Data from: Host identity, nest quality, and parasitism strategy: influences on body size variation in parasitoid bees and wasps
Data files
Jan 23, 2025 version files 44.45 KB
-
OIK-11052.R1_data.zip
25.01 KB
-
OIK-11052.R1_R_script.zip
16.79 KB
-
README.md
2.64 KB
Abstract
Body size determines mobility and fitness across taxa in various ways. Yet, drivers of body size in higher trophic invertebrate, especially parasitoids, including intra- and interspecific variations, are poorly understood due to complex interactions between parasitoid behaviour, the environment, and their hosts. We measured the body size of 393 individuals of four parasitoid species (collected from 2220 parasitized brood cells) sampled with trap nests for cavity-nesting bees and wasps in the Southern Black Forest, Germany. We related parasitoid body size to the size of 15 host species and the diameters of their nests along four environmental gradients (proportion of conifers, canopy cover, structural complexity, herb cover, and deadwood diameter). Host identity, nest diameter, and to a lesser extent, size differences within host species, were primary drivers of parasitoid body size, albeit responses varied among parasitoid species. For instance, when the host Black Wood Borer wasp, Trypoxylon figulus doubled in size, the Ichneumon wasp, Nematopodius debilis (parasitizing the host directly) increased by 37% in size, while the Blue Cuckoo wasp, Trichrysis cyanea (parasitizing food resources) increased by only 8%. Across host-parasitoid species combinations, parasitoid size corelated weakly with host size, and environmental gradients did not significantly influence host or parasitoid body size. Our findings highlight the primary factors influencing body size, with host identity and nest diameter emerging as influential factors within and between parasitoid species, although not uniformly. In contrast, the relationship between the top trophic level (parasitoids), the lower trophic level (hosts), and host size, with environmental gradients were less influential. Considering the environmental variables that directly affect body size, such as microhabitat conditions and biotic interactions, may further clarify the dynamics shaping the variation in body size at higher trophic levels and should be considered in future studies addressing how land management influences multitrophic interactions.
README: Data from: Host identity, nest quality, and parasitism strategy: influences on body size variation in parasitoid bees and wasps
https://doi.org/10.5061/dryad.rfj6q57m6
Description of the data and file structure
This README file describes the data package accompanying the associated publication. The folder *OIK-11052.R1_data.zip **containing six-data files in .csv format that were used in analysis, and folder **OIK-11052.R1_R_script.zip * containing R files to reproduce the analysis of above mentioned publication.
Files and variables
File: OIK-11052.R1_data.zip
Files list (files found within OIK-11052.R1_data.zip):
canopycover.csv
deadwood.csv
nest_data.csv
prop_conifer.csv
ssci&alttude.csv
understorey%.csv
Description:
1. canopycover.csv: data for thee mean value of canopy cover calculated in ImageJ using overhead hemispherical photos obtained at each sampling site.
2. deadwood.csv: data for the measurement of diameters at breast height (DBH) of standing deadwood above 7 cm in diameter were obtained during plot-level inventories.
3. nest_data.csv: data includes list of parasitoid species, parasitoid sex, body length of parasitoid in mm, list of host species, and the body length of host in mm. NA in the host length variable justify for non availale speciments for a direct measurements.
4. prop_conifer.csv: data for the proportion of coniferous trees were calculated from the inventory basal area of coniferous trees at each sampling site.
5. ssci&alttude.csv: The remotely sensed indices stand structural complexity index (SSCI) data, were obtained from terrestrial laser scanners performed at the northwest and southeast corners, as well as the centres of the sampling site.
6. understorey%.csv: Includes data on herbaceous covers and understory plant species richness, which were estimated from 5x5 m subplots. Missing data: NA
File: OIK-11052.R1_R_script.zip
File list files found within OIK-11052.R1_R_script.zip):
Revised_main_code_OIK-11052.R
Revised_main_plot_OIK-11052.R
Description:
7. Revised_main_code_OIK-11052.R: This is R file reproduces the analysis of the above publication.
8. Revised_main_plot_OIK-11052.R: this is R file reproduce the plot and visualisation of the above publication.
NOTE: all data in OIK-11052.R1_data.zip were measured at plot level and those each of the data contains "plot_ID" indicating the number of plots which data were collected.
Code/software
Primary software: R programming language software, ideal for processing and visualizing the data.
Methods
Study Area
The study was conducted in the Southern part of the Black Forest, Baden-Württemberg, Germany, as part of the Research Training Group 2123 - Conservation of Forest Biodiversity “ConFoBi” (Storch et al. 2020). This part of the Black Forest consists of mixed temperate forests within a low mountain range, spans roughly 5,000 km² (75% forests) and rises from 120 to 1,493 metres above sea level (m.a.s.l.). Norway spruce (Picea abies (L.) H. Karst.) account for about 42.8% of trees in the forest, and is especially prevalent in the northern and eastern regions as well as higher elevations. In the Southern and Western region of the Black Forest, silver fir (Abies alba Mill.) and European beech (Fagus sylvatica L.) comprise greater proportions of forest stands (18.5% and 15.3%). The forest has been managed under a close-to-nature forest management system since the 1990s. However, an increasing focus on conserving biodiversity led to the introduction of a retention forest program in 2010 (ForstBW, 2016), requiring state-owned forests and encouraging private forests to integrate deadwood and habitat tree groups—around 15 trees per 3 hectares—into their management strategies (ForstBW, 2016).
In the Black Forest, 134 one-hectare plots (mean distance between plot centres is 750m) were distributed across strictly protected forest reserves, and multi-functional forests actively managed by the State Forest Service (Figure 2). The selection of study plots was guided by two design gradients: (1) landscape-scale forest connectivity, determined by the percentage of forest within a 25 km² area surrounding the plot centres, and (2) retention-related forest structure at the plot level, including the richness of habitat tree species and deadwood per hectare. This approach ensured the representation of a broad spectrum of conditions prevalent in European montane forests. Plots were pre-selected based on criteria such as topography, forest stand age, absence of water bodies, and human infrastructure (Storch et al. 2020) to minimise variation due to confounding factors. Further verification was conducted to confirm the absence of forest operational activities, such as harvesting and road construction, during the ConFoBi funding periods.
Study species
The Wild Carrot wasp, Gasteruption assectator (Linnaeus, 1758) (Gasteruptiidae), parasitizes solitary bees that nest in cavities (Johansson and van Achterberg 2016). Larvae of G. assectator is also known as predator-inquiline, initially consume the host larva of the Yellow-Faced bee Hylaeus spp., and subsequently feed on the host’s provisions (similar to idiobiont kleptoparasitoid behaviour), such as pollen and nectar (Bogusch et al. 2018). This species is commonly found across Europe and thrives in a wide range of habitats, from agricultural landscapes to forest ecosystems (Bogusch et al. 2018).
Ichneumon wasp, Nematopodius debilis (Ratzeburg, 1852) (Ichneumonidae) is an idiobiont ectoparasitoid species found in the nest of Crabronid wasp species (Trypoxylon spp.) specialised in hunting spiders. The adult female paralyses the host during oviposition; subsequently, the parasitoid larva consumes the host and develops from outside its body. This species is distributed across Europe, but there is a lack of information on the specific habitat types in which it occurs (Broad et al. 2018).
Two Cuckoo wasp species, Omalus aeneus (Fabricius, 1787) and Trichrysis cyanea (Linnaeus, 1758), both members of the Chrysididae family, typically parasitize solitary wasps (Paukkunen et al. 2015, Wiesbauer 2020). Omalus aeneus and T. cyanea are kleptoparasitoids, meaning the parasitoid larvae appropriates the host's food resources (Wiesbauer 2020). Although they share similar parasitism types, each species targets distinct groups of host species. In our study, O. aeneus parasitized two species from the genus Passaloecus spp. (Crabronidae), while T. cyanea parasitized a broader range of hosts, including four species from the genus Trypoxylon spp. (Crabronidae), and two species from the genus Deuteragenia spp. (Pompilidae). Both species are distributed across central Europe in various habitats, from forests to open areas such as agricultural landscapes, parks, and gardens (Paukkunen et al. 2015, Wiesbauer 2020).
It is important to note that, while all host and parasitoid species involved in this study are broadly distributed across the Palaearctic region according to the Global Biodiversity Information Facility (GBIF), specific information on their native, non-native, endemic, or invasive status within our study area is not available. Based on their broad distribution, these species are likely native to the Palaearctic region, but precise information on their origin is lacking.
Collection, identification, and measurements of parasitoids and their hosts
Solitary cavity-nesting bees and wasps were collected using trap nests between March and October 2020. Each trap nest was constructed using hollow reed internodes (Phragmites australis (Cav.) Trin. ex. Steud). The length of hollow reeds was approximately 20 cm and were fitted into a PVC tube with a diameter about 11 cm. Each trap exposed an average of 150 cavities per trap side, with a wide diameter range between 1 mm to 10 mm. Traps were attached in pairs on 1.5 m high wooden poles, placed in open ground spaces halfway between plot centres and the Northwest and Southeast corners (within a radius of approximately 5 m), totalling four traps per plot. Each trap was oriented facing Southeast and Northwest, to promote nesting via sunlight exposure. After being occupied with nests, traps were removed and refrigerated at ~4°C during October-February to simulate winter diapause. Nests were then exposed to room temperature, and hatched bees, wasps and parasitoids were collected for species determination. Species identifications were done with identification keys for each taxonomic group (e.g. (Dahl et al. 2007), for host species; (Bogusch et al. 2018), for parasitoid species, see Rappa et al. 2023, 2024). The parasitoid and host specimens used in this study are a subset of specimens used in previous investigations on solitary bees and wasps by Rappa et al. (2023, 2024). The authors investigated the importance and effectiveness of structural elements within forest ecosystems in enhancing biodiversity conservation through retention forestry practices.
Body size (measured as body length) of both parasitoids and their hosts were measured from the top of the head to the end of the abdomen, excluding the ovipositor protrusion, as the metric for body size in this study. Each parasitoid that emerged from a trap was measured. Similarly, only hosts that emerged from similar nests (not from other nests in the same trap) as the parasitoids were measured. Severely damaged specimens were excluded from the measurements to enhance data reliability and ensure accuracy. Variables such as intertegular distance (ITD) and forewing length were difficult to measure for nearly one third of our samples in our case due to damage incurred during the pinning preparation process, such as being stabbed by pins, over glued, or having folded wings. As a consequence, we did not include these variables in our measurements.
We measured a total of 393 individual parasitoid specimens (n = 25 for G. assectator, n = 151 for N. debilis, n = 45 for O. aeneus, and n = 172 for T. cyanea) (Figure S1, Table S1). For the hosts, we were unable to obtain an equal number of measurements as we did for their parasitoids, because some hosts had been consumed by the developing parasitoids. Omalus aeneus parasitized two crabronid wasp species: Passaloecus insignis (Figure S1) (n = 24) and P. corniger (Figure S1) (n = 1); T. cyanea parasitized four species of Trypoxylon and two species of Deuteragenia wasps (Figure S1), with Trypoxylon figulus being the most parasitized host species (n = 85). There are three bee species from genus Hylaeus that were parasitized by G. assectator (Figure S2), only one individual being available for the measurement of body size. Nematopodius debilis parasitized five Trypoxylon species, with T. clavicerum (Figure S1) (n = 17) and T. figulus (Figure S1) (n = 34) being the two dominant hosts parasitized (Table S1). All measurements were conducted using a Leica stereo microscope M165 C supported with a Leica Application Suite 3.8 (LAS) imaging software system using a 10x magnification.
Nest quality and forest structures
We selected a variety of nest diameters to have enough choices for the hosts and parasitoids to select their preferred sizes. The quality of the nests, i.e. the diameter of the hollow reeds, was measured with digital callipers. The diameters of the nest entrance (one side of an internode only) each of 383 hollow reeds were measured. For forest structures, the proportion of coniferous trees were calculated from the inventory basal area of coniferous trees carried out between 2016 and 2018 (Storch et al. 2020). The mean canopy cover was calculated in ImageJ using overhead hemispherical photos obtained at each trap location in early fall 2020. The remotely sensed indices stand structural complexity index (SSCI), were obtained from terrestrial laser scanners performed at the northwest and southeast corners, as well as the centres of the plots. Diameters at breast height (DBH) of standing deadwood above 7 cm in diameter were obtained during plot-level inventories in 2017 and 2018. There were five decay stages applied in classification of deadwood following Hunter (1990): recently dead or raw wood (i), solid deadwood (ii), rotten wood (iii), mould wood (iv), and duff wood (v). As the host species for parasitoids in our study preferred fresh and/or moderately decomposed deadwood, only decay stages i-iii were used to calculate the cumulative diameter of standing deadwood at plot level. Herbaceous covers were estimated from 5x5 m subplots during 2017.
Statistical analyses
We analysed the relationship between parasitoid body size as a response variable, and host identity and nest diameter as fixed factors, with study plot as random factor using linear mixed-effects models (type III sums of squares). For a subset of host-parasitoid interactions, in which we had at least 10 replicates (for two host species in N. debilis, and for one host species in each of O. aeneus and T. cyanea), we also tested if differences in the body size of host individuals explained variance in the body size of parasitoids. Here, we used nest diameter, host size and, in the case of N. debilis, host identity and the interaction of host identity and size, as fixed factors in the analysis. Because G. assectator fed on host larvae, we did not perform this analysis, as we could not obtain data on the size of larvae from any host species that G. assectator parasitized (see Table S1). We used separate models for each parasitoid species because (1) parasitoid species parasitized different host species, (2) parasitoid species and parasitoid-host combinations strongly differed in their number of replicates, and (3) data on the length of hosts could not be collected for all host-parasitoid species combinations (e.g. G. assectator and its hosts). Together, these factors led to a strong context-dependency of the studied host-environments effects on parasitoids and the associated underlying data structure, making it impossible to analyse all four parasitoid species in one global model.
In addition, one host species (T. figulus) was parasitised by N. debilis and T. cyanea simultaneously. Here, we investigated if relationships between the size of T. figulus and the two parasitoids differed between the parasitoid species. We used a linear mixed effect model with the size of both parasitoids as response variables, with the identity and size of T. figulus and its interaction as fixed factor, and study plot as random factor.
To test the relationship between the size of the four parasitoid species and their 15-host species, we averaged the size values of hosts and parasitoids at the species level. This resulted in a total of eleven host-parasitoid size combinations, as data for some parasitoid-host combinations could not be retrieved. We adjusted body sizes of parasitoids for differences between the parasitoid species and their host combinations using a linear mixed intercept model, with parasitoid species as a random variable and parasitoid body size as response variable. The residuals from this model were then used in a standard major axis regression (SMA) with parasitoid and host body sizes as dependent variables.
To test for host-environment interaction effects on parasitoid size, we averaged values of body length for host and parasitoid individuals at the plot-level. First, we analysed relationships between forest variables (stand structural complexity, canopy cover, standing deadwood diameter, and herb cover) and host body length using linear models. Then, we tested relationships between parasitoid length and the forest variables. Because differences in nest quality and host size explained differences in the size of parasitoids, we also included host-related factors at the plot-level (nest diameter, host identity, host size and its two-way interactions) as fixed factors in these models. In T. cyanea, O. aeneus, and G. assectator, host size and nest diameter were correlated at the plot-level, and host-environment interactions could only be analysed for one host species. Therefore, we included only host size as a co-variable in their respective models (Table S2 a, b). Note that we did not find effects of sex on the parasitoid-host interactions, nor differences in the sex ratios across forest variables. Consequently, the sex variable was excluded from all our analyses.
The linear mixed models were performed using the package ‘glmmTMB’ (Brooks et al. 2017). To study assumptions of homogeneity and normality of residuals, and to check for overdispersion, we used the package ‘DHARMa’ (Hartig 2024). We further used Wald-χ² test from the package ‘car’ (Fox and Weisberg, 2019) to assess the statistical significance of each term in the model. For the standard major axis regression, we used the package ‘lmodel2’ (Legendre 2018). All analyses were conducted using R (4.2.1) within the RStudio environment (R Core Team 2022).