Effects of structural heterogeneity of a laboratory arena on the movement patterns of adult Eriopis connexa and Hippodamia variegata (Coleoptera: Coccinellidae)

We asked if the structural heterogeneity of a laboratory arena differentially affected the abandonment of the plot, resi­ dence time, locomotory rate, pause duration and turning rate of adult Eriopis connexa and Hippodamia variegata (Coleoptera: Coc­ cinellidae). We simulated an increase in heterogeneity by distributing vertically-oriented toothpicks in a circular arena: one control (without toothpicks), one uniform plot (toothpicks every 1 cm) and one random plot (randomly distributed toothpicks). No food was provided inside the arena. Coccinellids were released individually in the centre of each plot and their movement was videotaped. Fewer H. variegata adults left the plots as these become more heterogeneous. E. connexa did not discriminate between plots in their residence time, but H. variegata remained longer in the uniform and longest in the random plots. H. variegata only stayed longer than E. connexa in the random plots. This resulted because adults of H. variegata were stationary for longer periods, moved more slowly and less linearly and explored a higher number of toothpicks than adults of E. connexa in more heterogeneous environments. Thus, the physical structure of the environment differentially affects the movement patterns of insects. Eriopis connexa seems to be less sensitive to structural heterogeneity than H. variegata.


INTRODUCTION
Environments vary in structural heterogeneity.One way that organisms perceive and track the structural heterogeneity of environments is through their movement behaviour (Wiens et al., 1997;With et al., 1997).As the environment changes in structure, the resident organisms may also change their move ment pattern.Different species, or even different life stages of the same species, may perceive the environment in different ways, altering their movement patterns accordingly (With, 1994).Animal movement may determine the dynamics and spa tial distribution of entire populations and communities (Wiens et al., 1993a;Turchin 1998), as well as the outcome of many indi vidual ecological processes such as searching for food, oviposition sites, mates or shelter (e.g.,Jones, 1977).
The effect of the spatial distribution ofbiotic elements such as food resources or natural enemies on movement behaviour has been extensively studied, both theoretical and empirically (e.g.Ginsberg, 1986;Krakauer & Rodríguez-Girones, 1995).Never theless, the physical structure of the environment may modify this behaviour.Certain elements within the environment may facilitate movement, while others may interfere with this process (With, 1994;Ims, 1995;Wiens et al., 1997).This may be more important for ground moving animals than for flying ones.For instance, bare ground increases the speed of move ment of the tenebrionid beetle, Eleodes obsoleta, compared with individuals in patches ofgrass (Wiens et al., 1997).
Coccinellids are important natural enemies of pest species, particularly aphids, and have been used frequently as biological control agents (Obrycki & Kring, 1998).They spend most of their time searching for patches of aphids or scanning the envi ronment, using visual and olfactory cues for orientation (Bell, 1991;Ferran & Dixon, 1993;Lambin et al., 1996).They show area-restricted search in prey patches, but when prey are not available they assume an extensive searching behaviour, turning less frequently and moving more rapidly (Carter & Dixon, 1982;Ferran & Dixon, 1993).However, their searching behaviour (i.e., extensive vs. intensive search) may be influenced by the complexity of the substrate (e.g., plant architecture or the pres ence of trichomes, Obrycki & Kring, 1998), and not just by the presence of prey.
In this note, we experimentally analyse whether the structural heterogeneity of a laboratory arena, an artificial setting, differ entially affects the movement patterns of the adults of two coccinellid species: Eriopis connexa (Germ.)and Hippodamia variegata (Goeze).In order to identify the isolated effect of structural heterogeneity on the movement patterns of these coccinellid species, we excluded most other factors that also affect their movement in the environment (e.g., for a review see Ferran & Dixon, 1993).

MATERIAL AND METHODS
Eriopis connexa is a neotropical species native to Chile, whereas H. variegata was introduced to Chile for the biological control of cereal aphids (Zuniga et al., 1986).Both species are quite common, slightly elongated in shape and are 4-6 mm long.
We simulated three kinds of environments with increasing structural heterogeneity by setting 277 toothpicks, 1 cm high, in a white circular area 20 cm in diameter (i.e., plot): one control plot (without toothpicks), one uniform plot (toothpicks distrib uted every 1 cm), and one random plot (toothpicks randomly distributed) (Fig. 1).Toothpicks were supposed to alter coccinellid movement and may mimic a potential unit of foraging (e.g., a grass stick).The random plot was assumed to be more heterogeneous than the uniform one as random positions of toothpicks should result in a lower probability of finding any one of them.The average (variance) nearest distance between toothpicks was 1.00 cm (0) , and 0.61 cm (0.20) for the uniform and random plots, respectively.Coccinellids were collected from alfalfa crops and were kept overnight in the laboratory, with no food available.For testing the effect of structural heterogeneity on movement behaviour, coccinellids were released individually in the centre of each experimental plot (in a release area 4 cm2) (Fig. 1).Their move ments were videotaped with a VHS-C Panasonic NV-S750PN camera for up to 300 s or the exit of the insect from the plot.The tracks were reproduced in transparencies, identifying the position of the insect every 5 s (Wiens et al., 1993b).We used fixed-time intervals to define moves following Kareiva & Shigesada (1983).This time interval was selected because, although within approximately 5 s coccinellids reached the edge and abandoned the control plot (setting the highest time interval possible), in the other two treatments coccinellids stayed longer (sometimes more than 300 s), and with a frequency of 5 s we obtained a high number of consecutive movements (up to 60).
All experiments were carried out between 1300 and 1700 hr in a closed room, with artificial lighting.Temperatures varied between 22.4°C and 29.9°C.As these different temperatures may affect the movement patterns of insects (Ferran & Dixon, 1993), experiments for different treatments and species were interspersed in time, taking the possible effect of temperature as an experimental random error.Different individuals were used in each experiment, with a total of 48 individuals per species recorded (i.e., 16 in each treatment).The response variables analysed were: (a) percentage of insects that abandoned the plot before 300 s (usually insects stoped moving before that time period); (b) residence time inside the plot (s); (c) stop duration: percentage of the total time of each experiment that the insect remained stationary (this variable was measured for discrimi nating whether the residence time was due to constant move ment of the insect inside the plot or to long stationary periods); (d) locomotory rate (mm/s): D displacement distance every 5 s / total time inside the plot; (e) turning rate (deg/5 s): turning angles were measured between successive moves; this variable evaluates the linearity of movements; (f) number of toothpicks explored by coccinellids in the uniform and random plots.The percentage of insects that abandoned the plot before 300 s was analysed by y2 tests, variables b), c), d), and e) were analysed using a two factor MANOVA and post hoc Tukey HSD tests.Finally, the number of toothpicks explored by coccinellids was analysed using a two-way ANOVA.

RESULTS
Two out of 98 experimental individuals flew away from 25.0c 13.6 < 0.05 patches and were not considered in the analysis as we were interested in the effect of spatial heterogeneity on the walking movements of adult coccinellids.Structural heterogeneity dif ferentially affected the percentage of individuals that left a plot (Table 1).All individuals from both species abandoned the con trol plot before 8 s, but fewer individuals exited the plot in the other two treatments.This pattern was more evident in H. varie-  2, Fig. 2a).In the uniform and random plots, H. variegata remained stationary for longer periods of time than in the control plots.H. variegata had longer stationary periods, and moved significantly slower than E. connexa in the random plot (Table 2, Figs. 2b).Both species displayed a lower locomotory rate in the uniform and random plots, and H. variegata moved significantly slower than E. connexa in the random plot, but not in the others (Table 2, Fig. 2c).H. variegata had a higher turning rate in more heterogeneous plots and also when com pared to E. connexa.This was particularly so in the random plot, whereas E. connexa maintained the same turning rate in the three treatments (Table 2, Fig. 2d).
In the uniform and random plots, both species not only walked on the ground but also climbed the toothpicks, probably searching for food.H. variegata explored more toothpicks than E. connexa [F(1,60) = 12.67, P < 0.0001], but each species explored the same number of toothpicks in the uniform and random plots [F(1,60) = 0.007, P = 0.93].

DISCUSSION
The structural heterogeneity of the experimental arena differ entially affected the movement patterns of the experimental insects.As the arena become more heterogeneous, both species changed their movement patterns, but the change was stronger in H. variegata, as E. connexa was significantly affected by het erogeneity only in its locomotory rate, which diminished in the random plots.Changes in the movement patterns of insects related to environmental heterogeneity have been observed pre viously in field experiments (Crist et al., 1992;Wiens et al., 1997).Nevertheless, in field experiments many of the variables that determine the heterogeneity of an environment may covary.In our lab study, all factors other than presence and pattern of toothpicks were experimentally controlled, except for tempera ture which was considered as an experimental random error.Thus, the movement patterns of the coccinellids probably changed because of the variable physical structure of the arena.
The relationships between the structure of the environment and movement behaviour may depend on spatial scale (Wiens et al., 1993a) and on the life process involved (e.g., foraging, patch searching, dispersal, migration; Ims, 1995).Habitat selection by coccinellids may operate at a large scale, usually made by flying insects; thus, heterogeneity at this scale should be considered for studying this phenomenon.Food searching within habitats oper ates at a much finer spatial scale, usually according to the spa tial distribution of resources.At this spatial scale, coccinellids usually search their food by walking.The spatial scale used in this study attempted to simulate the foraging scale of coccinellids, and at this small spatial scale their movement patterns were influenced by the physical structure of the environment.
Our results also show that the effect of environmental struc tural heterogeneity is species-specific.In the more heteroge neous random plots, fewer individuals of H. variegata left the plots and they stayed longer inside these plots due to longer sta tionary periods, slower and less linear movements, and a higher number of toothpicks explored, adopting a more intensive search behaviour, compared to E. connexa.Thus, E. connexa seems to be less sensitive to environmental structural heteroge neity than H. variegata, and therefore may have a greater poten tial to survey a microhabitat.This could be advantageous since

Control
Uniform Random E. connexa may aggregate more easily in rich patches of prey or quickly abandon patches of low prey density, independently of the structural heterogeneity of the environment.From this point of view, E. connexa may be a better agent for biological control than H. variegata.However, our results come from a totally artificial setting, considering only one of a long series of factors affecting the movement of coccinellids, such as prey and host plant cues (Ferran & Dixon, 1993;Lambin et al., 1996).There fore, these relationships should be tested in more realistic experiments in the field.

Fig. 1 .
Fig. 1.Control, uniform, and random plots simulated with toothpicks.Each filled circle represents a toothpick.The square in the centre of each plot is the area were the insects were released.

Table 1 .
Percentage of individuals of E. connexa and H.

Table 2 .
Results of MANOVA on effects of species and structural heterogeneity treatments for residence time, stop duration, locomotory rate and turning rate of adult coccinellids E. connexa and H. variegata.whereasignificantlylowernumber of individuals left the uniform and random plots, compared to the control plot and to E. connexa (Table1).While the residence time of E. connexa was not significantly affected by heterogeneity, H. variegata remained longer in the uniform than in the control plot and longest in the random plot (Table2, Fig.2a).Moreover, H. variegata remained longer than E. connexa only in the random plot (Table