The locomotory activity of both hungry and satiated individuals of Carabus hortensis L. was studied both in pine and beech forest by tracking the beetles by using a portable harmonic radar system. The degree of locomotory activity as well as walking distances was recorded for both males and females. Hungry males and females show a similar degree of activity. When satiated, locomotory activity of females is significantly lower than that of males in both types of forest. These differences in locomotory activity are considered in the context of male-to-female ratios in pitfall catches. It is proposed to use the male-to-female ratio for C. hortensis, and possibly other species, as an indicator of food accessibility in forest habitat: the more males in pitfall catches in relation to females, the better the food situation in the forest.
THE FOOD CONDITIONS for a species in its habitat are of basic importance for its reproduction and survival. Therefore, it is expected that with bad feeding conditions, the individuals of a species will show a behavior different from that under better feeding conditions. Especially for surface-dwelling carabids, the locomotory activity in the reproductive period would be changed, and such changes may differ again between the sexes in connection with the different role of the sexes in that period.
The majority of studies dealing with carabid beetles are carried out using pitfall traps as described by Barber (1931). The possibilities to interpret catch numbers often is subject to controverses, because the numbers in catches are an expression of density of the carabids in the field as well as of their activities. Some authors doubt that catch numbers can ever be a reasonable estimate of local population densities (Müller 1984). However, Baars (1979) and Den Boer (1979) independently came to the conclusion that catches in pitfall traps for each species summed over the entire year give reliable, although relative, estimates of the densities of the species under study.
Concerning the male-to-female ratio in the traps, the catches will give a good impression of the de facto ratio in the field, only if both sexes show the same degree of locomotory activity. If there would be differences in locomotory activity, the sex with the higher activity will be represented in the traps above average. For Pterostichus oblongopunctatus F., Szyszko (1976) showed that forest stands of different age and/or different management of the soil in the past are characterized by different proportions of males to females in the traps. For the same species, Szyszko et al. (1996) suggested that the worse the feeding conditions for the adults, the higher the percentage of females in the traps. Basic differences of activity in relation to feeding state of Pterostichus melanarius Illiger have been shown by Fournier and Loreau (2001, 2002). M#x00FC;ller (1984) demonstrated sex specific differences in locomotory activity with respect to degree of satiation during the reproductive period of carabid beetles in the laboratory.
The present article deals with the question of whether the locomotory activities in the carabid species C. hortensis L. differ between males and females in relation to the feeding state of the beetles. Individuals of C. hortensis show a range of sizes of ≈22-30 mm (Freude et al. 1976). The species is mainly night-active. It prefers coniferous forests and is surface-active from the middle of summer to the middle of autumn (Larsson 1939). The beetles are unable to fly; therefore, a change of location has to be realized by walking.
During 3 yr, we tried to study the activities of individuals as directly as possible by tracing these. We recorded the degree of activity in relation to inactivity as well as the walking distances carried out by active specimens.
Materials and Methods
Our studies were carried out in both pine and beech forest in the Forest District Tuczno, Poland, shown in Fig. 1. The pine forest is growing on former arable land, and at the time of the experiments the trees were ≈25 yr old. The surface was covered with a rather dense growth of grassy plants. C. hortensis regularly occurs in this forest. At the time of the experiments, the beech forest was 76 yr old. C. hortensis was not recorded here in the past. In comparison to the pine forest, the surface cover with grassy plants is insignificant.
Individuals for the experiments were collected in the pine forest by using dry pitfall traps without preservatives. Every morning, the beetles were taken from the traps and stored in the laboratory at 6°C. Some of the animals were fed with larvae of Cerambycidae or beef.
According to the technique described by Van der Ent and Van Dijk (1991) AA 109 diodes and 4 cm of copper antennas were attached to the elytra, by which each beetle marked in this way could be tracked by application of harmonic radars (Mascanzoni and Wallin 1986). Every time the experimental setup was as follows: beetles with attached antennas were transported to the forest and released at sunset with their heads in the north direction at sites permanently marked by stakes. Experiments were made only on nights without rain, and the beetles were released alternatively once in the pine stand and once in the beech stand. In each experiment, we tried to use four specimens simultaneously (two males and two females).
After release the location of each beetle was recorded every 20 min. The actual positions of the individuals were flagged using 15-cm nails with a paper attached, indicating the identity of the individual, and 1 h 20-min period of the observation. The route of the beetles was traced in this way until dawn. At dawn the individuals were taken back to laboratory and stored again until subsequent experiments. When a beetle was lost, it was replaced by another one that was treated in the same way.
The angles of direction in which the beetles moved as well as the distances covered were recorded. Angles were noted in relation to the north direction with a compass. Sometimes the beetles climbed trees. In such cases, only the distances covered in that 20 min were recorded.
The studies were carried out in the period of the most intense activity of C. hortensis in the pine forest. In 1992, all experimental beetles were satiated, but in 1993 experiments with hungry individuals were carried out, too. Beetles were considered to be hungry if they were stored without food for at least 3 d. In 1994, experiments only occurred in the beech forest with hungry individuals. Then, higher numbers of individuals were observed simultaneously with a maximum number of 11 specimens. These individuals were not removed at dawn but were left in the forest until next sunset.
Observation periods ranged from 21 August to 5 October in 1992, from 7 August to 28 September in 1993, and from 9 August to 18 September in 1994.
Whenever an individual changed its location during an interval of 20 min, the recording was assessed as "active beetle" or else as "inactive beetle." The recordings were summed up concerning activity and inactivity for males and females, respectively. These summations were carried out for satiated beetles in 1992 and 1993, hungry beetles in 1993, and hungry beetles in 1994. The data were subjected to χ2 tests (Sachs 1984). Differences concerning activity patterns of males and females between the two forests were compared for the respective groups.
To estimate the degree of activity, the percentage shares of active beetles in the sum of active and inactive beetles were calculated for the respective groups. A sex ratio of activity was calculated dividing the percentage of active beetles for males by the percentage of active beetles for females.
To interpret walking distances covered, the active beetles were compared carrying out Mann-Whitney U tests (Sachs 1984). Males were tested against females for the pine forest as well as the beech forest for satiated beetles in 1992 and 1993, hungry beetles in 1993, and hungry beetles in 1994.
During 3 yr, 212 nightly observations were made on single individuals of C. hortensis(Table 1). The number of active beetles and inactive beetles, respectively, are shown in Tables 2-4.In all experiments the number of inactive beetles was by far higher than the number of active beetles. Both in the pine forest and in the beech forest satiated males showed significantly higher activities than females (Table 2). Among hungry beetles studied in the beech forest in 1994, when the individuals were not removed from the field at dawn, males also were significantly more active than females (Table 3), whereas the hungry individuals observed in 1993 did not show significantly different activities between the sexes in either of the forests (Table 4).
aχ2 test, P < 0.05.
b χ2 test, P < 0.001.
aχ2 test, P < 0.01.
aχ2 test, not significant.
Table 5 shows which percentages of active beetles were found in these experiments together with the ratios between the activities of males and females. Percentages of active beetles ranged from 8.7% (satiated females in the pine forest in 1992 and 1993) up to 25.0% (hungry males in the pine forest in 1993).
In general, satiated males showed a distinctly higher percentage of active beetles than satiated females in both forests. The ratio of activities (1.87 in the pine forest and 1.66 in the beech forest) indicates a stronger dominance of walking activities in males. For both males and females, a higher percentage of active beetles was found in the beech forest as compared with that in the pine forest.
The distributions of the walking distances are very similar. All distributions mainly existed of short distances with some outliers to much greater distances. Median values range from 28.5 cm (hungry males in the beech forest in 1994) to 78.5 cm (hungry males in the beech forest in 1993). The shortest distance covered was 2 cm (satiated males in the beech forest in 1992 and 1993), whereas the longest distance observed was 860 cm (hungry males in the beech forest in 1994).
In general, the distances covered by males and females of active beetles did not differ significantly, with the exception of hungry females in 1994, which covered significantly greater distances than males (Mann-Whitney U test, P < 0.05).
If we assume that in the field the density of males is similar to that of females (de facto male-to-female ratio is 1), a differing number between the sexes caught in pitfall traps should depend on differences in locomotory activities between the sexes. When individuals are satiated, males would be dominant in the catches. The experiments discussed in this article would predict a male-to-female ratio of 1.87 in the pine forest and 1.66 in the beech forest. For hungry individuals, these values might change toward values balanced around 1.0 (male-to-female ratio of 1.2 in the pine forest in 1993, male-to-female ratio of 0.98 in the beech forest in 1993, and male-to-female ratio of 1.44 in the beech forest in 1994).
The data in this article suggest that the differences in male-to-female ratio depend mostly on the locomotory activities of females with respect to their feeding state. Females and males show similar activities when hungry. However, females reduce activities when satiated, whereas the activities of males stay at about the same level (in the pine forest) or become a little lower only (in the beech forest). These findings strongly agree with results obtained by Müller (1984) on three carabid species. In laboratory experiments he showed that during the reproduction period males are constantly highly active independent of their feeding state. However, females were highly active when hungry but much less active when satiated. Wallin and Ekbom (1994) detected sex-specific differences in movement velocities for Pterostichus cupreus L., P. melanarius, and Pterostichus niger Schaller. The velocity of females of all three species was influenced by differences in hunger level as well as by prey density, whereas such effects were not found in males.
This conclusion agrees with the male-to-female ratio in our data from 1994 in the beech forest and also regarding the difference between satiated beetles in 1992 and 1993 and hungry beetles in 1993. In 1994, the beetles were set in the field hungry, but unlike in 1993 they stayed there more than one night and the increased time for potential feeding may have shifted the results toward those of satiated beetles. Our results contradict those of Heydemann (1962a, b) who suggested that changes in male-to-female ratio in catches have to be interpreted as changes in the de facto ratio. This opinion is shared by Müller (1970).
Walking distances did not distinctly differ between the sexes. It seems that walking distances are of minor importance in this respect. This is of interest in so far that the two forests had different densities of surface cover with grassy plants. Dense vegetation is supposed to impede the movements of carabid beetles ("Raumwiderstand," Heydemann 1957, Greenslade 1964). If the phenomenon of Raumwiderstand would apply, in our study we should have to expect a significantly lower median of walking distances in the pine forest, which was not confirmed by our data. The walking distances were not statistically different with a median value of 53 cm in the pine forest and 43 cm in the beech forest respectively (Mann-Whitney U test, not significant). Therefore, we conclude that the different surface cover of the forests is of minor importance concerning catches in pitfall traps. The feeding state of the individuals is much more important.
Our results indicate that the locomotory activity of females may be a good indicator of food accessibility in the habitat for adults. This will be found in the male-to-female ratio in catches in pitfall traps. Male-to-female ratios of ≈1 in such catches will indicate habitats with a rather bad feeding situation, whereas a domination of males would point to better feeding conditions. According to this assumption, the lower male-to-female ratios for the beech forest in the current study indicate a worse food situation. We propose the male-to-female ratio of epigeic carabid species in pitfall catches to be useful as an indicator of food accessibility.
The present work arose in collaboration with many people. The availability of the harmonic radar was made possible by Paul Odam and Rikjan Vermeulen (Institute for Forestry and Nature Research, Leersum, Holland. We thank Piet Den Boer (Wüster, Holland) for sharp but benevolent critical reading and remarks that stimulated our activity. Two anonymous referees made valuable comments on the manuscript. We also thank Radomir Dyjak (Department of Landscape Architecture, Agricultural University of Warsaw, Warsaw, Poland) for help with preparing the figures; Jan Krzyszkowski (Tuczno Forest Administration, Poland) for help in organization work, and Mark Bastian (Langhorne, Pennsylvania) for editorial comments.