First evidence for an aposematic function of a very common color pattern in small insects

Many small parasitoid wasps have a black-orange-black (BOB) color pattern, which is usually present in both sexes. A likely function of this widespread pattern is aposematic (warning) coloration, but this has never been investigated. To test this hypothesis, we presented spider predators (Lyssomanes jemineus), both field-captured and lab-reared individuals, to a species with the BOB pattern and a congeneric all-black species in each of four scelionid genera (Baryconus, Chromoteleia, Macroteleia and Scelio). Each spider/wasp trial was filmed for 40 minutes under controlled conditions and three behavioral responses (detect, attack, avoid) were recorded in each of 136 trials, never using the same predator and prey more than once. In order to better understand the results obtained, two additional studies were performed. First, the reflection spectrum of the cuticle of the wasp and a theoretical visual sensibility model of the spider were used to calculate a parameter we called “absorption contrast” that allowed us to compare the perception contrast between black and orange in each wasp genus as viewed by the spider. Second, acute toxicity trials with the water flea, Daphnia magna, were performed to determine toxicity differences between BOB and non-BOB wasps. By combining the results from the three types of experiments, together with a statistical analysis, we confirmed that BOB color pattern plays an aposematic role.


37
Many small (<10 mm) parasitoid wasps have a black head, an orange (or reddish 38 orange) mesosoma and a black metasoma. This color pattern has been found in 39 species belonging to 23 families of Hymenoptera (including phytophagous sawflies), 40 and is especially common in neotropical scelionid wasps (Platygastridae; formerly 41 Scelionidae), but is also common in evaniid wasps from diverse biogeographic of aposematism [3], whereby predators learn to associate particular color patterns 48 with noxious chemical defenses, although this learning process is much more 49 complex than simply developing an aversion to specific types of prey [4]. In some 50 larger insects contrasting black and orange color patterns are known to serve as 51 aposematic (warning) coloration for potential predators, mostly vertebrates [5], and it 52 is possible that the BOB pattern serves as a warning pattern for smaller (invertebrate) 53 predators, although this has not yet been tested. 54 To test whether the BOB pattern also serves as aposematic coloration for predatory 55 invertebrates, we chose a common jumping spider (Salticidae) as the predator for visual acuity in Lyssomanes may be somewhat lower than that of other salticids, such 68 as Portia [11]. Lyssomanes species are diurnal foliage dwellers and two hunting 69 behaviors predominate: a sit-wait strategy followed by springing from the underside 70 of the leaf to the upper surface (they usually sit on leaves that are exposed to the sun 71 waiting to ambush prey on the upper surface), and a hunting behavior that consists 72 of exploring both sides of the leaf, actively searching for passing insects [12]. Thus, 73 Lyssomanes, like most salticids, can be characterized as a hunting spider that does 74 not use webs, and which executes behavioral responses such as: prey detection by 75 sight, stalking, and attacking by jumping towards the prey [13] [14].

77
The objective of this study was to determine if the BOB color pattern in small 78 parasitoid wasps is aposematic, through toxicity tests and behavioral analysis of 79 invertebrate predators that depend on vision to capture prey. Our hypothesis is that 80 the BOB pattern in scelionids and other small wasps is aposematic, acting as a signal 81 for vision-dependent invertebrate predators. One of our predictions was that when 82 comparing wasp species with the BOB color pattern with others having a black color 83 pattern, the former would show greater toxicity. A second prediction was that in 84 predator choice experiments all-black prey would be preferred over prey with the 85 BOB pattern. Also, experienced predators (collected in the field) would avoid the BOB 86 pattern more than laboratory-bred spiders that were never exposed to it.  with two lateral circular areas (10 cm diameter) with small holes for aeration covered 137 with an ultra-fine mesh. Including vegetation in the cages helps mitigate the effects 138 of captivity [17], and we found that the best results were obtained by using the same 139 plant species that harbored the egg sac in the field; water was provided in test tubes 140 plugged with cotton. We occasionally observed advanced stage juveniles preying on 141 younger stages, but this behavior was sufficiently uncommon that it was deemed 142 unnecessary to separate specimens individually during the life cycle. However, 143 cannibalism increased as the number of individuals per cage increased and therefore 144 no more than 10 individuals were maintained in a single cage; they were generally 145 separated during the fourth week after hatching, which is when they usually begin to 146 disperse.

147
A survival rate of 53% was achieved by providing young spiderlings and juveniles contacting the prey; 4) piercing and ingesting the prey; 5) prey detection followed by 165 withdrawal from the prey; 6) prey undetected or ignored. In the results the first two 166 were combined as "detect" (although detection was not always followed by stalking), 167 three and four were combined as "attack" (although jumping was rarely followed by 168 ingestion), and five and six were combined as "avoid".

169
To facilitate observations of such small specimens we used a white background on 170 the floor, but natural light was allowed to enter the remaining walls of the acrylic box.  2) To ensure that the predator was motivated to feed during testing, we fed the 182 spider and then held it without prey for eight days before subjecting it to 183 testing. Eight days also ensured that specimens were not weak,        In the experiments with lures the acrylic box was placed inside a frame made of 261 pressed wood and containing an automated mechanism for moving the lure (Fig 2).

262
The lures were moved by a motor mechanism that allowed for two-dimensional 263 movements. Each lure was connected to a moving part through a nylon thread,  In order to elucidate whether the spider can distinguish between orange (BOB) and 286 black, and to determine whether the orange color varies between genera, the method  We refer to these differences as "absorption contrasts", in the sense that the only 302 assumption we are making about the physiology of the visual system of the spider is 303 the type of photosensitive pigment present and its maximum absorption wavelength 304 (as described above).  The response variable in this case was constructed by registering whether L. 323 jemineus responded similarly to black and BOB lures coded as 1, and whether L.

Experiments of spiders with live prey
A total of 136 trials with three groups of spiders (68 field-juveniles, 51 field-adults and 17 captive-adults) were analyzed. For each of the 136 trials consisting of a wasp with a spider, the size of each was measured (Fig 3). Although there is size variation within each group, predator (spiders) and prey (wasps) fall within the same general range, with field-captured adult spiders being the largest.   For the multinomial logistic regression that was fitted to a simplified response, the 380 most common activity for each spider was classified as "detect", "attack" or "avoid".  When the behavior of all spiders is analyzed during the 40 min and in the different 409 time slots a greater complexity of behaviors is observed (Fig 5), specifically in that 410 the field adults usually differed from the adults raised in captivity. There is a clear 411 effect of experience, with captive adults showing a lower detection and attack 412 capacity than field spiders (including juveniles). This may be due to the fact that 413 captive spiders were reared on a simple diet and were thus responding to unfamiliar 414 prey. However, there is some evidence for an innate response by spiders in captivity, 415 in avoiding attack responses and aversion to BOB wasps (Fig 5C). in field juveniles and adults (Fig 5B). Captive adults avoid wasps with the BOB pattern 427 during most time periods, which differs from the behavior of field-caught spiders, 428 which showed equal avoidance of both colors of wasps (Fig 5C).  The reflectance spectra of six blends of black paint and six blends of orange paint 477 were compared with that of the black and orange color, respectively, of BOB-colored 478 Baryconus (Fig 8). The best matches in terms of similarity to the black and orange     other insects) and has evidently evolved independently on numerous occasions [1].

626
To the best of our knowledge, the present study provides the first evidence that this 627 common color pattern has an aposematic function in these small insects.