Final hiding place

First, we describe results regarding the final hiding place, where the tested female scorpions were found in the morning, at the end of the experiment. We interpret this shelter to be the scorpion’s final decision where to hide during the day (see Data acquisition). Tested shelters are abbreviated by naming their properties, e.g. a small, prey-scented hide-out is referred to as “small, prey”.

Intact females

When comparing neutral shelters, that is, without chemosensory cues present, female E. italicus significantly preferred larger hides over smaller ones (93.3%: 6.7%, Fisher-Exact: P < 0.001, n = 15, Fig. 1a). They also preferred the larger shelter when they had to choose between a large neutral shelter and a small one provided with conspecific scent. Tested animals chose the large neutral hideout significantly more often, independent of male (72.2%: 22.2%, Fisher-Exact: P = 0.005, n = 18, 1 animal not hidden) or female scent (89.5%: 5.3%, Fisher-Exact: P < 0.001, n = 19, 1 animal not hidden) being associated with the smaller shelter. This preference for larger shelters held true when the larger hiding place was scented and the small one remained neutral (male scent: 82.4%: 11.8%, Fisher-Exact: P < 0.001, n = 17, 1 animal not hidden; female scent: 73.3%: 13.3%, Fisher-Exact: P = 0.001, n = 15, 2 animals not hidden). In a direct comparison of male and female scent, females decided in favor of female scent, no matter if both retreats were large (61.1%: 27.7%, n = 18, 2 animals not hidden) or small (45.0%: 25.0%, n = 20, 6 animals not hidden). If scented and neutral shelters were both large ones, tested females hid more often underneath the male-scented than the neutral one (52.9%: 35.3%, n = 17, 2 animals not hidden), and preferences were about equal in the case of neutral versus female scent (50.0%: 43.8%, n = 16, 1 animal not hidden). When both shelters were small ones, females tended to hide in the neutral one compared to female-scented shelters (50.0%: 35.0%, n = 20, 3 animals not hidden) and they were nearly indifferent concerning male scent versus neutral shelters (42.1%: 52.6%, n = 19, 1 animal not hidden). When tested females had to decide between “large, neutral” and “large, prey”, they hid more often in the prey-scented one (36.8%: 52.6%, n = 19, 2 animals not hidden, Fig. 1a). In comparison with “small, prey”, the animals tended to prefer “large, neutral” (26.3%: 57.8%, n = 19, 3 animals not hidden). Rosemary scent on a large retreat appeared to have an aversive effect as the tested animals hid more often underneath the neutral retreat, no matter whether it was large (22.2%: 50.0%, n = 18, 5 animals not hidden) or small (31.3%: 56.3%, n = 16, 2 animals not hidden).

Female scorpions of M. eupeus hid significantly more often underneath a large final hiding place than underneath a small one in neutral scent comparison (84.5%: 7.7%, Fisher-Exact: P < 0.001, n = 13, 1 animal not hidden; Fig. 1b). Even when one of the shelters was fragrant with conspecific scent, they showed a tendency toward the larger shelter. Especially when the larger hideout was scented by males, females significantly favored that one over a small neutral one (76.9%: 15.4%, Fisher-Exact: P = 0.003, n = 13, 1 animal not hidden). If there were two large shelters, females hid more often in the male-associated hiding place than in the large neutral one (68.5%: 31.5%, n = 13), and more often underneath “large, neutral” than “large, female” (66.7%: 33.3%, n = 12). Interestingly, it was the other way round for small versus small retreat tests: Here, females apparently preferred female scent over neutral (57.1%: 21.4%, n = 14, 3 animals not hidden), and neutral over male scent (53.8%: 30.8%, n = 13, 2 animals not hidden). In direct comparison, female scent was chosen more often when presented against male scent on large shelters (69.2%: 30.8%, n = 13). Prey scent was preferred in both cases tested. On large final hiding places (69.2%: 30.8%, n = 13), and even on small shelters, where the preference was significant (88.9%: 0.0%, Fisher-Exact: P < 0.001, n = 9, 1 animal not hidden) when tested against large neutral shelters (Fig. 1b). Rosemary oil applied on large retreats (40%: 40%, n = 10, 2 animals not hidden) appeared to have almost no influence on shelter choice, as they were chosen as final hiding place almost equally compared to a large neutral shelter. When the neutral shelter was small, however, tested representatives tended toward the larger rosemary-scented shelter (33.3%: 55.6%, n = 9, 1 animal not hidden).

Fig. 1

Final shelter choice in pairwise comparison. Females of (a) E. italicus, (b) M. eupeus. Ordinates show choices in percent. Pie charts above choice pairs indicate hidden (black) versus non-hidden (white) animal numbers, sample sizes noted above pie charts. Tests are sorted by stimulus condition (neutral = grey, male scent = blue, female scent = red, scent of both sexes, prey scent = orange, rosemary oil = green, mineral oil = brown) and shelter size (large-large = dark colors; large-small, small-large, small-small = light colors). Statistically significant differences indicated by asterisks: ** P < 0.01, *** P ≤ 0.001

Impaired females

With one impaired sensory organ, E. italicus still chose “large, neutral” over “small, neutral”, although results were no longer significant (Fig. 2a). In case of both organs impaired, tested animals significantly preferred the larger shelter compared to the smaller one (56.2%: 18.8%, Fisher-Exact: P = 0,039, n = 16, 4 animals not hidden). Female scorpions showed a clear tendency toward the small neutral retreat compared to “large, rosemary oil” when their pedipalps had been impaired (Fig. 2b).

With their pectines or their pedipalps severed, M. eupeus showed no significant preferences anymore: in neutral comparison, the scorpions seemed to make no difference between large and small shelters. Concerning prey odor, tested animals with impaired pectines preferred “large, neutral” over “small, prey” (Fig. 2d). This tendency was even more obvious with impaired pedipalps. When both sensory organs were impaired, tested individuals preferred the larger (unscented) shelter in both tests.

Fig. 2

Final shelter choice of female (a-b) E. italicus and (c-d) M. eupeus. One or two sensory organs impaired, ordinates show choices in percent. (a, c) Large neutral tested against small neutral shelters, (b) large shelters perfumed with rosemary oil tested against small neutral shelters, (d) small prey-scented shelters tested against large neutral shelters. Pie charts show portions of hidden (black) versus non-hidden (white) animal numbers, sample sizes on top. Statistically significant difference indicated: * P < 0.05. Color code: grey = neutral, orange = prey, green = rosemary oil; dark colors = large shelters, light colors = small shelters

Females not hidden

Intact animals appeared to select a final hiding place less often when both offered retreats were small (Fig. 1, Tables S2, S6). It further mattered whether both small hidings were perfumed with conspecific scent or just one of them was perfumed with female or male scent, tested against neutral. Added rosemary oil in a large-large test combination also increased the number of females not choosing a final shelter. There was a tendency towards more animals not hidden at all when the sensory organs had been impaired (Fig. 2, Tables S4, S8). Especially when the pectines had been cut, many animals refused to choose a final hiding place.

Contact times

Recording duration of the experiments was 13 h, thus covering the entire night phase. The longest median total contact times with shelters during these hours ranged from over 6 h to nearly no contact at all. For better comparison, we calculated the contact times as percentages of total observation time, rounded to one decimal place. The situation “on top” was defined by clearly visible contrast of any body part against the shelter, ending when there was no visible contrast anymore. Scorpions were considered “underneath shelter” as soon as approximately half of the body disappeared underneath the shelter, ending when the scorpion was visible in total. “Total contact time” was calculated as the sum of “on top” and “underneath shelter”.

E. italicus, median total contact times

Regarding neutral size comparison, larger shelters reached significantly longer median total contact times than small ones (46.2%: 0.7%, paired t-test: t = -3.892, n = 15, P = 0.002, Fig. 3). When conspecific scent was added, tested females showed median 37.7% total contact time for the large male-scented shelter, which was significantly longer than 1.4% for the small neutral shelter (Wilcoxon Signed Rank Test: W = 127.000, n = 17, P = 0.001). They also showed a significantly longer median contact time with “large, female” compared to “small, neutral” (35.5%: 0.4%, paired t-test: t = -3.477, n = 15, P = 0.004). The larger hideout was also favored in the comparison of “large, neutral” versus small with conspecific scent, no matter whether associated with male (48.6%: 0.4%, paired t-test: t = 3.643, n = 18, P = 0.002) or female scent (45.4%: 0.3%, Wilcoxon Signed Rank Test: W = -122.000, n = 19, P = 0.012). For male versus female scent added to small shelters, there was a statistical significance concerning contact times in favor of female scent (1.1%: 0.3%, Wilcoxon Signed Rank Test: W = -120.000, n = 20, P = 0.024).

The large hideout with prey scent was probed for longer times than the large neutral shelter (Fig. 3). When the prey scent was associated with the small shelter and tested against “large, neutral”, median total contact times where about equal. For large shelters scented with rosemary oil, there were almost no median total contact times compared to neutral shelters.

Fig. 3

Median total times in contact with shelters in female E. italicus. Ordinate, percent of total observation time. Tests are sorted by stimulus (neutral, male, female, both sexes, prey, rosemary oil, mineral oil) and size of shelters (large-large, large-small, small-large, small-small). Statistically significant differences are indicated by asterisks: * P < 0.05, ** P < 0.01, *** P ≤ 0.001. Error bars omitted for clarity (see Tab. S1). Sample sizes on top. Color codes: grey = neutral, blue = male, red = female, orange = prey, green = rosemary oil, brown = mineral oil; dark colors = large shelters, light colors = small shelters

E. italicus, median times “on top” and “underneath”

Walking across the presented shelters will give the tested female scorpions a first impression of the options provided. Therefore, we wanted to know the time the tested animals spent “on top” of a hiding place in comparison to the time span exploring the shelters in more detail by hiding “underneath” the plant saucers serving as shelter. We also calculated contact time in percentages of total recording time.

Median times on top of hiding places ranged below 1% of total recording time (Fig. 4a). As these time spans were comparably short and did not add much to the total contact times, we only describe striking differences like changed proportions. The first difference to the previously presented results was a longer contact time with the small neutral shelter compared to the larger neutral shelter (0.3%: 0.2%, n = 15). Significant differences in favor of large neutral hideouts were observed in four instances: in tests against “small, male” (0.4%: 0.2, Wilcoxon Signed Rank Test: W = -133.000, n = 18, P = 0.002), “small, female” (0.28%: 0.27%, Wilcoxon Signed Rank Test: W = -138.000, n = 19, P = 0.004), “small, prey” (0.4%: 0.2%, Wilcoxon Signed Rank Test: W = -152.000, n = 19, P = 0.001), and “large, rosemary oil” (0.3%: 0.1%, Wilcoxon Signed Rank Test: W = -129.000, n = 18, P = 0.003). Also, the median time on top of shelters was significantly longer for “large, female” (0.5%) compared to small neutral hideouts (0.3%, paired t-test: t = -4.475, n = 15, P < 0.001).

Median contact times underneath shelters were not much different concerning statistical significances compared to median total contact times, as noted above (Fig. 4b, Table S2).

Fig. 4

Median times in contact with shelters in female E. italicus, plotted separately for times (a) on top of shelter and (b) underneath shelter. Ordinate, percent of total observation time. Tests are sorted by stimulus (neutral, male, female, both sexes, prey, rosemary oil, mineral oil) and shelter size (large-large, large-small, small-large, small-small). Statistically significant differences indicated by asterisks: * P < 0.05, ** P < 0.01, *** P ≤ 0.001. Error bars omitted for clarity (see Table S1). Sample sizes on top. Color codes: grey = neutral, blue = male, red = female, orange = prey, green = rosemary oil, brown = mineral oil; dark colors = large shelters, light colors = small shelters

E. italicus, impaired sensory organs

With impaired sensory organs, median contact times mostly ranged below 1% of the total observation time. The only exception was the neutral comparison of shelter size when the tested animals had their pedipalps impaired (Fig. 5a). In this case, maximum median contact times reached 2.7% of total observation time. Proportions of the median total contact times still showed the tendency in favor of “large” over “small” in both test situations, neutral and rosemary oil. Only when the pedipalps were impaired, tested females spent longer time in contact with the small neutral shelter, especially in the neutral test combination. In the latter case, the longer median time in contact with the small shelter seems to result from the time spend underneath the small shelter.

Fig. 5

Median contact times of female E. italicus with impaired sensory organs. Ordinate, percent of total observation time. Shelter combinations, (a, c, e) large versus small neutral, and (b, d, f) “large, rosemary oil” versus “small, neutral”. (a, b) Total time in contact with shelter, (c, d) time on top of shelter, (e, f) time spent underneath shelter. Error bars omitted for clarity (see Table S3). Sample sizes on top. Color codes: grey = neutral, green = rosemary oil; dark colors = large shelters, light colors = small shelters

M. eupeus, median total contact times

Females of M. eupeus spent significantly longer median total times in contact with “large, neutral” versus “small, neutral” (paired t-test: t = 3.366, n = 13, P = 0.006, Fig. 6). Further, the combination of “large, male” versus “small, neutral” was statistically significant in favor of “large, male” (paired t-test: t = 3.434, n = 13, P = 0.005). Additionally, females spend significantly longer total median time in contact with a small but prey-scented shelter versus a large neutral one (paired t-test: t = -3.082, n = 9, P = 0.015).

Fig. 6

Median total times in contact with shelters in female M. eupeus. Ordinate, percent of total observation time. Tests are sorted by stimulus (neutral, male, female, both sexes, prey, rosemary oil, mineral oil) and shelter size (large-large, large-small, small-large, small-small). Statistically significant differences indicated by asterisks: * P < 0.05, ** P < 0.01. Error bars omitted for clarity (see Table S5). Sample sizes on top. Color codes: grey = neutral, blue = male, red = female, orange = prey, green = rosemary oil, brown = mineral oil; dark colors = large shelters, light colors = small shelters

M. eupeus, median times “on top” and “underneath”

For the median contact times spend on top of hiding places, statistically significant differences were present in favor of “large, neutral” in comparison to “small, neutral” (Wilcoxon Signed Rank Test: W = -91.000, n = 13, P < 0.001; Fig. 7a, Table S6). Tested animals also spent longer median times on top of “large, male” than on “small, neutral” (paired t-test: t = 3.753, n = 13, P = 0.003). The same held true when the large hideout was female-scented (paired t-test: t = 3.619, n = 13, P = 0.004). Median times spent on top of “large, neutral” were also significantly longer compared to “small, male” (paired t-test: t = -3.340, n = 13, P = 0.006). The individuals spent even more median time on top of “large, rosemary oil” than on top of “small, neutral” (Wilcoxon Signed Rank Test: W = -43.000, n = 9, P = 0.008). As the times on top of shelters ranged in minutes and therefore represent less than 1% of the total time, there were no notable differences between the median total times and median times spend underneath a shelter (Fig. 7b).

Fig. 7

Median times in contact with shelter in female M. eupeus, plotted separately for (a) times on top of shelter and (b) times underneath shelter. Ordinates percent of total observation time. Tests are sorted by stimulus (neutral, male, female, both sexes, prey, rosemary oil, mineral oil) and shelter size (large-large, large-small, small-large, small-small). Statistically significant differences indicated by asterisks: * P < 0.05, ** P < 0.01, *** P ≤ 0.001. Error bars omitted for clarity (see Table S5). Sample sizes on top. Color codes: grey = neutral, blue = male, red = female, orange = prey, green = rosemary oil, brown = mineral oil; dark colors = large shelters, pale colors = small shelters

M. eupeus, impaired sensory organs

The maxima of median total contact times with impaired sensory organs in M. eupeus ranged below 12% of total observation time (Fig. 8). Only when both, pedipalps and pectines were severed, tested females showed a tendency toward longer median total contact times with “large, neutral” compared to “small, neutral”. This tendency became significant for longer times spend on top the large shelter compared to the small shelter (0.2%: 0.1%, n = 10, P = 0.014, Table S8). Animals spent more time at the large neutral hideout than at the prey-scented one with impaired pectines (8.9%: 0.1%, n = 6, P = 0.031) as well as when both organs had been impaired (8.6%: 0.6%, n = 7).

Fig. 8

Median contact times of M. eupeus females with impaired sensory organs. Ordinates, percent of total observation time. (a, b) Total time in contact with shelter, (c, d) time on top of shelter, (e, f) time spent underneath shelter. (a, c, e) Left column, neutral test large versus small; (b, d, f) right column, “small, prey” tested versus “large, neutral”. Statistically significant differences indicated by asterisks: * P < 0.05. Error bars omitted for clarity (see Table S7). Sample sizes on top. Color codes: grey = neutral, orange = prey scent; dark colors = large shelters, light colors = small shelters

There was a general difference between the two scorpion species tested. Female M. eupeus investigated the shelters for longer median times by hiding underneath (see Figs. 4b and 7b), and thus also showed longer median total times (see Figs. 3 and 6). E. italicus, by contrast, spent longer median times on top of the shelters, while showing shorter median total contact times.

Congruence total contact time and final hiding place

Finally, we compared for congruence the total contact times per shelter with the final hiding place per animal, if it had chosen a final hiding place. We expected the total contact time for the final hiding shelter to be longer than the total time in contact with the competing shelter. It turned out that, if a scorpion had chosen a final hiding place, it was the shelter the animal had had the longest contact with in 85.6% of the observations for intact E. italicus and in 87.7% for intact M. eupeus.