Recently Brokaw and Busing argued that there is limited evidence for niche partitioning of tree species within forest gaps1. Consequently gaps appear to play a relatively minor role in the maintenance of tree species diversity in forests via traditional resource partitioning. This conclusion is strongly supported by the existing empirical evidence particularly for shade-tolerant tree species. Most studies of gaps however have failed to take into account plant groups other than trees2-4. Gaps may be a necessary habitat for the persistence of a large proportion of the vascular plant species other than shade-tolerant trees; specifically pioneer trees lianas herbs shrubs and herbaceous vines2-8. For example in a study on Barro Colorado Island (BCI) Panama gaps had higher liana and pioneer tree diversity on both a per area and per stem basis (thereby removing the effect of density) compared to the surrounding forest23. These two plant groups alone account for approximately 43% of the woody species in this tropical forest2. There is also evidence that many forest herbs are gap dependent56. The role of gaps in the maintenance of shrubs is less clear although there is some evidence that gaps promote shrub growth and reproduction78. Overall when the major vascular plant groups are considered as much as 65% of the flora of BCI may be gap dependent (Table 1). The specific mechanism that leads to the higher diversity of these groups in gaps remains unknown. Nonetheless because these vascular plant groups represent a majority of the plant species in tropical forests worldwide49 gaps may often play a strong role in the maintenance of species diversity. Brokaw and Busing also argued that gaps might maintain diversity via the density effect10; Specifically that gaps will have a higher diversity of trees solely because they have a higher density of trees compared to the surrounding forest. Tree density in gaps however declines (thins) with age and thus the density effect could maintain diversity in the mature forest primarily in two ways. First if individuals in gaps reach reproductive age prior to thinning then they could potentially colonize new gaps. Data are lacking however on whether trees reach reproductive age sooner (i.e. smaller size or age class) in gaps than in non-gap sites. Second there must be niche partitioning. Without niche partitioning thinning of individuals occurs randomly and the initial increase in diversity would be merely a transitory result of the short-term increase in plant density211. Consequently given the scanty evidence for niche partitioning and accelerated reproduction in gaps the evidence for the density effect as a viable mechanism to explain the maintenance of diversity in forests is equivocal at best. We argue that papers sounding the death knell for the role of gaps in the maintenance of diversity in forests (e.g. Ref. 11) may be premature. The focus of most previous research on the ability of tree species to partition resources in gaps may have caused us to overlook the importance of gaps for many other groups of vascular plants (Table 1). Future research is necessary to quantify further the proportion of species in these and other groups (e.g. epiphytes) that require gaps for persistence in the community.