Research ArticleECOLOGY

Adaptive switch to sexually dimorphic movements by partner-seeking termites

See allHide authors and affiliations

Science Advances  19 Jun 2019:
Vol. 5, no. 6, eaau6108
DOI: 10.1126/sciadv.aau6108
  • Fig. 1 Mating biology of subterranean termites, illustrated by R. speratus.

    In the mating season, alates (winged adults) fly off in large swarms to disperse. After dispersing, individuals shed their wings and walk to search for a mating partner without any previous locational information. Successfully encountered pairs then run in tandem to search for a suitable nest site, with the male following the female. Tandem running is quite synchronized, where a pair moves like a single individual. Nevertheless, they are sometimes separated and must search to find each other again. Thus, there are two search situations with different uncertainty about the location of their partner (before pair formation: uninformed search; after separation: reunion search). Photo credit: Nobuaki Mizumoto, Arizona State University.

  • Fig. 2 Pause behaviors of termite dealates across different periods of mate search.

    We first observed searching behaviors before pair formation. Then, we added a single mating partner of the other sex to observe their movements during tandem running. Last, we carefully removed the partner using an aspirator and observed how they change their searching behaviors. (A) Representative trajectories of dealates of R. speratus for two different trials, that is, focusing on a male (above) and a female (below). Trajectories are for 5 min for before pair formation and after separation, while they are for 3 min for tandem running. The movement trajectories of the added partner (above, female; below, male) were shown in dashed lines. Before pair formation, males and females showed similar movements, whereas they showed distinct sexually dimorphic movements after separation (movie S1). (B and C) Comparison of pause time between sexes in (B) R. speratus and (C) C. formosanus. Both species showed sexually dimorphic movements immediately after separation, where females often paused and males moved actively. Points with bars represent mean values with SDs. “*” indicates significant differences between sexes (Wilcoxon rank sum test with Bonferroni corrections, P < 0.05/70). Photo Credit: Nobuaki Mizumoto, Arizona State University.

  • Fig. 3 Searching efficiency of termite movements in two different searching situations.

    (A) Assumed search conditions before pair formation (uninformed search). Females and males search for partners without any locational information, which was simulated by periodic boundary conditions (size L × L) with random initial positions of females and males. (B and C) Searching efficiency under uninformed search conditions for (B) R. speratus (L = 223.6, φ = 7), where observed monomorphic movements and virtual monomorphic movements are overlapped, and (C) C. formosanus (L = 223.6, φ = 10). In both species, sexually monomorphic movements observed before pair formation achieved the higher encounter rates. (D) Assumed search conditions after a pair gets separated (reunion search). After separation, the distance between them is expected to be short. Such conditions were simulated by borderless continuous spaces with a distance d between a female and a male. (E and F) Searching efficiency of observed movement patterns under reunion search conditions for (E) R. speratus (d = 16.09, φ = 7), where the result after 180 s is inserted (10,000 replicates and data were obtained every 20 s until 3 × 105 s), and (F) C. formosanus (d = 22.97, φ = 10). In turn, sexually dimorphic movements observed after separation achieved the highest encounter rates during realistic searching period. The simulations performed by resampling empirical data to describe the durations of moves and pauses showed similar results (fig. S8). The results were obtained from means of 1,000,000 simulations.

  • Table 1 Parameters extracted from movement patterns of termites and comparison between sexes and searching schemes.

    The sinuosity corresponds to the scale parameter of wrapped Cauchy distributions, covering from 0 (most sinuous) to 1 (straight motion). In statistics, W corresponds to Wilcoxon rank sum test, where parameters were compared between sexes in each scheme and species, while V corresponds to Wilcoxon signed-rank test, where parameters were compared between before pair formation and after separation. P is the P value, and significant results are in bold. M and F indicate male and female, where the numbers in parentheses indicate sample size. The sample size of females in R. speratus is variable between before pair formation and after separation because some individuals only paused during analyzed duration after separation but moved before pair formation.

    A. R. speratus
    SpeedMean ± SE (mm/s)Wilcoxon rank
    sum test
    SinuosityMean ± SEWilcoxon rank
    sum test
    M (19)F (14, 16)WPM (19)F (13, 16)WP
    SchemeBefore14.7 ± 0.614.8 ± 0.71540.961Before0.85 ± 0.010.86 ± 0.011890.385
    After15.0 ± 0.85.4 ± 0.30<0.0001After0.77 ± 0.010.77 ± 0.061540.461
    Wilcoxon
    signed-rank
    test
    V1060V929
    P0.6790.0001P0.00010.2734
    B. C. formosanus
    SpeedMean ± SE (mm/s)Wilcoxon rank
    sum test
    SinuosityMean ± SEWilcoxon rank sum test
    M (22)F (20)WPM (22)F (20)WP
    SchemeBefore20.7 ± 1.422.3 ± 0.92790.142Before0.84 ± 0.010.85 ± 0.012740.18
    After19.0 ± 1.210.1 ± 1.143<0.0001After0.79 ± 0.010.82 ± 0.013100.0231
    Wilcoxon
    signed-rank
    test
    V1040V1535
    P0.4826<0.0001P<0.00010.007

Supplementary Materials

  • Supplementary material for this article is available at http://advances.sciencemag.org/cgi/content/full/5/6/eaau6108/DC1

    Supplementary Text

    Fig. S1. Duration of two different phases observed in tandem running.

    Fig. S2. Moving speeds of termite dealates across different periods in mate search.

    Fig. S3. Turning angles of termite dealates across different periods in mate search.

    Fig. S4. Comparison of the proportion of pausing times between sexes and conditions.

    Fig. S5. Inverse cumulative frequency distribution of the duration of moves and pauses.

    Fig. S6. Simulated encounter rates.

    Fig. S7. Histogram of the length of displacements between successive frames (0.2 s).

    Fig. S8. Simulation results using empirical data resampling to describe move-pause patterns.

    Table S1. Parameters on turning patterns extracted from turning angles both during moving and after pauses (reorientation).

    Table S2. Results of model fitting to moving and pausing time data.

    Movie S1. Sexual dimorphic movements after separation during tandem running in R. speratus.

    Movie S2. Sexual dimorphic movements after separation during tandem running in C. formosanus.

    Data file S1. All location data.

    Data file S2. Simulation codes.

    Reference (48)

  • Supplementary Materials

    The PDF file includes:

    • Supplementary Text
    • Fig. S1. Duration of two different phases observed in tandem running.
    • Fig. S2. Moving speeds of termite dealates across different periods in mate search.
    • Fig. S3. Turning angles of termite dealates across different periods in mate search.
    • Fig. S4. Comparison of the proportion of pausing times between sexes and conditions.
    • Fig. S5. Inverse cumulative frequency distribution of the duration of moves and pauses.
    • Fig. S6. Simulated encounter rates.
    • Fig. S7. Histogram of the length of displacements between successive frames (0.2 s).
    • Fig. S8. Simulation results using empirical data resampling to describe move-pause patterns.
    • Table S1. Parameters on turning patterns extracted from turning angles both during moving and after pauses (reorientation).
    • Table S2. Results of model fitting to moving and pausing time data.
    • Legends for movies S1 and S2
    • Legends for data files S1 and S2
    • Reference (48)

    Download PDF

    Other Supplementary Material for this manuscript includes the following:

    • Movie S1 (.mp4 format). Sexual dimorphic movements after separation during tandem running in R. speratus.
    • Movie S2 (.mp4 format). Sexual dimorphic movements after separation during tandem running in C. formosanus.
    • Data file S1 (.zip format). All location data.
    • Data file S2 (.cpp format). Simulation codes.

    Files in this Data Supplement:

Navigate This Article