This article is an open access article distributed under the terms and conditions of the Creative Commons Attribution license (http://creativecommons.org/licenses/by/3.0/).

This study examines the individual behavior of workers of the Formosan subterranean termite,

In social insects, such as ants and honey bees, elaborate systems of task allocation among castes have been documented extensively [

Subterranean termites construct extensive networks of underground tunnels. There is less information about task allocation among subterranean termite workers due to their cryptic lifestyle. A study examining foraging and mound building behavior by workers of

Numerous studies have examined the tunneling behavior of the Formosan subterranean termite [

When termites initiate construction of a new tunnel, only a single individual is able to enter the tunnel tip, and as the new tunnel gradually lengthens, more individuals become involved. Initially, only one or two individuals are involved in construction of a new tunnel. Tunnel width is related to the number of termites in the tunnel tip. Termites are more likely to widen the tunnel by engaging in lateral excavation when there are larger numbers present at the tunnel tip than when there are smaller numbers present at the tunnel tip [

Because only one or two individuals are involved in the construction of a new tunnel, it is possible that specific individuals play a role in the initiation of a new tunnel. The study presented here examined the tunneling behavior of marked individuals during the first hour of construction of a new tunnel on two consecutive days. The study examined whether specific individuals are more likely to tunnel than other individuals and whether their tunneling behavior on the second day can be predicted by their tunneling behavior on the first day.

Termites were collected from field colonies in an urban forest, City Park, New Orleans, LA, where termites were being monitored in over 100 underground traps using cylindrical irrigation valve boxes (NDS, Inc, Lindsay, CA) that were buried in the ground and filled with blocks of wood (spruce,

For each test, workers were uniquely marked on their dorsal abdomen with either a single color or a two color combination of eight different colors of enamel paint (Testor Corp., Rockford, IL). A group of 30 marked workers were introduced into a polystyrene, cylindrical screwtop container (4.5 cm high × 4.8 cm diameter). For each group of 30 marked workers, two trials were conducted on consecutive days using the same group of termites. There were two groups from each of two colonies.

The testing apparatus for tunneling assays.

Termites were able to enter a 2 cm length piece of PVC tubing (0.6 cm I.D. by 1.0 cm O.D. by 0.2 cm Wall) (Nalgene, Rochester, NY) inserted through a hole on the ventral side of the container and sealed in place with hot glue from a glue gun. A glass tube (10 cm length, 1 cm diameter) was filled with sand (Play Sand, Quikrete, Atlanta GA) and thoroughly moistened with distilled water. An indentation (2 mm length by 2 mm width) was made on the top of the sand in the glass tube in order to decrease the length of time required for termites to initiate tunnel construction and to ensure that termites constructed a tunnel at the top of the tube so that individuals could be easily identified. The indentation allowed a single termite to enter the sand-filled tube and initiate tunnel construction. The glass tube was attached to the distal end of the PVC tubing (

Tunneling behavior was defined by observing a termite picking up and moving sand particles. Before tunneling was initiated (the first time a termite picked up and moved sand particles), the number of times that a termite contacted the sand in the glass tube and left without tunneling was recorded. Once tunneling was initiated, the termites were observed continuously for 60 min and the time spent tunneling by each individual was recorded. The number of minutes spent tunneling by each individual on each day was determined for two consecutive days.

The total time spent tunneling by marked individuals in both trials was divided into time periods (none, 1–10 min, 11–20, 21–40, 41–60, >60) and compared using a Kruskal-Wallis one-way ANOVA on ranks. Means were separated using Tukey’s test with ranked sums. The number of individuals that did not tunnel, spent <25 min, or ≥25 min tunneling was compared for each day separately and for each group separately using a Pearson chi-square test. Also, the number of termites that did not tunnel on either day, tunneled on only one day, or tunneled on both days were compared for each group using a Pearson chi-square test. The total time spent tunneling by individuals on both days for each colony was compared using a Mann-Whitney U Rank Sum. The time taken to start tunneling and the number of times a termite contacted the sand before tunnel construction was initiated for each colony was compared using a t-test. The total numbers of termites that spent time tunneling on Day 1 and Day 2 were compared using a Pearson Chi-square.

Tunneling behavior of individuals was observed on consecutive days to determine if there was evidence for task allocation in tunneling behavior where key individuals did most of the work in the construction of a new tunnel on both days.

When comparing the total time spent tunneling by marked individuals on both days, the number of termites that tunneled different lengths of time was significantly different (Kruskal-Wallis: H = 17.48; df = 5; P = 0.004). There were significantly more termites that did not tunnel on either day than termites that spent either 41–60 min or >60 min tunneling (Tukey HSD Test: P ≤ 0.05) (

The number of individuals involved during the first hour of construction of a new tunnel was consistent for each trial. When the number of termites that did not tunnel, spent <25 min, or ≥25 min tunneling were compared for each day and each group separately, the proportion of individuals in each category was not significantly different in the different trials (Pearson chi-Square: 20.9; df = 14; P = 0.10) (

Mean (±SE) number of termites that spent time tunneling for different time intervals when the number of minutes each termite spent tunneling on both days was combined. Trials were conducted on consecutive days with the same group of 30 workers with two groups from each colony.

Time Spent Tunneling | Mean (±SE) Number of Termites |
---|---|

None | 14.0 ± 2.6a |

1–10 min | 5.0 ± 1.4ab |

11–20 min | 3.8 ± 0.9ab |

21–40 min | 3.8 ± 0.6ab |

41–60 min | 1.5 ± 0.3b |

> 60 min | 1.8 ± 0.5b |

Kruskal-Wallis: H = 17.48; df = 5; P = 0.004.

Number of termites in three categories of time spent tunneling (none, < 25 min, ≥25 min) in each 1-h tunneling trial conducted on consecutive days with the same group of 30 workers with two groups from each colony.

Colony-Group-Day | None | < 25 min | ≥25 min |
---|---|---|---|

1-1-1 | 19 | 7 | 4 |

1-1-2 | 15 | 9 | 6 |

1-2-1 | 26 | 3 | 1 |

1-2-2 | 24 | 2 | 4 |

2-1-1 | 22 | 4 | 4 |

2-1-2 | 16 | 7 | 7 |

2-2-1 | 18 | 9 | 3 |

2-2-2 | 23 | 5 | 3 |

Mean (±SE) | 20.4 ± 1.4 | 5.8 ± 0.9 | 3.9 ± 0.7 |

Pearson chi-Square: 20.9; df = 14; P = 0.10.

There was no significant difference in the total number of termites that spent time tunneling on Day 1 (35) compared with the total number of termites that spent time tunneling on Day 2 (43) (Pearson Chi-square: 1.2; df = 1; P = 0.27). There was no significant difference in the total number of minutes tunneled by termites on Day 1 (160 ± 40.2) compared with Day 2 (238.5 ± 55.1) (t-test: P = 0.29). Therefore, the disturbance involved in retesting the same individuals on a second day did not have an effect on the total number of termites tunneling, the total time spent tunneling, or the specific individuals involved in tunnel excavation.

Also, there were no significant colony differences in the tunneling behavior of individuals. When the number of minutes spent tunneling by individuals on both days was combined, there was no difference in the total time spent tunneling by termites in Colony 1 (12.3 ± 3.0) and Colony 2 (14.2 ± 2.7) (Mann-Whitney U Rank Sum: P = 0.36). There was no significant difference in time taken to start tunneling by Colony 1 (44.3 ± 13.8) and Colony 2 (31.0 ± 12.6) (t-test: P = 0.50). There was no significant difference in number of times a termite contacted the sand before tunnel construction was initiated between Colony 1 (14.8 ± 7.1) and Colony 2 (2.3 ± 1.7) (t-test: P = 0.14).

When the number of termites in three categories of time spent tunneling (none, one day only, both days) from both trials, was compared for each group, the proportion of individuals was not significantly different in the different groups (Pearson Chi-Square: 10.99; df = 6; P = 0.09) (

Number of termites in three categories of time spent tunneling (none, one day only, both days) from both trials conducted on consecutive days with the same group of 30 workers with two groups from each colony.

Colony-Group | None | One Day Only | Both Days |
---|---|---|---|

1-1 | 9 | 16 | 5 |

1-2 | 21 | 8 | 1 |

2-1 | 12 | 14 | 4 |

2-2 | 14 | 13 | 3 |

Mean (±SE) | 14.0 ± 2.6 | 12.8 ± 1.7 | 3.3 ± 0.8 |

Pearson Chi-Square: 10.99; df = 6; P = 0.09.

Number of termites in three categories of time spent tunneling (none, < 25 min, ≥ 25 min) in each 1-h tunneling trial conducted on consecutive days for four groups of 30 workers, two groups for each colony.

Time Spent Tunneling by Marked Individuals on Day 1and Day 2 | |||
---|---|---|---|

(Row Percentages of Termite Numbers) | |||

Day 2 | |||

Day 1 | None | < 25 min | ≥ 25 min |

None | 56 (72) | 17 (22) | 5 (6) |

< 25 min | 18 (78) | 4 (17) | 1 (4) |

≥ 25 min | 11 (58) | 2 (10) | 6 (32) |

Pearson Chi-square: 12.0; df = 4; P = 0.02.

When the tunneling behavior of individuals was compared on the two days, the probability of termites tunneling on Day 2 was not significantly affected by their tunneling behavior on Day 1 (did not tunnel on either day: 56; tunneled on Day 1 only: 29; tunneled on Day 2 only: 22; tunneled on both days: 13) (Pearson chi square: 0.10; df = 1; P = 0.752) However, the outcomes were significantly different than expected when comparing whether individuals that either did not tunnel, spent < 25 min, or ≥ 25 min tunneling on Day 1 behaved similarly on Day 2 (Pearson Chi-square: 12.29; df = 4; P = 0.015). While 72% of individuals that did not tunnel on Day 1 did not tunnel on Day 2 and 6% spent ≥ 25 min tunneling, 32% of individuals that spent ≥ 25 min tunneling on Day 1 also spent ≥ 25 min tunneling on Day 2 (

Individuals were ranked based on the time spent tunneling on Day 1. Figures show individuals ranked from the most to the least time spent tunneling on Day 1 compared to the time each individual spent tunneling on Day 2 for each group. There were individuals ranked in the top four on both days in three of the four groups. For Colony 1-Group 1, two of the four primary excavators on Day 1 and 2 were the same individuals (

The number of minutes spent tunneling by each individual marked termite in Colony 1-Group 1 on Day 1 and Day 2.

This study identified specific individuals that were more likely than other individuals to initiate the construction of a new tunnel, and to spend the most time tunneling during the first 60 min of tunnel construction. These results suggest that there is task allocation in the construction of new tunnels. Only one or two individuals were involved during the first few minutes of tunnel construction. Additional individuals gradually got involved over the first hour. On average, four individuals did most of the work during the first hour of tunnel construction. An individual that spent at least 25 min tunneling on Day 1 was significantly more likely to spend at least 25 min tunneling on Day 2 than individuals that spent < 25 min tunneling on Day 1. Because the number of individuals involved in tunnel construction increases as the tunnel expands [

The number of minutes spent tunneling by each individual marked termite in Colony 1-Group 2 on Day 1 and Day 2.

The number of minutes spent tunneling by each individual marked termite in Colony 2-Group 1 on Day 1 and Day 2.

The number of minutes spent tunneling by each individual marked termite in Colony 2-Group 2 on Day 1 and Day 2.

Task allocation among termite workers has been well documented for open air foragers and fungus-growing termites [

Results from this study indicate that group size and arena size may affect termite behavior in tunneling assays designed to test the efficacy of soil termiticides. Lenz [

In a study where

In the current study, only a small number of individuals were involved in tunnel excavation, while most of the individuals were inactive. However, under natural conditions, individuals that are not involved in tunnel excavation may perform other tasks, such as colonizing new food sources, tunnel maintenance and repair, or brood care. Also, individuals that were not active during the 60 min observation period may act as a reserve work force [

Increasing our understanding of the behavior of individuals in tunnel construction could eventually lead to a greater understanding of the physiological and behavioral processes that regulate termite tunneling behavior, termite sociality, and behavioral polyphenism. This knowledge could be used in the development of novel methods of termite control.

I would like to thank Erin Gallatin for her invaluable technical assistance on this project. I would also like to thank Mathew Tarver for making substantial improvements to an earlier draft of this manuscript.