Electronic Supplementary Material

Social Network Influences Decision Making During Collective Movements in Brown Lemurs (Eulemur fulvus fulvus)

Armand Jacobs · Cédric Sueur · Jean Louis Deneubourg · Odile Petit

Table SI Characteristics of individuals (name, age, sex, and index of initiations) and identity number in the model

Individual / Sex / Age / Network identity / Model identity / Index of initiations
Alain / ♂ / 13 / A01 / 1 / 1.294
Cryo / ♀ / 15 / C01 / 2 / 1.294
Candide / ♂ / 10 / C02 / 3 / 1.294
Calimero / ♂ / 4 / C03 / 4 / 0.647
Clement / ♂ / 3 / C04 / 5 / 0.647
Cedille / ♀ / 11 / C05 / 6 / 1.618
Celia / ♀ / 6 / C06 / 7 / 1.294
Cid / ♂ / 2 / C07 / 8 / 0.971
Choline / ♀ / 3 / C08 / 9 / 0.324
Cyrano / ♂ / 3 / C09 / 10 / 0.971
Clopin / ♂ / 1 / C10 / 11 / 0.647

Index of initiations of individual i was initiation rate for i (number of initiations of i per the total number of initiations) per the mean rate of initiations of the group. This correction allowed us to have a necessary weighted index in the model.

Table S2 Kinship matrix

1 / 2 / 3 / 4 / 5 / 6 / 7 / 8 / 9 / 10 / 11
1 / 0.000 / 0.000 / 0.000 / 0.000 / 0.000 / 0.000 / 0.000 / 0.000 / 0.000 / 0.000 / 0.000
2 / 0.000 / 0.000 / 1.833 / 1.833 / 0.458 / 1.833 / 0.917 / 0.917 / 1.833 / 0.917 / 0.458
3 / 0.000 / 3.141 / 0.000 / 1.571 / 0.396 / 1.571 / 0.785 / 0.785 / 1.571 / 0.785 / 0.396
4 / 0.000 / 2.749 / 1.374 / 0.000 / 0.346 / 1.374 / 0.687 / 0.687 / 2.749 / 0.687 / 0.346
5 / 0.000 / 0.879 / 0.443 / 0.443 / 0.000 / 1.758 / 3.517 / 0.879 / 0.443 / 0.879 / 1.758
6 / 0.000 / 1.833 / 0.917 / 0.917 / 0.917 / 0.000 / 1.833 / 1.833 / 0.917 / 0.917 / 0.917
7 / 0.000 / 1.048 / 0.524 / 0.524 / 2.095 / 2.095 / 0.000 / 1.048 / 0.524 / 1.048 / 2.095
8 / 0.000 / 1.467 / 0.733 / 0.733 / 0.733 / 2.933 / 1.467 / 0.000 / 0.733 / 1.467 / 0.733
9 / 0.000 / 2.749 / 1.374 / 2.749 / 0.346 / 1.374 / 0.687 / 0.687 / 0.000 / 0.687 / 0.346
10 / 0.000 / 1.692 / 0.846 / 0.846 / 0.846 / 1.692 / 1.692 / 1.692 / 0.846 / 0.000 / 0.846
11 / 0.000 / 0.879 / 0.443 / 0.443 / 1.758 / 1.758 / 3.517 / 0.879 / 0.443 / 0.879 / 0.000

The kinship index for individual i was the genetic coefficient between an individual i and an individual j per the mean genetic coefficient of individual i.

Table SIII Matrix of affiliative relationships

1 / 2 / 3 / 4 / 5 / 6 / 7 / 8 / 9 / 10 / 11
1 / 0.000 / 3.899 / 2.785 / 0.139 / 2.089 / 0.139 / 0.139 / 0.139 / 1.392 / 0.139 / 0.139
2 / 3.050 / 0.000 / 2.178 / 0.109 / 2.941 / 0.000 / 0.109 / 0.109 / 2.287 / 0.109 / 0.109
3 / 2.157 / 2.157 / 0.000 / 0.000 / 1.725 / 1.402 / 0.000 / 1.618 / 0.755 / 1.186 / 0.000
4 / 0.208 / 0.208 / 0.000 / 0.000 / 0.000 / 0.208 / 4.981 / 0.623 / 0.623 / 0.830 / 3.321
5 / 1.964 / 3.536 / 2.095 / 0.000 / 0.000 / 0.262 / 0.131 / 0.131 / 2.488 / 0.131 / 0.262
6 / 0.155 / 0.000 / 2.014 / 0.155 / 0.310 / 0.000 / 0.000 / 3.873 / 0.000 / 4.338 / 0.155
7 / 0.244 / 0.244 / 0.000 / 5.867 / 0.244 / 0.000 / 0.000 / 0.244 / 0.244 / 0.244 / 3.667
8 / 0.149 / 0.149 / 2.230 / 0.446 / 0.149 / 3.716 / 0.149 / 0.000 / 0.149 / 3.270 / 0.595
9 / 1.467 / 3.080 / 1.027 / 0.440 / 2.787 / 0.000 / 0.147 / 0.147 / 0.000 / 0.147 / 1.760
10 / 0.149 / 0.149 / 1.635 / 0.595 / 0.149 / 4.162 / 0.149 / 3.270 / 0.149 / 0.000 / 0.595
11 / 0.196 / 0.196 / 0.000 / 3.143 / 0.393 / 0.196 / 2.946 / 0.786 / 2.357 / 0.786 / 0.000

The contact index for the pair of individual i–j was the number of scans for which 2 individuals, i and j, were in contact per the mean number of scans in which individual i was observed.

Analyses of Departure Latency Distribution According to These Different Mimetic Coefficients

The equation for departure probabilities (Eq. 2d in the main text) allowed us to determine a range of mimetic coefficients from 0.001 to 0.004.

In these analyses, we determined which mimetic coefficient (from 0.001 to 0.004) fitted with the observed mimetic coefficient. We then calculated the departure latency of each joiner j for each mimetic coefficient as follows:

(Eq. 2b in the main text)

.

Results show that the best fitting mimetic coefficient was the one equaling 0.004 (Fig. S1; Table SIV).

According to this result; we fixed C at a value of 0.004 in our model.

Fig. S1 Observed and calculated distributions of departure latencies for joiners. Rank 2 indicated the first joiner, and rank n indicated the n – 1 joiner.

Table SIV Function comparison test (Kolmogorov-Smirnov) between observed and simulated distributions of departure latencies for joiners.

Z value / p value
C = 0.001 / 2.23 / < 0.00001
C = 0.002 / 1.78 / 0.001
C = 0.003 / 0.671 / 0.602
C = 0.004 / 0.671 / 0.623

Tests show that distributions with C = 0.001 and C = 0.002 were different from observed distributions. Between distributions with C = 0.003 and C = 0.004, distribution with C = 0.004 fit better with the observed distribution (see Fig. S1). A p < 0.05 showed that the 2 distributions were different.

Fig. S2 Distribution of direct joining indices for observed data (a) and for the 5 hypotheses (b–f). Only the data distribution for the affiliation hypothesis were exponential and similar to the distribution of observed data (see main text for details about results).

Fig. S3 Survival analysis of joining latencies for the 1st joining individual (rank = 2).

Fig. S4 Survival analysis of joining latencies for the 2nd joining individual (rank = 3).

Fig. S5 Survival analysis of joining latencies for the 3rd joining individual (rank = 4).

Fig. S6 Survival analysis of joining latencies for the 4th joining individual (rank = 5).

Fig. S7 Survival analysis of joining latencies for the 5th joining individual (rank = 6).

Fig. S8 Survival analysis of joining latencies for the 6th joining individual (rank = 7).

Fig. S9 Survival analysis of joining latencies for the 7th joining individual (rank = 8).

Fig. S10 Survival analysis of joining latencies for the 8th joining individual (rank = 9).

Fig. S11 Survival analysis of joining latencies for the 9th joining individual (rank = 10).

Fig. S12 Survival analysis of joining latencies for the 10th joining individual (rank = 11).