HERPETOLOGICAL JOURNAL, Vol. 10, pp. 137-142 (2000)

THE TERRESTRIAL SUMMER HABITAT OF RADIO-TRACKED GREAT CRESTED NEWTS (TRITURUS CRISTATUS) AND MARBLED NEWTS (T. MARMORATUS)

ROBERT JEHLE

Institute o/Zoology, University 0/ Vienna, Vienna, Austria

Thirty great crested newts (Triturus cristatus) and 25 marbled newts (T. marmora/us) were radio tracked for up to 31 days after leav ing breeding ponds in western France. Around the pond where most newts were radio-tracked. 95% of all summer refuges fell within a radius from the pond of63.0 m for T. cristatus and 59.5 m for T. marmora/us. The most frequen tl y used habitats were directly adjacent to th e pond shoreline. For the other two ponds, all summer refuges fell within radi i of between 26.2 m and 32.3 m from the ponds respectively. No significant differences among species or ponds were observed in the mean distance that newts moved away from th e breeding site. Five rad io-tracked T. cristatus and two radio-tracked T. marmoratliS that were moved back into their ponds migrated during the following night in almost identical directions to their initi al em igrat ions, but did not return to the original refuges. In eight cases, several (up to ten) untracked newts were found in th e refu ges of radio-tracked individuals.

Key words: radio-tracking, terrestria l sum mer habitat, site fide li ty, Tritllrlls crista/us, T. marmora/us

INTRODUCTION paper reports translocation experiments which tested European newts (genus Triturus) exhibit the the hypothesis that T. crislalus and T. marmoratus re­ biphasic life history pattern typical of amphibians and turn to particular terrestrial shelters. require both aquatic and terrestrial habitats (Griffiths, MATERIALS AND METHODS J 996). In Britain, there is currently a special concern about the conservation status of the great crested newt The Ihree study ponds are locntcd in the Trilurus crislalus (Gent & Bray, J 994; Beebee, 1997), Departement de Mayenne, Western France. Pond J which ha s become an important "umbrell a species" (approx. 50 m' area, at the edge of pastures with hedge­ whose protection serves to conserve a multitude of rows) and Pond 2 (approx. 20 m' area, an abandoned lesser-known, coexisting taxa. Nevertheless, protection sand quarry) lie near the village of Jublains; Pond 3 measures for T. crislatus sites often concentrate on their (approx. 150 m' area) lies in a pasture near the village breeding ponds on ly; this is mainly because empirical of Marcill e-la-Vi lle. In 1997 a radio-tracking study was data on their terrestrial ecology and behaviour are conducted at Pond I only, focusing on a comparison of scarce, owing to the difficulty of locating newts on migration and habitat utilization patterns between adult land. T. crista/us and T marmora/us (Jeh le & Arntzen, Lower fitness parameters for amphibians li ving in 2000). In the present paper some data from 1997 are fragmented landscapes corroborate the prediction of reanalysed alongside data coll ected in 1998 at all three the metapopulation concept that isolated popUlations ponds. At Pond 3 only T. cristallls were radio-tracked, are exposed to an increased risk of extinction (Hanski despite the presence of a small population of T. & Gilpin, 1997; Hitch in gs & Beebee, 1997; 1998), and marmora/us. demonstrate that stud ies on movement patterns are of The radio-tracking was conducted between June 21 vital importance for conservation issues. The recent ap­ and July 30, 1997 and between July I and August 3, plication of radio-tracking to adult T. crislalus and T. J 998. For a description of the radio-tracking procedure marmora/us enabled Jehle & Arntzen (2000) to fo ll ow in 1997, see Jehle & Arntzen (2000). In 1998 newts the movements of these newts over a period of approxi­ were captured in the ponds between June 30 and July 2 mately one month after leavin g a shared breeding site in with dip nets and kept in semi-terrestrial enclosures western France. In this paper, I extend the study and in­ made from 220-litre metal barrels, cut lengthwise and corporate data from a second year and two additional covered with netting to prevent escape. Water in the en­ ponds. The distance that radio-tracked adult T. crislalus closures was taken from the ponds where the newts and T. marmora/us move away from their ponds serves were captured. Newts that had moved out of the water to determine a terrestrial zone arou nd breeding site$ were assumed to have started their terrestrial phase, and which is used for summer refuges. Additionally, this were equipped with a transmitter and released at ap­ proximately 2100 hrs in the pond of their capture. In Correspondence: R. Jehle, Institute of Zoology, University 1998, Holohil Systems transmitters BD-2A (mass: of Vienna, Althanstr. 14, A-1090 Vienna, Austria. E-mail: 0.69 g - 0.78 g; battery li fe: 25-35 days) were used ex­ [email protected] clusively, with the external antenna twisted around the 138 R. JEHLE

TABLE I. Summary of movement data of the rad io-tracked adult newts at the three study sites. T. c.: crested newt (Triturus cristalus), T. m.: marbled newt (Tritu rlls marmora/us).

Pond 1 Pond 2 Pond 3 T c. T m. T c. T m. T c.

No. newts released 15 19 5 6 10 No. observed localisations 45 53 8 6 16 Tracking duration (days): minimum 3 3 4 4 5 maximum 28 31 13 6 10 median 16 15 5 6 8 Distance to pond (m): minimum 1.4 1.8 2. 1 1.8 13 maximum 95.7 146.0 26.2 30.0 32.3 median 9.3 12.2 14.8 13.7 15.9

transmitter and fi xed with "super glue" . Prior to trans­ 7 mm long) and surgical silk supplied for human medi­ mitter implantation, the newts were anaesthetised with cine. Bleeding rarely occurred, and no other MS 222 (Sandoz), until the muscular system was re­ complications were observed during the implantation laxed and the animals stopped moving (10-20 mins). procedure. All animals recovered from anaesthesia af­ The body cavity was opened at the ventro-Iateral side ter approximately 30 mins with no apparent ill effects. for 8-10 mm with a scalpel, and th e transmitter was in­ At the end of the tracking period, newts had their trans­ serted using forceps. The wound was closed with four mitters removed and were then released in the pond, as sutures using an iris cutting needle (a C-shaped needle recovering the newts would have damaged the

FIG. I . Aeria l photographs of Pond 1 and its surroundin g area. Th e circles around the pond (corrected using ground reference points as the picture is not taken exact ly above th e pond) encompass the areas where 50% and 95% of th e radio-tracked newts were observed. (a) Tritllrus crista/us, two localisations are outside Ihe image; (b) T. marmora/us, three localisations are outside the image. TERRESTRIAL HABITA TS OF NEWTS 139 microhabitat in which they were found. Some transmit­ (13%) the transmitter was found without the study indi­ ters were re-used. vidual, six transmitters (II %) were detected but could The 1997 season showed that the vast majority of not be recovered as they were inaccessible, three newts migratory activity took place during the first night after (6%) were found dead, three newts (6%) were eaten by the newts were released (Jehle & Arntzen, 2000). The snakes, and in three cases (6%) the signal was lost. All distance travelled from the pond edge after three days in dividuals but one left the pond on the night of release was not significantly different from that at the end of and moved between 2 m and 146 m away from the the tracking period, when the transmitter battery was pond. Newts were tracked for periods rangin g from due to expire (paired I-test with log-transformed data: three days to 31 days. Owing to a longer median track­ P>0.05). The collection of data in 1998 was therefore ing period in 1997, the number of localisations per restricted to one record per day over a period of up to 13 individual was higher at Pond I than at Ponds 2 and 3 days, in some cases including a translocation of the (Table I). newt back to the pond. Localisations of newts are de­ Sixty-n ine localisations of T crislaills and 59 fined as refuges where they were resident for at least 24 locali sations of T marmoraills were analysed (Table hrs, and were marked with a numbered flag. For ani­ I). The distance of the localisations from the pond! mals that were recovered dead, and for transmitters point of release did not vary sign ificantly between T found without the animal, data were included up to the crislalus and T marmoralus (F=0.04, P>O.05), or be­ last recorded movement. Distance measures were taken tween ponds (F=2.05, P>0.05). At Pond I , no to the nearest edge of the pond and to spatial reference significant differences between the sexes were detected points selected on the basis of good visibility and spac­ within species (T crista/us: 1=0 .93, T. marmora/us: ing (White & Garrott, 1990). Localisations were plotted 1= 1.73, P>0.05 in both cases). At Pond I, the circles en­ on aerial photographs. For Pond I, circles around the compassing 50% of all T. crista/us and T marmora/us pond were drawn on the photograph to plot the area where 50% and 95% of all summer refuges were ob­ served, as refuges were not expected to be evenly distributed in space. In line with the small er sample sizes, only circles circumscribing 100% of refuges were plotted around Ponds 2 and 3. Ponds I and 2 were considered to be approximately circular and the circles were centred on the middle of the pond. For Pond 3 the circles were centred on the point of release ofthe newts. Because the aerial photographs were not taken exactly above each pond, the circles were corrected using the reference points and plotted as ellipses. Distances of refuges from the pond centre (Ponds I and 2) or point of release (Pond 3) were compared between ponds and between species (simple factorial ANaYA with all ponds) and between sexes within species (I-tests for Pond I) using log-transformed data. Translocation experiments were conducted with two T criSlalus and two T marmoralus at Pond 2 and three T crislaills at Pond 3. Newts that had already been ra­ dio-tracked to their terrestrial refuges were recovered during daytime and released back into their ponds at 2100 hrs on the same day. Their migrations were subse­ quently fo llowed without changing the data collection procedure. The relationship between the direction (in azimuth) of initial emigration from the pond and emi­ gration direction after translocation was determined usin g linear regression (after Batschelet, 1981). Owing to the limited sample size, data from both species were combined. RESULTS FIG. 2. Aerial photograph of Pond 2 and its surrounding area. In 1997, 30 newts were radio-tracked at Pond I. In The ci rcles around the pond (corrected using gro und 1998, four, eleven, and ten newts were radio-tracked at reference points as the picture is not taken exact ly above the pond) encompass the areas where 100% of the radio-tracked Ponds 1,2, and 3, respectively (Table I). Thirty-three newts were observed. Sol id line: Trilllf'lls cristatus; broken (60%) newts were recovered alive; for seven newts line: T. marmoratus. 140 R. JEHLE

Twenty-three untracked newts were fo und sharing the refuges of eight radio-tracked newts, with up to ten newts in one refuge. The five T. cris/atus and the two T. marmora/us that were translocated back to their respec­ tive ponds left the pond on the day of translocation. They did not migrate to the refuge in whi ch they were recovered the day before (Fig. 4a), but showed a sig­ nifi cant tendency to depart in the same direction as before (Fig. 4b). Two T. cristatus spent four and seven days, respectively, in a transient refuge before moving closer to their initial site of recapture (Fig. 4a).

DISCUSSION FIG . 3. Aerial photograph ofPand 3 and its surrounding area. The circle around th e pond (corrected us in g ground reference The radio-tracking methodology is a direct monitor­ point s as th e picture is not taken exactly above th e pond) ing technique which is su itable for collecting data on encompass th e areas where 100% of the radi o-tracked the secretive terrestrial life of ad ult newts. The ob­ Triturus crislatus were observ ed. served losses were within the range of other te lemetric studies on urodeles (Mad ison , 1997; Madison & a) Farrand, 1998), and the frequ ent observations of radio­ N tracked newts sharing refu ges with untracked newts is in line with the assumption that the migration behav­ I iour of radio-tracked individuals was not abnormal. \ The time when the study individuals moved to the ter­ \ restrial part oftheir enclosures, prior to being implanted 10m with transmitters, coinc ided with a decreasing number ~""- Pond 2 of newts in the ponds; this supports the view that the study animals were ready to leave the water. Neverthe­ .... ,.,.,,, ....,',..r less, the procedures of imp lanting and remov ing a transmitter are invas ive and the ir fu ll conseq uences are 0" unknown; for example, no data are ava il ab le on poss i­ ." 90' ble adverse consequences for the newts after th e study ]'" b) period. One of the major lim itations of radio-tracking ! 360'/0' r=O.99, p<0.01, newts is the short lifetime of transmitters, precluding ~ n='7 • • th e collection of data over entire seasons. Th e tra nsmit­ ~ '" 270' • ters' mass was between 7. 0% and 14.3% (median: 8.7%) of the newts' body mass, and more powerful bat­ i 180' teries wou ld raise th e tran smitter mass:body mass ratio '13 to an unacceptab le leve l. Future tracking studies, also ~ 0 90' with regards to juveniles, may be possible using re­ ~ 90' 180' 270' 360'/0' 90' 0: motely detectable tags without internal energy sources « Angle of initial migration (Lovei, Stringer, Dev in e & Cartelli eri, 1997). Spatial movem ent patte rns were very similar be­ FIG . 4. Translocation experim ents of radio-tracked newts. (a) tween breeding sites. During the period of study, 50% Migrations of five Tl'ilurllS cristalus (filled arrowhead) and two T. marmora/lis (open arrowh ead) whi ch were released in of all newts moved only a few metres away from the Ponds 2 and 3, recovered, and put back into th e pond s. shore of Pond I, which suggests that the most important Arrows with the same origin represent one indi vidual. So lid area was directl y adjacent to the breeding site. As only line: initial emigration from th e pond; broken line: a few individuals perfonned the large-distance move­ emigration after tran slocation back into th e pond . (b) ments, th e sma ll er apparent summer hab itat range in Relationship between th e initi al direction of migration and the direction of migration after re-release in the ponds. Ponds 2 and 3 might be a consequence of the small er sample size. For th e purpose of defining a protected locali sations had radii of 12.2 m and 14 .8 m, respec­ area of terrestrial hab itat around a breed ing site. a sim­ tively; th e circles encompassing 95% of locali sations ple prescribed radius might not be sati sfactory as a had radii of 63.0 m (T. cristatus) and 59.5 m (T. general guid eline, as newts are li ke ly to prefer certain marmoratus) (Fig. I). The circles encompass in g all directions and habitat types. However, quantifying localisations around Pond 2 had radii of 30.0 m (T. habitat use is time-consuming, and radio-trackin g data crista/us) and 26.2 m (T. marmoratus, Fig. 2). At Pond as well as drift fence studies have shown that, although 3 a circ le of radiu s 32.3 m encompassed all mo vements are non -random in or ientat ion , narrow mi p localisations of T. crista/liS (Fig. 3). gration co rridors do not necessarily ex ist (Jehle, TERRESTRIAL HABIT A TS OF NEWTS 141

Pauli-Thonke, Tamnig & Hodl, 1997; Dodd & Cade, return to particular areas on land. Translocated newts 1998). moved away from their ponds in very similar directions The radius encompassing 95% of all T. crista/us and to those of their initial emigration . Although none of T. marmora/us refuges in this study was considerably the newts used exactly the same refuge again, that smaller than the 164.3 m buffer zone estimated for might have been due in part to disturbance associated North American ambystomatid salamanders on the ba­ with recovering the newt prior to translocation. Sepa­ sis of95% of recaptures (Semlitsch, 1998). Two factors rate observations have indicated that when newts were might account for this difference. First, T. crista/us and translocated to different terrestrial habitats they subse­ T. marmora/us have an aquatic phase which lasts ap­ quently moved only short distances «5 m) and in no proximately 5 and 3 months, respectively (Bouton, particular direction (unpublished data). Apparently, 1986; Griffiths & Mylotte, 1987), whereas they were disorientated when translocated to an un­ ambystomatids are more terrestrial, spending 86-99% known site, as observed by Madison & Farrand (1998) of the year on land (Semlitsch, 1998). Second, with translocated, radio-tracked Ambys/oma /igrinum. ambystomatids exhibit bimodal migratory activity with The question of whether newts become disorientated peaks in AprillMay and OctoberlNovember, and when through massive alterations of the terrestrial area radio-tracked over the entire year move up to 286.5 m around establi shed breeding ponds, for example due to away from their ponds (Madison, 1997; Madison & urban development, is still open. Farrand, 1998). Drift-fence studies on T. crista/us also Personal observation suggested that ample refuges suggested an increase of migration activity in autumn were availab le to the newts in my study area; this im­ (Verrell & Halliday, 1985), and mark-recapture data plies a social behavioural mechanism behind the showed that within about one year single T. crista/us frequent detection of several newts sharing a refuge. can migrate over much larger distances than those ob­ The basis of such behaviour in T. crista/us has been ad­ served in this study (Kupfer, 1998). With the present dressed by Hayward e/ at. (2000), but its implications limitation of battery life, the task of collecting radio­ have not been investigated in the context of alteration tracking data of large-bodied newts after summer of habitat or translocation of animals to make way for would require that animals be caught and tagged during roads or other developments. their terrestrial phase. The data from the present study ACKNOWLEDGEMENTS are not sufficient for the designation of a general terres­ trial buffer zone encompassing the space required over This study was financed by the Austrian Science all life stages (cf. Sem litsch, 1998). Foundation (FWF grant P-1l852 BlO to W. Hodl), and Owing to successional processes and the relatively conducted under the permits 97/204 & 981226/AUT of small size of typical amphibian ponds, many such the Ministere de l'Environment, Paris. The Austrian ponds are short-li ved on an ecological timescale. The Academy of Sciences (KIOS) provided additional colonisation of newly-formed breeding sites, in combi­ transmitters. Thanks are due to M. Sztatecsny for help nation with abandonment of sites that become in the field, D. Bardou for hospitality, and the landown­ unsuitable, is a major component of amphibian popula­ ers for access to their property. Pim Arntzen initiated tion dynamics. The importance of hedgerows as my work in Mayenne, and his suggestions significantly corridors in the "bocage" landscape typical of western improved a previous draft of the manuscript. Critical France is well documented (for example carabid bee­ remarks by C. Cummins and three anonymous review­ tl es: Burel, 1989), and, being among the preferred ers helped to avoid unwarranted conclusions. habitat types (Jehle & Arntzen, 2000), it is very likely REFERENCES that they also serve as main corridors for dispersal of large-bodied newts. However, a study on T. crista/us in Arntzen, J. W. & Wallis, O. (1991). Restricted gene flow England revealed a preference for deciduous wood­ in a moving hybrid zone of newts (TritllYliS cristatus lands (Latham, Oldham, Stevenson, Duff, Franklin & and T. marmoratus) in western France. Evolution 45, Head, 1996), whereas in Mayenne a local expansion of 805-826. the species' distribution was associated with the re­ Batschelet, E. ( 1981). Circufar Statistics in Biology, moval of shrubs and small woodlands (Arntzen & London: Academic Press. Wallis, 1991); this suggests that habitat preferences Beebee, T. J. C. (1997). Changes in dew pond numbers and may vary across the species' range. The mUltiple func­ amp hibian diversity over 20 years on chalk downland tions and the varying quality of different types of in Sussex, England. Bioi. Cons. 81 , 215-2 19. hedgerow as lifelines between populations of mammals Bennett, A. F., Henein, K. & Merriam, O. (1994). Corridor has been demonstrated (Bennett, Henein & Merriam, use and the elements of corridor quality: chipmunks 1994), but the role of hedgerow quality for connecting and fencerows in a farmland mosaic. BioI. Cons. 68, newt breeding sites is not yet quantified. 155-156. That European newts return to their breeding ponds Bouton, N. (1986). Donnees sur 1a migration de Triturus after being translocated has been documented before cristatus et T. marmoratus (Urodela, Salamandr idae) (Joly & Miaud, 1993), but according to the best of my dans Ie Departement de la Mayeone (France). Bull. knowledge th is is the first study to show that they also Soc. Herp. Fr. 40, 43-51. 142 R. JEHLE

Burel, F. (1989). Landscape-structure effects on carabid loly, P. & Miaud, C. (1993). How does a newt find its spatial patterns in western France. Landscape Ecology pond? The role of chemical cues in migrating newts 2, 2 15-226. (Tritllrus alpestris). Ecol. Ethol. Evol. 5,447-455. Dodd, C. K. & Cade, B. S. (1998). Movement patterns and Kupfer, A. (1998). Wanderstrecken einzelncr Kammolche the conservation of amphibians breeding in small, (Trilurus crista/us) in einem Agrarlebensraurn. temporary wetlands. Cons. BioI. 12,33 1-339. Zeitschriftfor Feldherpetologie S, 238-242. Gent, A. & Bray, R. (Eds., 1994). Conservation and Latham, D. M., Oldham, R. S., Stevenson, M. I ., Duff, R., management 0/ great crested newts, English Nature Franklin, P. & Head, S. M. (1996). Woodland Science Report No. 20. Peterborough: English Nature. management and the conservation of the great crested Griffiths, R. A. (1996). Newts and Salamanders of newt (Trilurus cristatus). Aspects of Applied Biology Europe. London: T. & A. D. Poyser. 44, 451-459. Griffiths, R. A. & Mylotte, V. I. ( 1987). M icrohabitat Lovei, G. L., Stringer, 1. A. N., Devine, C. D. & selection and feeding relations of smooth and warty Cartellieri, M. (1997). Harmonic Radar - a method newts, Triturus vulgaris and T., crista/us, at an upland using inexpensive tags to study in vertebrate movement pond in mid-Wales. Hofarcl. Ecol. 10, 1-7. on land. New Zealand J. Ecol. 21, 187-193. Hanski, I. & Gilpin , M. E. (Eds., 1997). Metopoplilation Madison, D. (1997). The emigration of radio-implanted Biology: Ecology, Genetics, and Evolulion. San spotted salamanders, Ambysloma maculatum. J. Diego: Academic Press. Herpetol. 31, 542-55 1. Hayward, R., Oldham, R. S., Watt, P. I. & Head, S. M. Madison, D. M. & Farrand Ill, L. ( 1998). Habitat use (2000). Dispersion patterns of youn g great crested during breeding and emigration in radio- implanted newts (Tritllrus crista/us). Herpetol. J. 10, 129-136. tiger salamanders, Ambys/oma ligrinum. Copeia 1998, Hitchings, S. P. & Beebee, T. I. C. (1997). Genetic 402-4 10. substructuring as a result of barriers to gene now in Semlitsch, R. D. ( 1998). Biological delineation of urban Rana temporaria (common frog) populations: terrestri al buffer zones for pond-breeding implications for biodiversity conservation. Heredity salamanders. Cons. Bioi. 12, 1113- 111 9. 79, 117-127. Verrell, P. A. & Halliday, T. R. (1985). The population Hitchings, S. P. & Beebee, T. I. C. (1998). Loss of genetic dynamics of th e crested newt Trill/I'll.!> cristatus at a diversity and fitness in common toad (Bulo bula) pond in southern England. !-lolm·ct. Ecol. 8, 151- 156. populations isolated by inimical habitat. J. Eval. Bioi. White, G. C. & Garrott, R. A. (1990). AnalYSis of Wildlife 11, 269-283. Radio-Tracking Dala. San Diego: Academic Press. lehle, R., Pauli-Thonke, A., Tamnig, I. & Hodl, W. (1997). Phanologie und Wanderaktivitat des Donau­ kammmolches (Triturlls dobragicus) an einem Gewasser auf der Wiener Donauinsel). In Populations­ biologie von Amphibien: eine Langzeitsludie alii der Wiener Donauinsel, 119- 132. Ht>dl , W., Jehle, R., & Gollmann, G. (Eds.) Linz: Stapfia 51. lehle, R. & Arntzen, I . W. (2000). Post-breeding migrations of newts with contrasting ecological requiremcnts. J. Zool., Land. 2SI, 297-306. Accepted: 24.8.00