since changes

Figure 4.Average resistance forcemeasurements at varyingwater contents and compactions. (A) Resistance force sharply increased asW increased fromW=0 to W 0.005= , then increasedmore gradually. For a givenW, a highϕ state (dark red) had a 50%higherresistance force compared to the lowϕ state (dark blue). (B) Average drag forcewith changing compaction forW=0.01 (red), 0.03(green) and 0.05 (blue). The dashed lines show the best linearfits to the data.

8

Phys. Biol. 12 (2015) 046009 S S Sharpe et al

in resistance force could influence locomotion strategyand/or kinematics. The average drag force in the wetmedia preparations displayed larger standard devia-tions among different trials compared to the dry forcemeasurements. We attribute this deviation to the largeforce fluctuations during drag in wet media. Althoughwe measured mean force over one fluctuation period,changes in the fluctuations (such as the frequency and/or minima) may have skewed mean values (seefigure 3(B)). A more in-depth study which examinesthese force fluctuations with water content could leadto more consistent results and reveal principles of theresponse of wet GM to localized intrusions.

In dry GM at the speeds the animal operates (a fewcm s−1) drag forces are insensitive to drag speed[2, 41]. We found similar results in slightly wet GM(W 0.1< ): rod drag experiments at different speeds(1, 5, and 10 cm s−1) also revealed that force was insen-sitive to speed (see supplementary material, figure S1stacks.iop.org/pb/12/046009/mmedia). We note thatthese experiments were conducted in wet sand pre-parations that were create via hand-mixing and there-fore compaction was not controlled. However, sincecompaction does not increase speed sensitivity in dryGM, we expect that for a given W and compaction,drag forces will remain insensitive to speed. However,we suggest that future studies should test thishypothesis.

We comment briefly on the applicability of ourrod drag experiments to infer properties of drag on theanimalʼs body. Although the stainless steel cylindricalintruder used to estimate resistance force has different

properties than the Ocellated skinkʼs body, we expectthat for fixed body properties the trend of increasingforces with water content can be extended to the forcesexperienced by the lizard. The diameter of the rod iscomparable to the body diameter of the Ocellatedskink. Furthermore, we have shown that in dry mate-rial, force scales with the projected area of the intruder.While theOcellated skinkʼs skin-particle friction (0.15± 0.02 on the ventral surface in the forward directionand 0.13 ± 0.01 on the dorsal surface in the forwarddirection, see [4, 42] for discussion of the technique) isabout half that of the rod-particle friction, normaldrag force is independent of the friction of the rod.The rod was moved in the GM with its long axis nor-mal to the direction of travel in order to maximizenormal drag force and minimize tangential forcewhich scales with friction. Finally, we expect wateradhesion properties to be similar enough between therod and Ocellated skinkʼs body such that force char-acteristics are the same. Skin wettability for lizards thatinhabit arid regions have a range of adhesion proper-ties but all tend to have water contact angle

measured normal and tangential forces on a small rodmoving through GM, assumes steady-state motion ofthe animals. However, because the Ocellated skinkemploys a start-start motion we did not attempt to

apply RFT to understand the forces imparted on theskink’s body during movement through dry GM.Future developments of RFT in dry GM shouldaccount for such transient phenomena (which are

Figure 6. Subsurface bodymidlines during burial in (A)wetmedia and (B) drymedia. The color of the curves correspond to the timein seconds, where dark blue is time= 0 s and dark red is time= 33 s in (A) and time= 30 s in (B). Thewhite region indicates theportion of the body that is above surface and the pink shows the subsurface region. (C) and (D) Total displacement of the 0.4 SVLmarker on theOcellated skink during burial intowet and drymedia, respectively, displaying start-stopmotion. Pink horizontal barsindicate the duration that theOcellated skinkwas stationary and green regions indicatemovement. (E) and (F)Distributions of theOcellated skinkmovement times inwetmedia and drymedia, respectively. Themean andmedianmove time inwetmedia is 0.6 and0.3 s respectively. Themean andmedianmove time in drymedia is 0.6 and 0.4 s, respectively. (G) and (H)Distributions of timesbetweenmovement intervals inwetmedia and drymedia, respectively. Themean andmedian stop timewas 1.6 and 1.1 s, respectively,in wetmedia and 2.2 and 1.2 s, respectively, in drymedia.

10

Phys. Biol. 12 (2015) 046009 S S Sharpe et al

known to lead to inaccuracies in RFT in larger particles[15]) to allow confident application of RFT to suchlocomotor modes. In slightly wet GM like that studiedhere, we are not aware of similarly predictive models;development of a theoretical framework to accountfor resistive forces on intruders (broadening the dragmeasurements presented here) will be critical toexplain the biological and physical observations.

4.3. Burial behaviorThe use of a homogenous wet GM preparation withcontrollable compaction allowed us to discover thatOcellated skinks bury into ground using periodic bodybending movements, the forelimb on the convex side,and head oscillation. Although burial speeds are muchslower than some specialized burrowers like thesandfish ( 0.5≈ s for the sandfish [44] compared to20–30 s for the Ocellated skink) the general burialstrategy enabled the Ocellated skink to burrow intoboth wet and dry media preparations. While thesandfish burial and swimming strategy is effective indrymedia, sandfish in captivity either avoid wetmedia(as in nature) or use a limbed digging strategy todisplace the wet sand (from personal observations).This may indicate that in wet sand the sandfishswimming strategy is ineffective. We expect that thisburial capability is vital for the Ocellated skinkʼssurvival in nature, because it allows thermoregulation

and conceals the skink from other animals. Attum et al[40] reported thatOcellated skinks weremore inclinedto run toward vegetation to escape predators in thewildwhereas sandfish preferred burial intoGMduringescape. This difference in preference may be due toOcellated skinkʼs inability to rapidly submerge. Also,we noticed during kinematic recordings in which themedia had been disturbed and large voids werepresent, the Ocellated skink tended to move towardthe voids during subsurface locomotion. Althoughthis was not investigated in detail, these preliminaryfindings could indicate that the head oscillation maybe used to sense less resistance and attract theOcellated skink to that area. In nature, the Ocellatedskinks may use existing void spaces which exist in wetsubstrates to decrease resistance force acting on theirbodies and improve performance. The Ocellatedskink more readily buries into LP wet substrate thanCP which may be due to the higher resistance forcesin CP media. This may also indicate that in nature,Ocellated skinks preferentially bury into loosersubstrates.

This study has revealed an interesting pattern ofsubsurface locomotion which is different from thesandfish lizard. The Ocellated skink uses a start-stopmotion, head oscillations, and forelimbs to move sub-surface with low slip. The function of the head oscilla-tion is still unknown. The head oscillation could serve

Figure 7.Head lateral excursion duringOcellated skink burial. (A) Trajectories ofmarkers along the body during subsurfacemovement inwetmedia. (B) Angle of excursion of the headwas calculated by comparing the low pass time-filtered snout positions tothemaximum snout positions (i.e. angle of excursionwould bemeasured frommaximumexcursion at t1 and t2 to thefilteredposition, see text for details). (C) The averagemaximumangle of excursion duringmovement in drymedia is greater than inwetmedia.

11

Phys. Biol. 12 (2015) 046009 S S Sharpe et al

as (1) a force sensing mechanism (as described above)to steer motion toward low resistance regions, (2) apropulsion mechanism in which to push against thesurrounding substrate, and or (3) a force reduction

mechanism to ‘crack’ GM and reduce resistance tointrusion similar to the activity of the annelid wormNereis virens which burrows into mud by extendingcracks [45].

Figure 8.Difference in entry angle and depth duringmovement inwet and drymedia. (A)Diagrammatic representation of thecharacteristic orientations of theOcellated skinks in dry andwetmedia, where burial depthwasmeasured from the surface of thesubstrate to the ventralmidpoint of the animal and burial angle ismeasured relative to the horizontal. (B)Ocellated skinks buried toan average depth of 5.0 cm in dry (green)media and 1.6 cm inwet (blue)media. (C) Burrowing angle was on average 31.6° in drymedia (green) and 8.0° inwet (blue). (D)Average drag force (n=3 trials per point) on a cylinder in LP drymedia (green) andW 0.03= media (blue). Themean compaction in thewet substrates was 0.549 ± 0.006.Note that the standard deviation bars arewithin themarker size for drymedia. Dashed lines show the best linearfits to the dry (slope = 0.83 N cm−1) andwet(slope = 1.47 N cm−1) data. Vertical dashed lines show the average burial depth in experiment and±1 standard deviations are shownby the colored regions forwet (blue) and dry (green)media. Blue and green horizontal lines show the estimated drag force around themidpoint of the body at thefinal average and estimated (using 1 std)maximal burial depths inwet and drymedia, respectively.

12

Phys. Biol. 12 (2015) 046009 S S Sharpe et al

To investigate whether the head oscillation couldreduce resistance force during forward progression,we performed a simple experiment in which we sub-merged a cylindrical intruder into GM and rotated theintruder by +20° and−20° prior to dragging it throughthe granular substrate (see figure 9(B); n = 3 trials). Inthese experiments, the rod started andmoved forwardwith its long axis oriented parallel to the direction ofmovement to more closely resemble the movement ofa head, although we note that both the shape andmovement pattern were not intended to exactly repli-cate that of the Ocellated skinkʼs movement. Wefound that in wet media, a rotation before dragreduced the yield force by ≈45% at the initial draginstant (figure 9(C)) in comparison to drag withoutthe rotation (figures 9(A) and (C)) and approachedthe non-rotation drag force with increasing drag dis-tance. After 5 cm of horizontal movement, the forcesin the rotated condition matched the forces in thenon-rotated condition. In dry media, the forces weresimilar throughout the entire horizontal drag in bothconditions, indicating that the rotationwas not advan-tageous. Unlike these simple experiments, the headoscillation in the Ocellated skink occurs during for-ward locomotion. While the oscillation could still actto reduce forces in wet media prior to the next move-ment, a more intriguing possibility is that the headcould help to pull the animal forward. Future

computational and theoretical modeling work will becritical to reveal the purpose of the head oscillationduring forward motion. It is interesting to note thatthe maximum head angle of excursion is higher in drymedia compared to wet (even though there is littleforce reduction in a dry substrate, figure 9(D)). Weattribute the higher angle to the lower resistance forcein dry media. This may imply that the kinematic pat-tern is an open loop strategy in which the animal gen-erates a constant torque at the head and kinematicschange based on substrate resistance.

Ocellated skinks possess elongated bodies andreduced limbs in comparison to the sandfish. The fore-limb movement on the convex side of the body revealsthat limbs aid in propulsion subsurface. We receivedone animal that wasmissing nearly all of one of its fore-limbs. In subsurfacemovement tests in dry GM (poppyseeds) we found that the animal was still able to moveforward when the side with the missing limb was con-vex (i.e. no limb movements occurred during this per-iod). The animal was able to move forward using onlyhead oscillation and undulatory bodymotion. Thismayindicate that although these limbs aid in forward pro-pulsion, head oscillation and body bending are suffi-cient to produce forward thrust subsurface. Morerigorous experiments which constrain limbs duringsubsurface locomotion in dry and wet substrates areneeded to further investigate the role of limbs.

Figure 9.Exploring force reduction by rotation prior to forward drag. (A) In the normal drag condition, a cylindrical rod(diameter = 1.6 cm, length= 3.81 cm)moved 15.2 cm inGMsubmerged to a depth of 3.81 cm. The rodwas oriented such that its longaxis was parallel to the direction of drag. (B) In the rotated condition, the cylinder rotated about the support rod clockwise 20° (step 1in (B)), then rotated 40° counter-clockwise (step 2), then 20° clockwise to the original orientation. The cylinder was thenmovedforward 15.2 cm (step 3). (C) The force inwetmedia for the normal drag condition (black) and rotated condition (gray) for arepresentative trial. Yield forcewas reduced by 45% for the rotated condition and themean drag forces were similar after ≈ 5 cmofmovement. (D)Drag force in LPGM for the normal drag condition (black) and rotated condition (gray). The reduction in yield forcewasminimal < 15%. The vertical dashed lines in (C) and (D) show the approximate distances inwhich the drag force in the rotatedcondition becomes similar to the non-rotated condition.

13

Phys. Biol. 12 (2015) 046009 S S Sharpe et al

4.4. Burial limited by resistive forceDuring burial, the Ocellated skink employed a loco-motion strategy inwhich cyclic body bendingwas usedto enter the substrate, a strategy that enabled burialinto both wet and dry substrates. However, theOcellated skink remained near the surface of thesubstrate during burial into LP wet media. Weattribute the lower angle and burial depth to the higherresistance forces in wet GM (figure 8(D)). Empiricallymeasured resistance forces on a cylindrical intruderwere used to estimate the forces on the skinkʼs body atvarying depths. At the Ocellated skinkʼs final burialdepth (defined as the depth of the ventral midpoint)resistance forces were 4.5 N in wet media and 4.1 N indry media; at the burial depth extremes (+1 standarddeviation) resistance forces were identical: 5.2 N inboth wet and dry media. To estimate the peak forceson the animal (at the head), using the average bodylength and average angle of entry inwet and drymedia,we calculated the depth of submersion of the head inwet and dry media and forces near the snout (asmeasured in rod drag experiments). In wet media, theresistance forces were lower (on average 6.7 N) thanthe resistance forces at the final calculated snout depthin dry media (on average 8.8 N). These measurementsallow us to estimate that the Ocellated skink is capableof generating forces that are 40 times greater than itsbody weight, well within the range expected fromanimals of its size [46].

The arguments above thus indicate that the Ocel-lated skink is limited in its burial by the inability of itstrunk (and possibly limb) muscles to conquer thebasic resistive properties of the GM (with possiblereduction in drag through head oscillation). It is aninteresting question for further research to determineif similar limits to burial performance are seen in lim-bed and limbless squamates with similar body plans tothe Ocellated skink (e.g. see examples in [4]). In fact,other (non-squamate) animals have evolved mechan-isms to bury to depthswithin substrates in which resis-tance forces are clearly greater than forces the animalcan generate; suchmethods include crack propagationby the annelid worm Nereis virens [45], fluidizationand anchor method used by the Atlantic razor clam(Ensis directus) [25], and substrate ingestion by theearthworm (Aporrectodea rosea) [47]. Regarding thelatter, the razor clam can produce 10≈ N of force [48]to pull itself into saturated GM which should onlyallow burial into a few centimeters in saturated GM;however, this animal can submerge to depths greaterthan 70 cm [25]. Winter et al [25] found that itsunique valve movements fluidize the surroundingmedia allowing it to bury deeply while reducing theenergy required to reach large depths. As anotherexample, measurements on certain earthworms(Aporrectodea rosea) revealed these animals can gen-erate a maximum force of 3≈ N using longitudinalmuscles in any one segment [47], but can submergedeep within the ground (≈ 50cm [49]). To cope with

larger forces at larger depths or in compacted soils,earthworms may switch burial strategy and ingest soiltomove through themedium.

5. Conclusions

In conclusion, we have developed a new technique tocreate relatively large controlled preparations of wetgranular media. We used this to make the firstmeasurements of the movement pattern of a generalistlizard burying within such substrates. The ability tocontrol properties and generate repeatable states thatwere homogeneous allowed us to monitor the burialprocess. The Ocellated skinkʼs movement patterndiffered from the sand-specialist sandfish lizard (theonly other lizard studied at this level of detail) in drygranularmedia (the only substrate inwhich the sandfishrapidly buries). And while for the Ocellated skink therewere some differences in burial kinematics (e.g. headoscillation amplitude, speed, and burial entry angle)between dry and wet media, the same general locomo-tion strategy was employed in both conditions. Wetherefore hypothesize that the Ocellated skink uses asimilar kinematic burial pattern, or ‘template’ (see [50]which defines a template as a target of neuromechanicalcontrol). In this case, resistive force properties of thesubstrate lead to differences in burial depth. To test for atemplate, future studies could perform electromyogra-phy as the animal is subjected to different substrateconditions; this approach was used in studies of atemplate for rapid burial in dryGM[3, 51].

However, our previous discoveries of templates indry sand-swimming were only made possible by mod-els of intrusion and drag in GM, which unlike truefluids, have no validated equations of motion thatdescribe all scenarios. In those studies RFT (see forexample [52]) and multi-particle discrete elementmethod simulations (see for example [15]) revealedthat by following a template, the sandfish could max-imize swimming speed and minimize energy use[4, 16]. Therefore we argue that better models of wetGM during intrusion are needed to aid in modelingthe locomotion of the vast diversity of animals thatbury. Perhaps the approach using soil mechanics [25]could be of use to develop a more general ‘terrady-namics’ (see [14] for discussion). Our sieving prepara-tion technique should be useful in constructingmodels of intrusion: the homogeneity and repeat-ability will aid in direct comparison with computa-tional and theoretical models which can be used togenerate hypotheses and increase understanding ofbiological principles.

Acknowledgments

We thank Tamunodiepriye George who began thekinematic work with the Ocellated skinks, KevinDaffon for assistance with investigation of techniques

14

Phys. Biol. 12 (2015) 046009 S S Sharpe et al

to prepare repeatable wet granular states, and AndrasKarsai for his help in data analysis.We also thank FeifeiQian and Kaitlin Judy for assistance in collectingpreliminary wet granular media force data. This workwas supported by The Burroughs Wellcome FundCareer Award at the Scientific Interface, NSF Physicsof Living Systems grants PHY-0749991 and PHY-1150760, Army Research Office Grant No. W911NF-11-1-0514, and the Army Research Laboratory (ARL)Micro Autonomous Systems and Technology (MAST)Collaborative Technology Alliance (CTA) undercooperative agreement numberW911NF-08-2-0004.

References

[1] NeddermanR1992 Statics andKinematics of GranularMaterials (Cambridge: CambridgeUniversity Press)

[2] Maladen RD,Ding Y, Li C andGoldmanD I 2009Undulatoryswimming in sand: subsurface locomotion of the sandfishlizard Science 325 314–8

[3] Sharpe S S, Ding Y andGoldmanD I 2012 Environmentalinteraction influencesmuscle activation strategy during sand-swimming in the sandfish lizard Scincus scincus J. Exp. Biol.216 260–74

[4] Sharpe S S, Koehler SA, KuckukRM, SerranoM,Vela PA,Mendelson J III andGoldmanD I 2015 Locomotor benefits ofbeing a slender and slick sand-swimmer J. Exp. Biol. 218440–50

[5] MarviH et al 2014 Sidewindingwithminimal slip: snake androbot ascent of sandy slopes Science 346 224–9

[6] MazouchovaN,GravishN, SavuA andGoldmanD I 2010Utilization of granular solidification during terrestriallocomotion of hatchling sea turtlesBiol. Lett. 6 398–401

[7] Li C,Hsieh S T andGoldmanD I 2012Multi-functional footuse during running in the zebra-tailed lizard (Callisaurusdraconoides) J. Exp. Biol. 215 3293–308

[8] IrschickD J and Jayne BC 1999Afield study of the effects ofincline on the escape locomotion of a bipedal lizard,Callisaurus draconoides Physiol. Biochem. Zoology 72 44–56

[9] KorffWL andMcHenryM J 2011 Environmental differencesin substratemechanics do not affect sprinting performance insand lizards (Uma scoparia andCallisaurus draconoides) J. Exp.Biol. 214 122–30

[10] Maladen RD,Ding Y,Umbanhowar PB, KamorA andGoldmanD I 2011Mechanicalmodels of sandfish locomotionreveal principles of high performance subsurface sand-swimming J. R. Soc. Interface 8 1332–45

[11] Maladen RD,Ding Y,Umbanhowar PB andGoldmanD I2011Undulatory swimming in sand: experimental andsimulation studies of a robotic sandfish Int. J. Robot. Res. 30793–805

[12] Maladen RD,Umbanhowar PB,Ding Y,Masse A andGoldmanD I 2011Granular lift forces predict verticalmotionof a sand-swimming robot IEEE Int. Conf. onRobotics andAutomation (ICRA) pp 1398–403

[13] Li C,Umbanhowar PB, KomsuogluH,KoditschekDE andGoldmanD I 2009 Sensitive dependence of themotion of alegged robot on granularmedia Proc. Natl Acad. Sci. 106 3029

[14] Li C, Zhang T andGoldmanD I 2013A terradynamics oflegged locomotion on granularmedia Science 339 1408–12

[15] Ding Y, Sharpe S S,Masse A andGoldmanD I 2012Mechanicsof undulatory swimming in a frictional fluidPLoS Comput.Biol. 8 e1002810

[16] GoldmanD I 2014Colloquium: biophysical principles ofundulatory self-propulsion in granularmediaRev.Mod. Phys.86 943

[17] European-Environmental-Agency 2012Global surface soilmoisture content based on remote sensing data (available:http://eea.europa.eu/data-and-maps/)

[18] Fournier Z et al 2005Mechanical properties of wet granularmaterials J. Phys.: Condens.Matter. 17 S477

[19] Mitarai N andNori F 2006Wet granularmaterialsAdv. Phys.55 1–45

[20] Herminghaus S 2005Dynamics of wet granularmatterAdv.Phys. 54 221–61

[21] Jung S 2010Caenorhabditis elegans swimming in a saturatedparticulate systemPhys. Fluids 22 031903

[22] Trueman ER 1970Themechanismof burrowing of themolecrab, emerita J. Exp. Biol. 53 701–10

[23] Juarez G, LuK, Sznitman J andArratia P E 2010Motility ofsmall nematodes inwet granularmediaEurophys. Lett. 9244002

[24] Ducey PK, FormanowiczDR Jr, Boyet L,Mailloux J andNussbaumRA1993 Experimental examination of burrowingbehavior in caecilians (Amphibia: Gymnophiona): effects ofsoil compaction on burrowing ability of four speciesHerpetologica 49 450–7

[25] Winter AG,Deits R LHandHosoi A E 2012 Localizedfluidization burrowingmechanics of Ensis directus J. Exp. Biol.215 2072–80

[26] GidmarkN J, Strother J A,Horton JM, Summers A P andBrainerd E L 2011 Locomotory transition fromwater to sandand its effects on undulatory kinematics in sand lances(ammodytidae) J. Exp. Biol. 214 657–64

[27] Kornilios P, Kyriazi P, PoulakakisN, KumlutaşY, Ilgaz Ç,MylonasMand Lymberakis P 2010 Phylogeography of theocellated skinkChalcides ocellatus (squamata, scincidae), withthe use ofmtdna sequences: a hitch-hikers guide to theMediterraneanMol. Phylogenetics and Evol. 54 445–56

[28] ScheelM, SeemannR, BrinkmannM,DiMichielM,Sheppard A, Breidenbach B andHerminghaus S 2008Morphological clues towet granular pile stabilityNat.Mater. 7189–93

[29] RichefeuV, El YoussoufiMSandRadjai F 2006 Shear strengthproperties of wet granularmaterials Phys. Rev.E 73 051304

[30] DickinsonWWandWard JD 1994 Lowdepositional porosityin eolian sands and sandstones, Namib desert J. SedimentaryRes. 64 226–32

[31] GravishN,Umbanhowar PB andGoldmanD I 2010 Forceandflow transition in plowed granularmediaPhys. Rev. Lett.105 128301

[32] Marston JO,Vakarelski I U andThoroddsen ST 2012 Sphereimpact and penetration intowet sand Phys. Rev.E 86 020301

[33] Jackson S J,WhyteMA andRomanoM2009 Laboratory-controlled simulations of dinosaur footprints in sand: a key tounderstanding vertebrate track formation and preservationPalaios 24 222–38

[34] Schubert H,HerrmannWandRumpfH 1975Deformationbehaviour of agglomerates under tensile stress Powder Technol.11 121–31

[35] Vandewalle N, LumayG, Ludewig F and Fiscina J E 2012Howrelative humidity affects randompacking experiments Phys.Rev.E 85 031309

[36] Gioia G, CuitiñoAM, Zheng S andUribe T 2002Two-phasedensification of cohesive granular aggregates Phys. Rev. Lett. 88204302

[37] Fiscina J E, LumayG, Ludewig F andVandewalleN 2010Compaction dynamics of wet granular assemblies Phys. Rev.Lett. 105 048001

[38] Carranza S, Arnold E,Geniez P, Roca J andMateo J 2008Radiation,multiple dispersal and parallelism in the skinks,chalcides and sphenops (squamata: Scincidae), withcomments on scincus and scincopus and the age of the SaharadesertMol. Phylogenetics and Evol. 46 1071–94

[39] Greer A E,CaputoV, Lanza B and Palmieri R 1998Observations on limb reduction in the scincid lizard genusChalcides J. Herpetology 244–52

[40] AttumO, Eason P andCobbsG 2007Morphology, nichesegregation, and escape tactics in a sand dune lizardcommunity J. Arid Environ. 68 564–73

[41] Albert R, PfeiferMA, Barbási A L and Schiffer P 1999 Slowdrag in a granularmedium Phys. Rev. Lett. 82 205

15

Phys. Biol. 12 (2015) 046009 S S Sharpe et al

http://dx.doi.org/10.1126/science.1172490http://dx.doi.org/10.1126/science.1172490http://dx.doi.org/10.1126/science.1172490http://dx.doi.org/10.1242/jeb.070482http://dx.doi.org/10.1242/jeb.070482http://dx.doi.org/10.1242/jeb.070482http://dx.doi.org/10.1242/jeb.108357http://dx.doi.org/10.1242/jeb.108357http://dx.doi.org/10.1242/jeb.108357http://dx.doi.org/10.1242/jeb.108357http://dx.doi.org/10.1126/science.1255718http://dx.doi.org/10.1126/science.1255718http://dx.doi.org/10.1126/science.1255718http://dx.doi.org/10.1098/rsbl.2009.1041http://dx.doi.org/10.1098/rsbl.2009.1041http://dx.doi.org/10.1098/rsbl.2009.1041http://dx.doi.org/10.1242/jeb.061937http://dx.doi.org/10.1242/jeb.061937http://dx.doi.org/10.1242/jeb.061937http://dx.doi.org/10.1086/316641http://dx.doi.org/10.1086/316641http://dx.doi.org/10.1086/316641http://dx.doi.org/10.1242/jeb.045682http://dx.doi.org/10.1242/jeb.045682http://dx.doi.org/10.1242/jeb.045682http://dx.doi.org/10.1098/rsif.2010.0678http://dx.doi.org/10.1098/rsif.2010.0678http://dx.doi.org/10.1098/rsif.2010.0678http://dx.doi.org/10.1177/0278364911402406http://dx.doi.org/10.1177/0278364911402406http://dx.doi.org/10.1177/0278364911402406http://dx.doi.org/10.1177/0278364911402406http://dx.doi.org/10.1073/pnas.0809095106http://dx.doi.org/10.1126/science.1229163http://dx.doi.org/10.1126/science.1229163http://dx.doi.org/10.1126/science.1229163http://dx.doi.org/10.1371/journal.pcbi.1002810http://dx.doi.org/10.1103/RevModPhys.86.943http://eea.europa.eu/data-and-maps/http://dx.doi.org/10.1088/0953-8984/17/9/013http://dx.doi.org/10.1080/00018730600626065http://dx.doi.org/10.1080/00018730600626065http://dx.doi.org/10.1080/00018730600626065http://dx.doi.org/10.1080/00018730500167855http://dx.doi.org/10.1080/00018730500167855http://dx.doi.org/10.1080/00018730500167855http://dx.doi.org/10.1063/1.3359611http://dx.doi.org/10.1209/0295-5075/92/44002http://dx.doi.org/10.1209/0295-5075/92/44002http://dx.doi.org/10.1242/jeb.058172http://dx.doi.org/10.1242/jeb.058172http://dx.doi.org/10.1242/jeb.058172http://dx.doi.org/10.1242/jeb.047068http://dx.doi.org/10.1242/jeb.047068http://dx.doi.org/10.1242/jeb.047068http://dx.doi.org/10.1016/j.ympev.2009.09.015http://dx.doi.org/10.1016/j.ympev.2009.09.015http://dx.doi.org/10.1016/j.ympev.2009.09.015http://dx.doi.org/10.1038/nmat2117http://dx.doi.org/10.1038/nmat2117http://dx.doi.org/10.1038/nmat2117http://dx.doi.org/10.1038/nmat2117http://dx.doi.org/10.1103/PhysRevE.73.051304http://dx.doi.org/10.1103/PhysRevLett.105.128301http://dx.doi.org/10.1103/PhysRevE.86.020301http://dx.doi.org/10.2110/palo.2007.p07-070rhttp://dx.doi.org/10.2110/palo.2007.p07-070rhttp://dx.doi.org/10.2110/palo.2007.p07-070rhttp://dx.doi.org/10.1016/0032-5910(75)80037-4http://dx.doi.org/10.1016/0032-5910(75)80037-4http://dx.doi.org/10.1016/0032-5910(75)80037-4http://dx.doi.org/10.1103/PhysRevE.85.031309http://dx.doi.org/10.1103/PhysRevLett.88.204302http://dx.doi.org/10.1103/PhysRevLett.88.204302http://dx.doi.org/10.1103/PhysRevLett.105.048001http://dx.doi.org/10.1016/j.ympev.2007.11.018http://dx.doi.org/10.1016/j.ympev.2007.11.018http://dx.doi.org/10.1016/j.ympev.2007.11.018http://dx.doi.org/10.2307/1565304http://dx.doi.org/10.2307/1565304http://dx.doi.org/10.2307/1565304http://dx.doi.org/10.1016/j.jaridenv.2006.07.010http://dx.doi.org/10.1016/j.jaridenv.2006.07.010http://dx.doi.org/10.1016/j.jaridenv.2006.07.010http://dx.doi.org/10.1103/PhysRevLett.82.205

[42] HuDL,Nirody J, Scott T and ShelleyM J 2009Themechanicsof slithering locomotion Proc. Natl Acad. Sci. 106 10081–5

[43] Comanns P, Effertz C,Hischen F, Staudt K, BöhmeWandBaumgartnerW2011Moisture harvesting andwater transportthrough specializedmicro-structures on the integument oflizardsBeilstein J. Nanotechnology 2 204–14

[44] Sharpe S 2013Control of burial and subsurface locomotion inparticulate substratesPhDDissertationGeorgia Institute ofTechnology

[45] DorganKM, Jumars PA, JohnsonB, Boudreau BP andLandis E 2005 Burrowingmechanics: burrow extension bycrack propagationNature 433 475

[46] Alexander RM1985Themaximum forces exerted by animalsJ. Exp. Biol. 115 231–8

[47] McKenzie B andDexter A 1988Radial pressures generated bythe earthwormAporrectodea roseaBiol. Fertility Soils 5 328–32

[48] Trueman ER 1967The dynamics of burrowing in Ensis(bivalvia) Proc. R. Soc.B 166 459–76

[49] JoschkoM,Diestel H and LarinkO1989Assessment ofearthwormburrowing efficiency in compacted soil with acombination ofmorphological and soil physicalmeasurementsBiol. Fertility Soils 8 191–6

[50] Full R J andKoditschekDE 1999Templates and anchors:neuromechanical hypotheses of legged locomotion on landJ. Exp. Biol. 202 3325–32

[51] Ding Y, Sharpe S S,Wiesenfeld K andGoldmanD I 2013Emergence of the advancing neuromechanical phase in aresistive force dominatedmedium Proc. Natl Acad. Sci. 11010123–8

[52] Zhang T andGoldmanD I 2014 The effectiveness of resistiveforce theory in granular locomotiona) Phys. Fluids 26101308

16

Phys. Biol. 12 (2015) 046009 S S Sharpe et al

http://dx.doi.org/10.1073/pnas.0812533106http://dx.doi.org/10.1073/pnas.0812533106http://dx.doi.org/10.1073/pnas.0812533106http://dx.doi.org/10.3762/bjnano.2.24http://dx.doi.org/10.3762/bjnano.2.24http://dx.doi.org/10.3762/bjnano.2.24http://dx.doi.org/10.1038/433475ahttp://dx.doi.org/10.1007/BF00262141http://dx.doi.org/10.1007/BF00262141http://dx.doi.org/10.1007/BF00262141http://dx.doi.org/10.1098/rspb.1967.0007http://dx.doi.org/10.1098/rspb.1967.0007http://dx.doi.org/10.1098/rspb.1967.0007http://dx.doi.org/10.1007/BF00266478http://dx.doi.org/10.1007/BF00266478http://dx.doi.org/10.1007/BF00266478http://dx.doi.org/10.1073/pnas.1302844110http://dx.doi.org/10.1073/pnas.1302844110http://dx.doi.org/10.1073/pnas.1302844110http://dx.doi.org/10.1073/pnas.1302844110http://dx.doi.org/10.1063/1.4898629http://dx.doi.org/10.1063/1.4898629

1. Introduction2. Materials and methods2.1. Dry media preparation2.2. Wet media preparation2.2.1. Wet media mixture2.2.2. Apparatus development2.2.3. Technique characterization and capabilities

2.3. Force measurements2.4. Ocellated skink experiments

3. Results3.1. Penetration and drag forces in a wet substrate3.2. Ocellated skink burial in wet and dry media

4. Discussion4.1. Preparation of states of wet GM4.2. Resistance forces4.3. Burial behavior4.4. Burial limited by resistive force

5. ConclusionsAcknowledgmentsReferences