http://www.ncbi.nlm.nih.gov/pmc/articles/PMC1450323/Vitamin B8 (biotin): Egg yolk, avocado, salmon, bananas, apples, mackerel and strawberries contain most vitamin B8.
According to this study, banana's and apples are not a rich source of biotin.
Biotin content (ng biotin/g food) of select foods determined by HPLC/avidin binding assay:
Chicken liver, cooked 1872
Beef liver, cooked 416
Egg, yolk, cooked 272
Egg, whole, cooked 214
Peanuts, roasted, salted 175
Salmon, pink, canned in water 59
Egg, white, cooked 58
Almonds, roasted, salted 44.07
Walnuts, fresh 25.9
Pecans, fresh 20.0
Strawberries, fresh 15.0
Avocado, fresh 9.61
Tomatoes, fresh 7.01
Orange juice, canned, from concentrate 4.13
Banana, fresh 1.33
Apple, fresh 0.20
According this study this should be a more appropiate list:
Vitamin B8 (biotin): Egg yolk, salmon, allmonds, walnuts, pecans and strawberries contain most vitamin B8.
Although I'm not sure how cooking/roasting affects the values of egg yolks/salmon/almonds.
This study don't give any information of mackerel.
I don't know if chicken liver, beef liver are included in Wai Diet?
This study concludes that previously published values for biotin content of foods are likely to be inaccurate:
The differences between published values and values determined here may arise from one or more of the following: (1) Heterogeneity in foods assayed might have arisen from differences in growing conditions, biotin content of animal feeds or fertilizers, fortification, processing, season, and geographic location. Food samples presented here were not pooled from more than one source. (2) Analytical problems are likely to have contributed to the differences with respect to published values. Published biotin values likely both over- and underestimate true biotin content of particular foods because the majority of these studies used bioassays. Bioassays characteristically suffer from interference by unrelated substances, nonspecificity for biotin versus its metabolites, incomplete release of biotin from its protein-bound state, destruction of biotin during release from protein binding, and release of substances that cause assay artifacts during acid hydrolysis (Mock et al., 1992). Bioassays may measure ‘sparing factors’ and thereby overestimate biotin. Bioassays do not consistently discriminate biotin from its inactive metabolites. Biotin in food is largely protein bound to both endogenous and exogenous biotinyl proteins. The proportion of free (water extractable) biotin versus protein-bound (released by acidic or enzymatic hydrolysis) biotin varies among foods; the majority of biotin in meats and cereals appears to be protein bound (Zempleni and Mock, 1999). Furthermore, published biotin values rely heavily on values for compound foods reported by the food manufacturer (Pennington and Church, 1985; Pennington, 1989). These manufacturers do not necessarily specify methods for biotin analysis and may over- or underestimate biotin content in food values due to the same factors listed previously.