. | No parts benefit . | One part benefits . | Both parts benefit . |
---|---|---|---|
Layer 4 | 80.5% | 4.5% | 15.0% |
Layer 4 Shuffled | 89.3% | 2.7% | 8.0% |
p-value of difference | 0.041 | 0.214 | 0.033 |
p-value < 0.05 | Yes | No | Yes |
. | No parts benefit . | One part benefits . | Both parts benefit . |
---|---|---|---|
Layer 4 | 80.5% | 4.5% | 15.0% |
Layer 4 Shuffled | 89.3% | 2.7% | 8.0% |
p-value of difference | 0.041 | 0.214 | 0.033 |
p-value < 0.05 | Yes | No | Yes |
Notes. Shuffling decreases the probability that the component parts will both benefit from fusion; that is, the probability of symbiosis as persistent mutualism. The whole is fitter than both parts 15% of the time for Layer 4, but only 8% of the time for Layer 4 Shuffled. This shows that increased fitness does not come merely from the increase in area when fusing two seeds; a significant part of the increased fitness comes from the structural properties of the two seeds that are fused together. Layer 4 Shuffled is less likely to have the proper structure than Layer 4. The p-values in the table are calculated using a two-tailed Welch t-test for samples with unequal variance (heteroscedastic variance).