Evan L. Preisser, Christopher J. Dugaw, Brian Dennis, and Donald R. Strong. 2006. Plant facilitation of a belowground predator. Ecology 87:1116–1123.

Appendix B. Experimental production of Heterorhabditis marelatus IJs within hosts.

Infective juveniles (IJs) are the soil dwelling, predatory, mobile propagules of entomopathogenic nematodes. We counted the number of H. marelatus IJs produced by H. californicus larvae of different sizes. We collected large larvae from lupine roots and raised small larvae, which are difficult to find in the field, from eggs (Preisser 2003). Prior to infection, each larva was observed for two days in a darkened container with lupine wood that had been pasteurized in order to kill any pre-existing EPNs. We used 40 larvae, ranging in weight from 0.4 mg (1^st instar) to 760 mg (9^th–10^th instar). Each larva was weighed and placed in a three cm diameter petri plate on filter paper wetted with a solution containing approximately 10,000 H. marelatus IJs obtained from white traps (Kaya and Stock 1997).The distinctive orange-gold color symptomatic of H. marelatus infection developed in 33/40 ghost moth larvae within 2–4 days. We placed infected larvae on individual white traps to measure IJ production. We counted IJs in the water surrounding the white trap; when IJ production was at its peak, we used subsamples to measure IJ density.

Statistical analysis: Since many of the 1^st- and 2^nd-instar larvae did not produce any H. marelatus IJs, we fit (IJs produced per cadaver + 1) as a function of H. californicus larval weight (g) using JMP-IN v. 5.1 (SAS 2004). We determined the relationship between larval instar and length using data published elsewhere (Wagner 1985), then used this relationship in conjunction with our data measuring ghost moth larval length, weight, and nematode production to estimate nematode cohort size as a function of larval instar.

Results: Large ghost moth larvae produced far more IJs than small larvae (F_1,31 = 135, P < 0.0001). The best fit to the data was a square-root transformation: (sqrt(IJs + 1) = 1105.9*host weight (g) – 16.1; r² = 0.814). There was a huge increase in weight and subsequent IJ production between the third and fourth instar. Third instar larvae had a mean weight of 27.5 mg (lower and upper 95%: 21.8 and 33.9 mg), producing a mean of 206 IJs (lower and upper 95%: 64 and 459 IJs). For fourth instar larvae, the corresponding values were 92.0 mg (95%: 73.2 and 112.9), producing 7,338 IJs (95%: 4,211 and 11,836 IJs). Only three of 10 1^st- and 2^nd-instar larvae weighing 2.6 mg or less produced IJs; these three weighed 0.8, 2.1, and 2.6 mg and produced 121, 2, and 5 IJs, respectively. All 23 of the 2^nd to last (9^th–10^th) instar larvae weighing above 2.6 mg produced IJs. The smallest five weighed 24.3 + 4.5 (SE) mg and produced 2809 + 1743 IJs, while the largest five weighed 672 + 34.1 mg and produced 643,618 + 101,495 IJs.

LITERATURE CITED

Kaya, H. K., and S. P. Stock. 1997. Techniques in insect nematology. Pages 281–324 in L. A. Lacey, editor. Manual of techniques in insect pathology. Academic Press, San Diego, California, USA.

Preisser, E. 2003. Field evidence for a strongly cascading underground food web. Ecology 84:869–874.

SAS. 2004. JMP-IN v.5.1. Duxbury Learning, Pacific Grove, California, USA.

Wagner, D. L. 1985. The biosystematics of Hepialus F. s. lato, with special emphasis on the californicus-hectoides species group. Ph.D. Thesis, U.C. Berkeley, Berkeley, California, USA.