A cleistogamous species consists of individuals that produce both chasmogamous (open, CH) and cleistogamous (permanently closed, CL) flowers, which facilitates a mixed-mating system. In contrast to what one might expect, CL (obligately selfed) seeds and the plants derived from them can be more fit than CH (potentially outcrossed) seeds and the plants they give rise to. Our aim was to review some theoretical aspects of mixed mating in relation to retention of both CH and CL in cleistogamous species and to determine if data on germination support the notion that CL is advantageous over that of CH. Based on germination (or seedling emergence) of CH vs CL seeds in 29 species in 21 genera and 11 families of monocots and eudicots, CL seeds germinated better in 107 and equally well as in 64 of 252 case studies as CH seeds (67.9%), and the (CH < CL):(CH > CL) ratio was 107/81 (1.32). We conclude that our study lends support to the notion that production of CL seeds by cleistogamous species is advantageous over that of CH seeds. Retention of CH by CL species may be due to the need to prevent complete selfing (s= 1.0) and thus total inbreeding depression (δ), which theory predicts would decrease reproductive success. Some caveats concerning the results of comparative studies on the germination biology of CH vs CL seeds of amphicarpic sensu stricto Fabaceae species and Commelina benghalensis and the aerial cleistogamous grass Danthonia spicata are discussed.

Theory predicts that not all plant species will be equally affected by disruption to their pollinator mutualisms because traits such as breeding system and mutualism strength can affect their response. We investigated these traits in three species of Afromontane mistletoe Globimetula braunii, Agelanthus brunneus and A. djurensis in Ngel Nyaki Forest Reserve, Nigeria, to test whether the traits were reliable indicators of relative reproductive success and could be used to predict relative vulnerability to pollinator loss. For each mistletoe species, insect and bird visitors were identified during a 160–240-min observation period of 4–10 plants and their roles in flower opening and fruit set were investigated using exclusion experiments applied to 250–500 flowers. We found that all three mistletoes are self-compatible but not capable of autonomous self-fertilization. The pollinator assemblage comprised four species of sunbird (Cyanomitris spp., Cinnyris spp.) and a small social wasp (Vespinae). None of the mistletoes requires birds for flower opening: G. braunii flowers self-opened in the absence of pollinators, whereas insects opened both Agelanthus spp. Irrespective of flower opening, each mistletoe species requires sunbirds for effective pollination and fruit set. Only G. braunii demonstrated pollen limitation (pollen limitation index = 0.504) which may be an early indication of mutualism breakdown. We suggest that mistletoes be considered as indicators of habitat condition and functioning within Afromontane forest ecosystems.

There is now considerable evidence for disruption of plant–pollinator mutualisms in fragmented landscapes. The causes can be myriad, and vary from lack of nesting sites for key insect pollinators, to a decline in pollinator visits when plant population sizes are reduced by fragmentation. One important and well-documented consequence of disruption in pollination systems is reduced seed production in plant populations in fragments as a result of pollen-limitation. When reduced seed production, in turn, has a negative effect on seedling recruitment, populations may enter a spiral towards extinction. Some pollination systems, on the other hand, are surprisingly resilient to anthropogenic changes, and pollen flow between trees in fragments separated by several kilometres has been reported recently. Understanding how reproductive attributes of plants, such as breeding systems and specificity of pollination systems, affect the outcome of habitat fragmentation remains a key challenge.