Data from: Testing the role of body size and litter depth on invertebrate diversity across six forests in North America

Data files for manuscript titled "Testing the role of body size and litter depth on invertebrate diversity across six forests in North America". Metadata is contained within excel file that describes all variables for each tab.

Methods

Study sites Samples were collected from six forests in North America: Andrews LTER, Oregon (AND); Barro Colorado Island, Panama (BCI); Coweeta LTER, North Carolina (CWT); Harvard Forest LTER, Massachusetts (HFR); Luquillo LTER, Puerto Rico (LUQ); and Niwot Ridge LTER, Colorado (NWT). At each site, 21 1-m2 plots were arrayed in a cross design with plots placed 1, 10, 50, 100, and 200 m at each cardinal direction from a central plot.

Invertebrate sampling Invertebrates were collected from each plot by sifting litter and approximately 0.5 cm of mineral soil through 1 cm mesh screens. Invertebrates were then extracted from the siftate over 48 hours in Tulgren funnels with 25 W bulbs. Individuals were sorted taxonomically, counted, and assigned to morphospecies. Mites (Acari), springtails (Collembola), and spiders (Arachnida) were the most abundant invertebrate groups collected within most plots and we focus on differences amongst these three groups.

Body size and litter depth measurements For each plot, all individuals from the three focal groups listed above were photographed within a 10 mm x 10 mm box set at 10x magnification using a Leica EC3 digital camera with FireCam software version 3.4.1 mounted on a Leica S8 APO stereo microscope. After photos were cropped and resized using Adobe Photoshop CS3 version 10.0.1, the length and width of each individual, representing the major and minor axes respectively, were measured in mm using ImageJ. As the body shape of most invertebrates roughly resembles an ellipse, which may be a better correlate of size (i.e. mass), we also calculated ellipse area using the equation: Area = Pi (length/2) (width/2). To validate our methods, digital measurements from a subset of mites were compared to those from an ocular micrometer and found to be similar (r = 0.998, P < 0.0001). We then dried 84 of those mites to constant mass at 60°C for 48 h and weighed them to the nearest 0.001 mg on a Cahn microbalance. Of the three body size measurements, ellipse area was the best correlate to mass (r = 0.940, P < 0.0001) and subsequently used in all hypothesis tests. Litter depth, our measure of habit size for invertebrates, was measured 1 cm from the four corners of each plot and averaged. We exclude two plots from our analysis, one at AND due to litter depth not being recorded and the second at NWT due to the absence of any individual from our selected taxa. This changes the total number of potential plots used to 124. Of those 124 plots, invertebrates were available for measurement in 123, 102, and 78 plots for mites, springtails, and spiders respectively.

• Resource Title: Data for "Testing the role of body size and litter depth on invertebrate diversity across six forests in North America".

  File Name: MACROplots - Final Data File.xlsx

  Resource Description: Excel file with 4 tabs: Metadata, master data, Ocular-ImageJ validation, and Mass-ImageJ validation