Skip to main content
. 2014 Oct 8;9(10):e109001. doi: 10.1371/journal.pone.0109001

Table 1. ANOVA on fitness measures for crop-wild hybrid sunflowers grown under competitive conditions in Kansas, USA.

Prob. Emerged1 Prob. Surv. Reprod.1 Seeds per Repr. Plt2 Seeds per Emgd Sdling2 Seeds per Seed Plted2
Source ndf3 ddf3 F P-value F P-value F P-value F P-value F P-value
Frequency of Hybrids 1 55 0.23 0.6371 0.14 0.7102 6.19 0.0159 7.57 0.008 5.46 0.0231
Interspec Competitors 1 55 0 0.9707 0.74 0.3947 61.06 <.0001 50.19 <.0001 36.72 <.0001
Density of Seeding 2 55 0.06 0.9456 33.41 <.0001 292.8 <.0001 315.6 <.0001 215.53 <.0001
Freq*Interspec 1 55 0.04 0.8514 0.26 0.6138 0.52 0.4742 0.23 0.63 0 0.9872
Freq*Dens 2 55 1.07 0.3499 0.08 0.9254 1.81 0.1736 1.85 0.167 2.56 0.0864
Interspec*Dens 2 55 0.05 0.9498 0.18 0.8333 2.9 0.0636 3.14 0.0513 3.14 0.0511
Freq*Interspec*Dens 2 55 0.47 0.6251 0.92 0.4063 0.08 0.9229 0.01 0.9907 0.36 0.6993
Cross Type 3 179–186 65.09 <.0001 1.45 0.2305 197.9 <.0001 159.5 <.0001 168.17 <.0001
Freq*Cross 3 179–186 0.24 0.8667 1.55 0.2036 1.66 0.1778 4.09 0.0078 2.81 0.0412
Interspec*Cross 3 179–186 1.18 0.3178 0.75 0.5227 1.06 0.3679 1.8 0.1492 1.88 0.1341
Dens*Cross 6 179–186 0.59 0.7406 1.48 0.1875 14.68 <.0001 13.39 <.0001 11.52 <.0001
Freq*Interspec*Cross 3 179–186 0.29 0.8338 0.68 0.5654 0.93 0.4292 0.39 0.7576 0.46 0.7125
Freq*Dens*Cross 6 179–186 0.59 0.7414 0.49 0.8174 0.69 0.6548 0.48 0.8259 0.63 0.7057
Interspec*Dens*Cross 6 179–186 0.31 0.9306 0.33 0.9185 2.35 0.033 2.19 0.046 1.4 0.2167
Freq*Interspec*Dens*Cross4 6 179–186 0.94 0.4668 – – 0.12 0.9945 0.22 0.9687 0.55 0.7677
Meristem Damage 3 1941 – – – – 3.95 0.0485 0.11 0.7377 2.53 0.1138

Notes: These analyses explore the effects of experimental treatments on the probability that focal seeds emerged and survived to reproduce, the number of seeds produced by reproductive plants, the number of seeds produced by emerged seedlings, and the number of seeds produced by each seed planted the previous fall (see text for details). P-values>0.05 are bolded to emphasize significance.

1

Binary life cycle data (probability of emergence and survival to reproduction) analyzed using residual pseudo-likelihood and the logit link function in SAS Glimmix.

2

Seeds per reproductive plant, Seeds per emerged seedling, and Seeds per planted seed were all natural log transformed (ln(y+1)).

3

ndf and ddf, numerator and denominator degrees of freedom for each trait, respectively.

4

This 4-way interaction was left out of the probability of survival to reproduction analysis because the model could not otherwise converge.