Growth rate of parasitized and non-parasitized caterpillars from oak trees in France and Poland

Overview We collected and reared spring-feeding caterpillars from the canopy of mature oak trees (Q. petraea, Q. robur and hybrids) in the Forêt de Rennes in Brittany, western France (2012 and 2019), and the Puszcza Zielonka forest in western Poland (2021 and 2022). Data on caterpillar growth were collected with other aims in mind, and methods differed between locations and years. Caterpillar rearing in France 2012 We sampled branches from 19 oak trees (Q. petraea - Q. robur hybrids) shortly after budburst [11th April to 11th May 2012 as timing of budburst varies between trees]. We transported the branches in plastic bags to the lab and reared the associated caterpillars individually in Petri dishes at room temperature using locally obtained Q. robur leaves. We photographed the caterpillars every alternate day with a standard and then determined their length from the photographs (24th April to 24th May). We assigned the caterpillars to surrogate species and linked these later to species using individuals that we reared to adulthood and identified. We noted when a parasitoid exited from a caterpillar or if the caterpillar pupated. Caterpillar rearing in France 2019 We collected branches from 24 oak trees (Q. petraea, Q. robur) in spring (26th March to 23rd May 2019) and reared the caterpillars in the lab. We reared all caterpillars separately in Petri dishes in a climate room mimicking the average diurnal change in temperature, daylight, and relative humidity of the study forest. We fed the caterpillars with freshly emerged Q. robur leaves. In addition to classifying caterpillars into surrogate species and linking these to reared adults, we used DNA barcoding to provisionally identify caterpillars that did not reach adulthood. The caterpillars were weighed on a semi-microbalance on average twice a week. Caterpillar rearing in Poland 2021 We collected branches from the canopy of four sessile oak trees (Q. petraea) shortly after budburst (May 14th 2021). We cut the branches into pieces and transported them to the laboratory in plastic bags. The next day, we searched the branches for caterpillars and placed each individually in a 50 ml cup. Operopthera brumata (L.) (Geometridae) caterpillars could be putatively recognised, and other morphologies were given surrogate-species names. We reared the caterpillars individually on leaves from these trees in a climate chamber with a constant temperature of 20°C and relative humidity of 75%. Three days later, we removed the frass and placed a new leaf inside each cup. We weighed the caterpillars the day after collection before adding food, and six days later, as they were transferred to a new cup. Thus, on both occasions, caterpillars could not feed for a few hours before weighing, removing variation caused by stomach content. Caterpillars were henceforth fed locally obtained Q. petraea leaves, and we noted death, parasitoid exit, pupation, and eclosion dates. We identified all eclosed Lepidoptera to species using dissection of genitalia when needed to then link caterpillar surrogate species to species. Caterpillar rearing in Poland 2022 The study in 2022 was similar to that in 2021, but sampling ten trees (six on May 11th and four on May 16th), and caterpillar mass change was measured over a five-day period. Unfortunately for the current purpose, this time we avoided rearing caterpillars that appeared parasitised. Data analysis To estimate caterpillar mass from their length (data from France 2012), we assumed that caterpillars had a cylindrical shape and thus utilised the formula ‘π × (height / 2)2 × length’ to calculate their volume. For each species, we inferred height from length by determining the height-length ratio from lateral photographs of caterpillars. We thus estimated that the height-length ratios were 0.1563 for Cosmia trapezina (L.) (Noctuidae), 0.1364 for O. brumata, and 0.1507 for Orthosia cruda (Den. & Schiff.) (Noctuidae). We finally assumed that the mass density of caterpillars is similar to water, to translate estimated volume into estimated body mass. Quantifying caterpillar growth rate is complicated because caterpillar growth occurs in phases, and absolute mass gain increases with body size. Specifically, growth pauses during moulting (change of instar), and caterpillars lose considerable body mass while preparing to pupate [23]. Detailed studies showed that caterpillar growth rate can be approximated as ‘(final mass-initial mass)1/3 / development time’ [including the moulting, Supplement 1]. We visualised the growth curves of oak-feeding caterpillars in our study for individuals with at least three size measurements (France 2012 and 2019; Supplement 1). To calculate growth rate, we took the first measurement as the initial mass. In an attempt to avoid the mass loss associated with pupation, we selected as final mass the measurement before mass gain decreased when multiple measurements were available for an individual (2012 and 2019; Supplement 1). For 2012, we also excluded caterpillar measurements that were longer than 16 mm, as caterpillars then approach pupation. When we had only two measurements (Poland 2021, 2022), we used a short period (five days in 2022, six in 2021) over which to calculate growth rate to avoid inclusion of mass loss in preparation for pupation. We finally excluded caterpillars that lost rather than gained mass from analysis. Caterpillars that experienced mass loss may have been pupating or unhealthy, but possibly also included instances of mass loss induced by parasitoids. We also excluded the few instances of dipteran parasitoids, as they might not manipulate caterpillar growth.