33 European Arachnology 2000 (S. Toft & N. Scharff eds.), pp. 33-38. © Aarhus University Press, Aarhus, 2002. ISBN 87 7934 001 6 (Proceedings of the 19th European Colloquium of Arachnology, Århus 17-22 July 2000) Distance of approach to prey is adjusted to the prey’s ability to escape in Yllenus arenarius Menge (Araneae, Salticidae) MACIEJ BARTOS University of Łódź, Department of Invertebrate Zoology and Hydrology, Banacha 12/16, 90-237 Łódź, Po- land. Present address: University of Łódź, Laboratory of Teaching Biology and Studies of Biological Di- versity, Banacha 1/3, 90-237 Łódź, Poland ([email protected]) Abstract The aim of the study was to investigate, whether Yllenus arenarius, a dune dwelling salticid, can ad- just its jumping distance when hunting prey of high or low escapability risk. It was found that the spiders possess a conditional hunting strategy, depending on the prey’s potential ability to escape. The spiders jumped from significantly longer distance on prey that can escape than on prey that cannot escape, thus decreasing the risk of detection and the escape of prey. There were found no significant differences in relative jumping distances within prey types between juveniles (in first and second year of life) and adults (in third year of life), suggesting, that flexibility in attack behaviour is inherited rather than learned. Key words: behaviour; predation; jumping distance; conditional strategy; Salticidae; spiders INTRODUCTION while others practically cannot escape (e.g. This article is part of a wider study (Bartos Thysanoptera and larvae of Lepidoptera). Irre- 2000) concerning predatory versatility of a sal- spective of prey type, spiders approach the ticid spider, Yllenus arenarius Menge, 1868. The prey and jump on it (Bartos 2000). Close ap- aim of the following research was to find out, proach is advantageous because of more pre- whether the spider can adjust its jumping dis- cise identification of prey and more precise tance when attacking prey of different ability to jumping and grasping. However, close ap- escape. proach is also connected with a high risk of Yllenus arenarius is a salticid inhabiting being noticed by the prey, compared to jump- open, sandy dunes of mainly Central and East- ing from longer distances. Jumping distance is, ern Europe (Żabka 1997). It seems to be a good therefore, a trade-off between several factors model for the study, because it is a predator (Bear & Hasson 1997). It is most profitable to stalking the prey in a habitat with very few attack the prey that can easily escape from a places to hide. Thus, it has to depend solely on longer distance, and the prey that cannot es- its cryptic coloration and prey approaching cape from a shorter distance. Since the spiders tactics. encounter different types of prey randomly, Y. arenarius hunts a wide variety of inverte- they should apply both of the behavioural tac- brates (Bartos 2000), which among other things tics flexibly and quickly. differ in mobility. Some of them can easily es- Flexibility in behavioural tactics, in order to cape (e.g. Diptera, Homoptera, Orthoptera) optimise the outcome is known as a conditional 34 European Arachnology 2000 strategy (Alcock 1993), and the phenomenon (height: 10 cm, Ø: 10 cm) with a 2 cm layer of has already been reported for salticids (Jackson dune sand on the bottom. Spiders from three 1978, 1992; Edwards & Jackson 1993, 1994; Bear age groups were used in the experiments: juve- & Hasson 1997). Jackson (1977ab, 1978) found, niles in first year of life (juv-I), juveniles in sec- that a male’s mating behaviour in Phidippus ond year of life (juv-II) and adults in third year johnsoni depends on female maturity and loca- of life (ad). Spiders were assigned to the age tion. Behavioural flexibility was also found in groups on the basis of their size and maturity the genus Portia, where spiders were found to according to a previously developed method tune the mode of hunting to specific condi- (Bartos 2000). To include all three age groups in tions, such as prey type and its location the study the data had to be standarised. There- (Jackson & Blest 1982; Jackson & Hallas 1986a, fore the jumping distance (JD) was divided by b; Jackson 1992). the abdomen length (AL) and the relative Most cases of predatory versatility have jumping distance (JD/AL) was further included been reported for the subfamily Spartaeinae. in the tests. Abdomen length was used to stan- However, other studies suggest that condi- darise the jumping distance to correct not only tional strategies are likely to be found also in for body size but also for the condition of dif- other salticids (Edwards & Jackson 1993, 1994; ferent spiders of the same age. Spiders in bad Bear & Hasson 1997). condition have shorter abdomens than spiders in good condition. The worse the condition the MATERIALS AND METHODS shorter jumping distance, therefore the use of A two-year-long diet analysis was carried out abdomen length as a standarising factor re- prior to the experiment. On the basis of that duces the influence of spider condition on the research five groups of insects were chosen for relative jumping distance (JD/AL). Abdomen the experiments. These were: Homoptera, Dip- length was measured with a stereomicroscope tera, Orthoptera, Thysanoptera and Lepidop- on living spiders. A linear relationship was tera larvae (Bartos 2000). Insects from the three found between jumping distance and abdomen former orders are capable of efficient escape, length (r = 0.70; df = 222; P = 0.001), which al- and were therefore regarded as prey with high lowed employing the relative jumping distance escape risk. Thysanoptera and Lepidoptera lar- for the analysis. Spiders as well as prey items vae are unable to move quickly and were re- were chosen randomly for experiments and garded as prey with low escape risk. In both used only once in the whole set of experiments. groups of different escape potential, the prey All the experiments were carried out within types which were most common in the spider’s a white cardboard arena (height: 15 cm, Ø: 20 diet, were chosen for the experiments. cm) with a 1cm sand layer on the bottom. At- By sweep-netting dune grass (Corynephorus tack behaviour was recorded with a camera canescens) prey items were collected either on placed above the arena and connected to a the day of experiment or the day before. They computer. A scale was also recorded, which were brought to the laboratory and kept indi- allowed measuring of the jumping distance. vidually. In order to reduce mortality of the The significance of the differences in jump- prey, insects were stored in a refrigerator ing distance was tested with one-way ANOVA (temp. 5 °C) and taken out 15 min. before the and the Tukey test with unequal sample sizes. experiment started. To test differences between the three age Spiders were collected on the day of experi- groups one-way ANOVA was applied. The ment or the day before in order to reduce the significance of skewness (G1) was also calcu- influences of rearing conditions on the spider’s lated. Data are presented as: mean ± SD. behaviour (Carducci & Jakob 2000; Bartos un- publ.). They were kept in glass chambers Bartos: Approach to prey in a jumping spider 35 RESULTS Age differences Prey differences There were no significant differences in relative Significant differences in relative jumping dis- jumping distance between spiders in different tance on the five types of prey were found age groups attacking Homoptera (one-way (one-way ANOVA: F0.05;4.219 = 85.56; P = 0.001). ANOVA: F0.05;2.64 = 1.57; P = 0.2), Diptera (one- Homoptera, Diptera and Orthoptera were at- way ANOVA: F0.05;2.74 = 1.53; P = 0.2) or Lepi- tacked from significantly longer distances than doptera larvae (one-way ANOVA: F0,05;2;14 = Thysanoptera and larvae of Lepidoptera (Fig. 1, 0.33; P = 0.7). Regarding other prey types, the Table 1). The preferences in jumping distance number of data in age groups was not big are also seen for relative jumping distance dis- enough to be tested. Thysanoptera were tribution (Fig. 1). In cases of hunting prey with hunted only by juveniles in the first year of life, a high ability to escape, all distributions are and Orthoptera were eaten only by adults. positive; they are significant, however, only for Diptera (G1 = 2.70) and Homoptera (G1 = 2.62) DISCUSSION (Table 2). Significant difference (P = 0.02) was The data suggest that Y. arenarius is able to dis- also found between relative jumping distance criminate between different types of prey and on Thysanoptera and Lepidoptera larvae applies different behavioural tactics to hunt (Table 1). them. Spiders jumped on Diptera, Homoptera and Heteroptera from significantly longer dis- 9 n=67 n=77 n=26 7 5 3 n=36 n=18 1 Min.-Max. Relative distance jumping 25%-75% Median -1 Homoptera Orthoptera Lepidoptera (larvae) Diptera Thysanoptera Prey taxa Fig. 1. Relative jumping distance (JD/AL) on 5 prey taxa. Homoptera, Diptera and Orthoptera are prey with high risk of escape, whereas Thysanoptera and Lepidoptera larvae are prey with low risk of escape. Table 1. Results of the Tukey test with unequal sample sizes showing significant differences (marked with asterisk) in pair-wise comparison between relative jumping distances (JD/AL) on different prey types. Lepidoptera Prey type Diptera Orthoptera Thysanoptera larvae Homoptera P = 0.9 P = 0.4 P = 0.0001* P = 0.0001* Diptera - P = 0.5 P = 0.0001* P = 0.0001* Orthoptera - P = 0.0001* P = 0.0001* Thysanoptera - P = 0.02* 36 European Arachnology 2000 Table 2.