Bee Foraging Patterns: New Statistical Methods and the Effect of Species Richness and Sub-Lethal Pesticide Exposure Open Access

Ayers, Carolyn (2016)

Permanent URL:


My dissertation consists of two distinct components, each represented by two chapters. In the first component, I develop new statistical techniques to quantify and test the significance of trapline foraging, a behavior in which foragers repeatedly visit spatially fixed resources in a predictable sequence. Though traplining is taxonomically widespread, the few metrics and null models that exist to statistically test traplining have substantial drawbacks. In my first chapter, I present a modified version of determinism from recurrence quantification analysis as a standard metric for quantifying traplines. Using empirical data to compare metrics, I find that determinism offers an improvement over other metrics since it does not depend on the arrangement of resources or experimental design, which allows for comparisons between differing environments. In my second chapter, I present a spatially explicit, individual-based null model designed to test whether resource layouts and realistic forager movements alone can account for suspected traplines. Using empirical data, I find that my null model is less prone to Type I or II statistical error relative to existing models. In the second component, I use a foraging enclosure with artificial flowers to examine the effects of bee species richness and sub-lethal neonicotinoid pesticide exposure on functionally important bee foraging behaviors. Pollinator diversity is declining worldwide, yet it is relatively unknown how species losses will affect plant pollination services. In my third chapter, I examine how bee species richness drives patterns of bee specialization, which is important for conspecific pollen transfer. I find that species-level specialization and complementarity increase with bee species richness. The focus of my fourth chapter is exposure to neonicotinoid pesticides, which have been implicated as a potential driver of bee declines and have been shown to affect bee foraging behaviors at low concentrations. I examine how field-realistic neonicotinoid exposure interacts with lost bee diversity to affect bee behaviors important for bee fitness. I find that neonicotinoid exposure decreases total flower visits and bee energy gains in a multiple species context. These findings indicate that neonicotinoid exposure and bee species losses may negatively affect bee and plant fitness more greatly than previously anticipated.

Table of Contents


Traplining. 1

Biodiversity Ecosystem Functioning Relationships. 3

Pollinators as a Study System for BEF. 4

Neonicotinoid Pesticides. 6

Overview of Dissertation. 8

Literature Cited. 11

PART I. 18

Chapter 1: Determinism as a Statistical Metric for Ecologically Important Recurrent Behaviors with Trapline Foraging as a Case Study. 18

Abstract. 18

Introduction. 19

Methods. 23

Statistical Analysis of Determinism. 28

Worked Examples. 30

Final Remarks. 32

Literature Cited. 35

Tables. 44

Figures. 46

Chapter 2: Statistically testing the role of individual learning and decision-making in trapline foraging behavior: a spatially explicit, individual-based null model approach. 50

Abstract. 50

Introduction. 51

Methods. 55

Results. 60

Discussion. 63

Literature Cited. 67

Figures. 72


Chapter 3: The Effect of Bee Species Richness on Complementarity and Resource Specialization on the Individual and Species Level. 77

Abstract. 77

Introduction. 78

Methods. 83

Results. 88

Discussion. 89

Literature Cited. 94

Tables. 100

Figures. 103

Chapter 4: The Effect of Sub-lethal Neonicotinoid Pesticide Exposure on Functionally-Relevant Bee Foraging Behaviors in a Single versus Multiple Species Context. 107

Abstract. 107

Introduction. 108

Methods. 111

Results. 115

Discussion. 118

Literature Cited. 123

Tables. 129

Figures. 130



Summary. 139

Calculation of Determinism. 139

Calculating the Prevalence of Reverse Sequences. 143

Calculation of Simulated Sequences. 144

Sensitivity Analysis: the Effect of Resource Abundance on Determinism. 145

The Effect of Bee Foraging Experience on the Degree of Traplining. 146

Tables. 148

Figures. 152


Sensitivity Analysis of Resource Density, Detection Distance, and Movement Type. 155

Decision Rules for Selecting Amongst Multiple Resources. 157

R Code: Sample Randomization Null Model. 158

Netlogo Code: Null Model for Testing the Statistical Significance of Traplines. 162

Bee Foraging Sequences from "Emory" Dataset. 162

Sensitivity of Determinism to Minimum Trapline Length. 162

Tables. 163

Figures. 168

About this Dissertation

Rights statement
  • Permission granted by the author to include this thesis or dissertation in this repository. All rights reserved by the author. Please contact the author for information regarding the reproduction and use of this thesis or dissertation.
Subfield / Discipline
  • English
Research Field
Committee Chair / Thesis Advisor
Committee Members
Last modified

Primary PDF

Supplemental Files