Floral traits are associated with the quality but not quantity of heterospecific stigmatic pollen loads

BMC Ecology

Table 2 Importance of measured floral traits for module separation based on random forest analysis

Floral trait	EB II	EB III	HT I	HT II	HT III	RB I	RB II	RB III
Stamen length	13.76	4.94	1.58	9.84	11.08	2.94	14.11	12.01/13.77
Inflor. diameter	17.45	2.52	2.40	7.53	14.12	2.87	10.86	16.40/18.35
Nectar tube depth	10.38	1.27	3.88	8.01	7.11	6.67	13.64	11.88/13.01
Nectar tube width	13.28	3.93	2.81	10.93	5.06	3.12	9.11	15.36/15.66
Display size	16.50	3.73	2.12	9.32	11.99	4.74	13.63	17.36/18.36
Style length	18.19	2.29	3.87	11.62	8.96	6.45	17.94	12.83/15.96

Larger numbers indicate greater importance for module classification. Inflor. diameter = Inflorescence diameter. Location: Eselsburger Tal (EB), Hirschtal (HT), Reichenbach (RB). Sampling date: early summer (I), mid-summer (II), late summer (III). For RB III, two solutions of the modularity algorithm [43] are shown as both were equally likely: solution one with four modules / solution two with five modules because one module was split into two. Eselsburger Tal I is not shown as no modules were detected

ISSN: 1472-6785