Skip to main content
Dryad logo

Impacts of female body size on cannibalism and juvenile abundance in a dominant arctic spider

Citation

Koltz, Amanda; Wright, Justin (2020), Impacts of female body size on cannibalism and juvenile abundance in a dominant arctic spider, Dryad, Dataset, https://doi.org/10.5061/dryad.dz08kprtz

Abstract

1. Body size influences an individual’s physiology and the nature of its intra- and interspecific interactions. Changes in this key functional trait can therefore have important implications for populations as well. For example, among invertebrates, there is typically a positive correlation between female body size and reproductive output. Increasing body size can consequently trigger changes in population density, population structure (e.g., adult to juvenile ratio), and the strength of intraspecific competition.

2. Body size changes have been documented in several species in the Arctic, a region that is warming rapidly. In particular, wolf spiders, one of the most abundant arctic invertebrate predators, are becoming larger and therefore more fecund. Whether these changes are affecting their populations and role within food webs is currently unclear.

3. We investigated the population structure and feeding ecology of the dominant wolf spider species (Pardosa lapponica) at two tundra sites where adult spiders naturally differ in mean body size. Additionally, we performed a mesocosm experiment to investigate how variation in wolf spider density, which is likely to change as a function of body size, influences feeding ecology and its sensitivity to warming.

4. We found that juvenile abundance is negatively associated with female size and that wolf spiders occupied higher trophic positions where adult females were larger. Because female body size is positively related to fecundity in P. lapponica, the unexpected finding of fewer juveniles with larger females suggests an increase in density-dependent cannibalism as a result of increased intraspecific competition for resources. Higher rates of density-dependent cannibalism are further supported by the results from our mesocosm experiment, in which individuals occupied higher trophic positions in plots with higher wolf spider densities. We observed no changes in wolf spider feeding ecology in association with short-term experimental warming.

5. Our results suggest that body size variation in wolf spiders is associated with variation in intraspecific competition, feeding ecology, and population structure. Given the widespread distribution of wolf spiders in arctic ecosystems, body size shifts in these predators as a result of climate change could have implications for lower trophic levels and for ecosystem functioning. 

Methods

Field sampling

Wolf spider communities were sampled from two arctic tundra sites, Toolik (68°65′N, 149°58′W) and Imnavait (68°62′N, 149°30′W) in the summer of 2012. The sites are 10.4 km apart in the northern foothills of the Brooks Range, Alaska, near Toolik Lake Field Station. At each site, we sampled wolf spiders two weeks, one month, and two months after snowmelt had occurred (Table S1 contains the dates of sampling, spring snowmelt, and fall snow accumulation during the study year).

At two weeks and one month past snowmelt, we sampled each site using pitfall traps placed one meter apart on a grid of 10 x 10 meters. Due to concerns about depleting the local population and potential low sample sizes during the later part of the season, at two months past snowmelt, we sampled adjacent areas at each site using four 5 x 5 meter grids of pitfall traps; each grid was separated by at least thirty meters. Thus at each sampling time point, there were 100 pitfall traps per site. Pitfall traps contained 75% ethanol and were left out for 24 hours during each sampling period, except during the sampling at one month past snowmelt at Imnavait, where traps were open for 48 hours due to unforeseen poor weather. We therefore only used pitfall catch data from 15 and 60 days past snow melt to compare abundances of females between sites (see below).

Sample processing

We identified all adult female specimens from Toolik and Imnavait to species according to Dondale and Redner (1990), except for the classification of Pardosa concinna and Pardosa lapponica, which are likely to be the same species (Sim et al. 2014). Juveniles cannot be identified to the species level by morphology alone. As in Sackett et al. (2008), we assume that the relative abundance of juveniles of each species reflects that of the adult community. To compare body sizes (and hence, reproductive potential) at each site and to relate body size to trophic position, we measured the carapace width of every adult female and juvenile wolf spider using digital calipers (Diesella, Kolding, Denmark). We also dissected all egg sacs present from females sampled at Toolik and Imnavait in order to estimate parasitism rates at the two sites.

Mesocosm experiment

We conducted a fully factorial mesocosm experiment during June-July 2013 near Toolik Field Station to measure the effects of altered wolf spider densities and warming on wolf spider feeding ecology (Fig. S1). The mesocosms were 1.5 meters in diameter and enclosed with aluminum flashing that was buried 20 cm belowground and stood 20 cm above the soil surface. Plots were distributed among five blocks and randomly assigned to one of six spider density/warming treatments for a total of 30 plots (see Koltz et al. 2018 PNAS for more detailed methods on the mesocosm experiment). 

For the warming treatment, we covered the mesocosm openings with heavy gauge plastic sheeting that had regularly spaced, small openings to allow the passage of air and rainfall. We measured temperature at the soil surface in the plots every half hour from June 12 – July 22nd using iButtons (Maxim Integrated, San Jose, CA). Temperatures during this period were significantly higher in the experimentally warmed plots than in the ambient temperature plots (mixed effects model with plot nested in block; warming treatment: Estimate ± SE = 2.223 ± 0.381, df = 27, t = 5.845, p < 0.0001). Mean temperatures in the ambient temperature plots were 15.23 (± 2.35 SE)°C and those in the warmed plots were 17.45 (± 2.95 SE)°C (Fig. S2).

The spider density treatments included: 1) reduced wolf spider density; 2) control wolf spider density, and 3) high wolf spider density. We removed all wolf spiders from the reduced density plots at the beginning of the summer and continued to check and remove individuals throughout the experiment. High-density plots received enough additional spiders at the start of the experiment to bring wolf spider densities to approximately double the early season average density of control plots. We used pitfall traps at the end of July to verify that we had successfully manipulated wolf spider densities (ANOVA: F2,27 = 6.20, p = 0.0061). Post-hoc Tukey tests for these data showed that high density plots had significantly more spiders than either low density (p = 0.0061) or control plots (p = 0.044), although densities did not differ between low and control density plots (p = 0.68; Fig. S3a). Density estimates from the pitfall traps showed that there was an average of 0.3 ± 0.48 wolf spiders in low density plots, 0.6 ± 0.7 wolf spiders in control plots, and 1.5 ± 1.08 wolf spiders in the high density plots. At the conclusion of the experiment, we destructively sampled the plots in order to opportunistically catch any remaining wolf spiders. We measured the body sizes (carapace width) of these specimens and of the wolf spiders caught earlier in pitfall traps and removed the legs from a subset of individuals for stable isotope analyses.

Wolf spider feeding ecology

We used stable isotope analyses of carbon (C) and nitrogen (N) to quantify the functional feeding roles of a subset of juveniles and non-egg sac carrying adult female P. lapponica from the Toolik and Imnavait field populations. We also conducted stable isotope analyses on all captured wolf spiders from the mesocosm experiment. While sample storage in ethanol can affect carbon stable isotope ratios in invertebrates, we assume here that any effects of ethanol storage were similar across samples and experimental treatments. We also measured the stable isotope ratios of a subset of Collembola that were removed from pitfall traps to use as site-specific isotopic baselines for Toolik and Imnavait (see Ponsard and Arditi 2000). We used Collembola as our isotopic baseline because this group is an important prey source for wolf spiders (Wise 2004, Koltz et al. 2018b), and because as detritivores, they should reflect differences in δ13C between the two sites due to any differences in vegetation. Values of δ13C and δ15N for the field-collected wolf spiders from Toolik and Imnavait are expressed as the measured values minus the site-specific mean value for Collembola.

Wolf spider and Collembola specimens were dried for 48 hours at 60 °C and homogenized with a mortar and pestle prior to stable isotope analysis. We used whole bodies of specimens from the Toolik and Imnavait field populations and between 4-6 legs from wolf spiders collected from the mesocosm experiment. Because of their small size, up to 13 Collembola were pooled per sample. Stable isotope ratios of animals were then determined at the Duke Environmental Stable Isotope Laboratory (DEVIL) using a Carlo-Erba NA1500 elemental analyzer feeding a Thermo Finnigan Delta Plus XL continuous flow mass spectrometer system. Ratios were calculated as δX in per ml (vs. atmospheric N2 for 15N and vs. VPDB for 13C) as ((Rsample− Rstandard)/Rstandard) x 1000, whereby R represents the heavy to light isotope ratio (i.e. 13C/12C or 15N/14N) and X is the target isotope. Reference materials from The National Institute of Standards and Technology (NIST) and the U.S Geological Survey (USGS), as well as internally calibrated standards were used for 3-point normalization of raw isotope data. The ratio of standards to samples analyzed was approximately 1:6.

Funding

National Science Foundation, Award: 1210704

National Geographic Committee for Research and Exploration

Research and Education Opportunities International (CREOi)

Alaska Geographic

Kappa Delta Foundation

Lewis and Clark Fund

Arctic Institute of North America

Discover Denali Research Fellowship

National Geographic Committee for Research and Exploration

Research and Education Opportunities International (CREOi)

Alaska Geographic

Kappa Delta Foundation

Lewis and Clark Fund

Discover Denali Research Fellowship