The sampling characteristics of electivity indices.

Area of Science:

• Ecology
• Animal Behavior
• Quantitative Ecology

Background:

• Electivity indices quantify food selection by comparing diet proportions (r) to environmental availability (p).
• Numerous electivity indices exist, derived from Ivlev's work, each with unique algorithms and properties.
• Assessing the strengths and weaknesses of these indices is crucial for accurate ecological interpretation.

Purpose of the Study:

• To compare the performance of various electivity indices.
• To evaluate host tree preferences of gypsy moths (Lymantria dispar) using different electivity indices.
• To recommend the most suitable electivity index for ecological studies.

Main Methods:

• Graphical analysis of electivity index values across all combinations of diet proportion (r) and availability (p).
• Evaluation of indices based on criteria: value at r=p, symmetry, range, linearity, sensitivity to error, testability, and stability.
• Application of selected indices to gypsy moth feeding data on 19 host tree species in Quebec.

Main Results:

• No single electivity index perfectly satisfies all evaluation criteria.
• Different indices produced conflicting results regarding host tree preference and avoidance.
• Rank order of host tree preference was largely consistent across indices, except for Strauss' linear index (L).
• Gypsy moths showed strong preference for Populus grandidentata, Quercus rubra, Ostrya virginiana, and Amelanchier.
• Gypsy moths avoided Prunus serotina, Acer pensylvanicum, A. rubrum, Betula lutea, and Fraxinus americana.

Conclusions:

• The choice of electivity index significantly impacts the interpretation of feeding preferences.
• The Vanderploeg and Scavia's relativized electivity index (E*) is recommended for its robust performance.
• Understanding index limitations is vital for accurate ecological assessments of consumer-resource interactions.