Ata Theriologia 39 (2): 25-2, 994. PL ISSN -7 5 FRAGM EN TA TH ERIO LOGICA Dormie Glis glis ativity and hazelnut onsumption Grazia RODOLFI Rodolfi G. 994. Dormie Glis glis ativity and hazelnut onsumption. Ata theriol. 39: 2 5-2 2. The iradian and monthly ativity rhythms of dormie Glis glis (Linnaeus, 766) and their foraging for a preferred food soure (hazels) have been studied. Dormie showed a monophasi pattern of ativity from June to early November. The daily ative period started from or after sunset, and ended usually before sunrise. In June and Septem ber, however, the foraging was prolonged a hr after sunrise. This suggests that the ative phase is limited by the endogenous yle of dormouse rather than by photoperiod. Hazelnut intake began in June before the fruits were ripened, and went on until the supply was exhausted, in August. Dipartim ento di Biologia Evoluzionistia e Sperimentale, sede di Zoologia, Via San Giaomo 9, 426 Bologna, Italy K ey words: Glis glis, ativity rhythm, hazelnut foraging Introdution Dormie Glis glis (Linnaeus, 766) are gregarious, noturnal rodents, found in both deiduous woods and areas inhabited by man, sometimes ausing extensive damage to fruit trees (Platt and Rowe 964, Santini 983). They an be very numerous in ountry settlements too, entering above the wooden-beam eilings and damaging the buildings. Despite their eonomi importane, few studies on dormouse eology and behaviour in Italy have been published. This paper presents a field study on the ativity rhythm of a dormouse population of about 6 individuals. I also onsidered their seasonal onsumption of hazelnuts, as the favourite food of this speies (Holisova 968, Grodzinski and Sawika-Kapusta 97, G^bzyhski et al. 972): hazelnuts have a high alori ontent (7984 alg in dry weight seeds minus oat; f Grodzinski and Sawika-Kapusta 97), a high edibleinedible share ratio (Parisi 977) and great metabolizable energy (88. ±.7% of the gross hazelnut energy; f G^bzyriski et al. 972). Moreover, dormie quikly handle and open hazelnuts, yielding great energy prodution per unit of time. All these fators make hazels an attrative food soure. Methods The study was arried out between September 987 and August 989, in the Centro Inter- dipartimentale Sorelle Clarke, situated in Bagni di Lua, Tusany ( 9 6.2-2 3.8 m a.s.l.), a 26 ha [2I5
G. Rodolfi 26 area of hilly, mixed woodland and a human-inhabited area, formed by a garden and some ultivated plots. There were several food items for rodents both in the woodland (pine one seeds, hestnuts) and in the anthropogeni area (apples, herries, pears, hazelnuts, walnuts, magnolia fruits and ypress one seeds). However, from September 987 I onentrated the field work in the man-ultivated sites, as during Ju n e-a u gu st 987 a series of preliminary observations showed that there was very little dormouse ativity in the woodland. The ativity rhythm of the population was studied in September 987, M ay-o tober 988, M ay, June and August 989, for ten days every month. The observations were arried out ontinously, reording the sounds of ativity (sounds aused by dormie moving into the tree leafy branhes when foraging, travelling or during reprodutive ativity, as well as the sounds of their alls), time, number o f trees, loation and behaviour - when possible - of any anim al seen. The method used was ontinuous reording (M artin and Bateson 988). To onfirm the ativity report, dormie were floodlighted only at the end of the preagreed observation period, so as not to interfere with their behaviour. The daily ativity rhythm was expressed eah month, all the data olleted during the period under onsideration, using both sightings and ativity signals (sounds). Ativity was alulated as the number o f dormie seen per hour of observation time. Effets of population size were removed from the level of ativity, dividing ativity per hour by the number of dormie present in the population and multipled by (Ativity index; Thompson 976). Data were analysed monthly for eah hour, as the mean of all ativity indies. To define the am ount of hazelnuts onsumed by dormie, several daily sam plings of hazelnut remains were arried out from June to August 988 and 989, with the use of a green waterproof loth suspended ju st under the anopy of the hazel trees. I olleted the hazelnut remains that fell on this loth when all the dormie returned to their nests. Sampling ourred for 3 days every one or two weeks. Hene the data were analysed per 5 day sampling, the amount of hazelnuts onsumed being expressed as a perentage of all food remains. Results and disussion Dormie showed an annual yle of ativity lasting about five months, from early June to early November. This period of ativity is similar to the findings for dormie in different parts of their distribution area: Gaisler et al. (977) reported that dormie first ourred in nest boxes on 3 May, the last ourrene being on 4 Otober (Czehoslovakia, 4-95 m a.s.l.). Pilastro (99, 992) stated that dormie used nest boxes from mid-may to early November, and that all adults started hibernating before the last week of Otober, three weeks before the most reent young of the year (Colli Berii, Vienza, Italy, - m a.s.l.). The results refer to September 987, May-Otober 988, May, June and August 989. In May 988 no ativity was observed either at the beginning or the end of the month. Ativity started in June, maintaining the same level of ativity per hour (average of ativity index). July showed slightly higher values in the first part of the night, with a derease in the last three ative hours (Fig. ). From the first days of the month, dormie began to redue the ativity duration until August, when the minimum number of ative hours per day (four hours) was reahed, as was the greatest mean ativity index. As there were no apparent exogenous fators exerting an ativity redution, so the rapid derease may have been a onsequene
Ativity and foraging in dormie 27 of births and nursing the young a this time (Gaisler et al. 977, Pilastro 992). In September, ativity inreased onsiderably, reahing the highest number of ative hours ( hours) and the minimum number of sightings (22.2 per day). In spite of the length of the ative period, a rather high and ontinuous ativity per hour was maintained, probably aused by intensive feeding prior to hibernation (Mrosovski 966, Frano 99). By early Otober the ativity values and the number of ative hours began to derease rapidly. Moreover, Otober showed an ativity pattern with a gap of two hour in the middle of the night, between a 5-hour and a 2-hour ative period. This was reorded both by the absene of voalizations and by diret sightings, but also by observations on some dormie oming bak to the nests and oming out after about 2 hours. By the seond half of November no more ativity signs were reorded. Overall, the ativity pattern light darkness light June o o> II July II CO O x. a) o "D Aug n = >. o OJ Ion Sep \ n = 222 Ot n = 37 l l i l f I I... 2 3 4 5 6 7 8 9 2 2 3 4 5 6 7 8 9 2 pm Time am Fig.. Ativity rhythm over a year (988), expressed as mean ativity index. The times of sunset and sunrise (Civil Twilight) are indiated by dotted lines. The number of dormie sighted (n) per days of field work is given in the top right-hand of the figures.
28 G. Rodolfi of 988 was monophasi, the only exeption being Otober, possibly due to a rest break. Data from May 989 showed a very slight indiation (2 hours) of animal arousal. June 989, in omparison with 988, showed a derease both in the length of the ative period (9 hours in 988; 6 hours in 989) and of the mean ativity indies per hour (6.7 in 988; 8.3-. in 989). Also in August 989 the level of ativity per hour was redued (-26.7 in 988; 3.3-8.8 in 989) as well as in September 987 (-2.7 in 987; 6.7-28.3 in 988). There was no orrelation between the lenght of night-time and the 988 monthly perentage of ative period (Sperman rank orrelation: n = 6, rs ns), as well as the onset and the end of ativity were not linked to sunset and daybreak respetively. Dormie were ative before sunset in September 987. They were ative after sunrise in June and September 988. These observations suggest that the endogenous yle may beome unoupled from the daybreak and nightfall zeitgeber. In the summer, dormie appeared to prefer hazelnuts, whih began to ripen at this time of the year. Unlike other small rodents, dormie were able to tear, handle and eat away whole groups of hazelnuts, without separating fruits one by one, Table. Hazelnut intake in 988 and 989. Number of samples refers both to groups of hazelnuts (groups of 4, 3, 2 nuts) and single nuts. Date Sample size (n) Nuts intake Groups of n % 4 nuts 3 nuts 2 nuts nut 9 8 8 ;year rop = 72 nuts June - 5 n samples 8 43 237 6-54.6 n nuts 32 29 474 887 July - 5 352 422 29.2 of these: 6-3 6 73 5.6 opened 25 4 4 887 August -5 83 39 9.6 not opened 7 5 54 6-3 Total 75 446 Perent of year rop 84.5 989; year rop = 92 nuts June - 5 n samples 6 4 3 6-3 223 266 24.5 n nuts 24 262 742 July 5 445 5 46.9 of these: 6-3 9 26 9.9 opened 98 228 742 August -5 6 95 8.7 not opened 4 22 34 6-3 Total 99 87 Perent of year rop 9.2
Ativity and foraging in dormie 29 gnawing and then tearing away the nut periarp (G. Rodolfi, pers. obs.). Of the five hazels present on the woodland border line, only the one produing the greatest number of fruits, was used by the dormouse population. This tree produed a greater rop in 988 (72 nuts) than in 989 (92 nuts), onsequently the total number of meal samples (single or grouped hazelnuts), was greater this same year for the dormouse population (Table ), while there was no differene between the years as to the ratio of hazelnuts opened by dormie between single and groups of fruits (Wiloxon mathed pairs signed-ranks test: n = 8, T = 6.5, ns, two-tailed). Hazelnut onsumption by the dormouse population ourred from early June to late August, in both years of field work. During the first half of June they seldom appeared at any of the ultivated plots, being ative espeially in the garden, where they exploited ypress Cupressus sempervirens ones, but by the seond half of June, when the fruits were still unripe but easily opened beause the periarps were still tender, they began to feed on hazelnuts. Food intake over the last fifteen days of June in 989 was almost twie as muh as in 988 (Table ). Moreover, the peak intake in 989 was fifteen days earlier (5 July) than in 988 (3 July). Feeding went on every night until the food supply ran out, whih ourred by the end of August, both in 988 and in 989, while other hazels had a still intat rop. Total dormouse intake of nuts orresponded to 84.5% and 9.2% of the two-year rops of the preferred hazel tree. Dormouse intake was always onspiuous, reahing a mean of 3.7 and.5 hazelnuts per day in the seond part of July 988 and in the first part of July 989, respetively. A pressing hazelnut intake ourred when fruits were still unripe or not yet fully ripe, exhausting the resoure in a short time. From late August, when hazelnuts were no longer available, the dormouse population was again fairly uniformly dispersed in ypresses, a garden horse-hestnut tree, magnolias and fruit trees on the ultivated sites, but it was the ypress ones - amply distribyted over the whole anthropogeni sites - that beame the most important food supply. Other fruits were utilized by dormie only to a lasser extent. Referenes Frano D. 99. Feeding habits of a dormouse population (M yoxus glis) of the Asiago Plateau (Venetian prealps). Hystrix 2: -22. Gaisler J., Holas V. and Homolka M. 977. Eology and reprodution of Gliridae (M am m alia) in Northern Moravia. Folia Zool. 26: 23-228. Gębzyński M., Góreki A. and Drożdż A. 972. Metabolism, food, assimilation and bioenergetis of three speies of dormie (Gliridae). Ata theriol. 2: 27-294. Grodziński W. and Sawika-Kapusta K. 97. Energy values of tree-seeds eaten by small mammals. Oikos 2: 5 2-5 8. Holisova V. 968. Notes on the food of dormie {Gliridae). Zool. Listy 7: 9-4. Martin P. and Bateson P. 988. Measuring behaviour. Cambridge University Press, Cambridge: -. Mrosovski N. 966. Aeleration of annual hibernating yle to 6 weeks in aptive dormie. Can. J. Zool. 44: 9 3-9.
2 G. Rodolfi Parisi V. 977. Ambiente e alimentazione. Etas libri, Milano: -37. Pilastro A. 99. Studio di una populazione di ghiro (Glis glis, Linnaeus) in un ambiente forestale dei Colli Berii. Lavori So. Ven. S. Nat. 5: 45-55. Pilastro A. 992. Communal nesting between breeding females in a free-living population of fat dormouse (Glis glis, L.). Boll. Zool. 59: 6 3-6 8. Platt F. B. W. and Rowe J. J. 964. Damage by the edible dormouse (Glis glis) at Wendorer Forest (Chilterns). Quarter. J. Forestry 58: 2 2 8-2 3 7. Santini L. 983. I roditori italiani di interesse agrario e forestale. C.N.R. Progr. Fin. Prom. Qual. Amb. A Q 232 Padova: 8 9-9 4. Thompson D. C. 976. Diurnal and seasonal ativity of the grey squirrel (Siurus arolinensis). Can. J. Zool. 55: 85-89. Reeived 6 June 993, aepted 26 April 994.