POLLINATORS POLLINATION AND FOOD PRODUCTION

THE ASSESSMENT REPORT ON POLLINATORS, POLLINATION AND FOOD PRODUCTION
the soil (Hopwood et al. 2015). A recent study on prescribed
burning and the imperiled mardon skipper (Polites mardon)
in California showed substantially fewer butterflies in the
burned areas of meadows compared to unburned areas
after 1, 2, 3 and 5 years following the burn event (Black
et al., 2014). Queen bumble bees overwintering in small
cavities just below or on the ground surface are at risk, as
are ground-nesting bee species that nest in shallow burrows
(Cane and Neff, 2011). Solitary bees nesting in stems or
twigs are unlikely to survive the heat of burns (Cane and
Neff, 2011), and stem-nesting bee populations will only
recover postfire when the availability of suitable stems
increases over time (Potts et al., 2005). The loss of bees due
to a burn can lead to reduced fruit set in plants in burned
areas (Ne’eman et al., 2000).
Recovery of pollinators following a burn varies between
guilds. Though losses of bees following a fire can be
catastrophic, bees may be able to recolonize burned sites
and recover within a few years (Potts et al., 2003). Habitatdependent
or -specialist species and those that are less
mobile are most likely to be negatively affected immediately
by a fire (Panzer, 2002; Vogel et al., 2010). A pollinator
species’ ability to cope with regular burns is dependent on
there being adequate unburned adjacent areas that can
provide sources of colonizers into the burned land cover
type (e.g., Harper et al., 2000; Hartley et al., 2007; Panzer,
2002; Swengel 2001). Isolated populations of pollinators
in small fragments may not survive repeated prescribed
burns (Panzer 2002) because there are often no source
populations available for recolonization once a population
has been locally extirpated. Burning a small fragment in
its entirety could risk eliminating some species because of
limited recolonization from adjacent patches (Harper et al.,
2000). This accentuates the need to leave substantial land
cover patches when using fire as a management tool. Land
cover patches should not be burned completely; rather,
a mosaic of burned and unburned areas is ideal. Besides
controlled grazing and mowing, prairies (ecosystems
considered part of the temperate grasslands, savannas,
and shrublands biome, typically in North America) can be
managed through prescribed burning. A large experimental
study demonstrated that different butterfly species have
varied responses to prairie management through fire.
While prairie specialists responded negatively to burning,
generalists were largely benefited by this action (Swengel
2001). Moreover, greater durations without burning
benefited specialists but reduced generalists (Swengel
1996). However, there may be some geographic variation
in these results, as it has been shown that burns in oak
savannas in the USA do not harm butterfly diversity
(Siemann et al., 1997).
In the Cape Floristic Region of South Africa, nectarfeeding
bird abundance and species richness was found
to decrease in post-fire vegetation, and floral arrays within
burnt vegetation received no visits by nectar-feeding
birds (Geerts et al., 2012). Some studies, however, have
shown that fire-dependent communities have indirectly
and positively impacted pollinators by altering plant density
and distribution (Van Nuland et al., 2013, Charnley and
Hummel, 2011). Moreover fires in Mediterranean climates
are necessary for seed dispersal and germination (Pausas
and Vallejo, 1999).
2.2.2.2.4 Transformation of agroforestry systems
Agroforestry refers to the practice of integrating trees and
other large woody perennials into farming systems and
throughout the agricultural landscape (Schroth et al., 2004).
While a considerable number of papers show the positive
effects of plant diversity in agroecosystems for bees and
other insect pollinators (see Nicholls and Altieri, 2013, for
a review), considerably less attention has been paid to
understand the effects of agroforestry for bees and other
pollinators. Willemen et al. (2013) revealed a high diversity
of Tree-Based Ecosystem Approaches, including trees
in croplands, trees in grasslands, forest-based systems,
complex multi-strata agroforestry and homegardens. They
report positive impacts for food security and climate change,
but very few of these studies evaluated the impacts of these
systems for pollinators.
Studies in temperate landscapes are particularly infrequent,
although agroforestry has been flagged as a practice
favourable to beekeeping (Hill and Webster, 1995). In
Québec, Alam et al. (2014) estimated the value of ten
ecosystem services in an agroforestry system (tree-based
intercropping), in particular the value of pollinators, and
found that yield and profit could be maximized with the
presence of tree and shrub cover in agricultural landscapes.
Instead, the relatively few field studies on this topic have
been performed in tropical landscapes, where agroforestry
systems are the major agroecosystems that resemble
natural forest, and potentially have high biodiversity and
pollinator conservation value (Tscharntke et al., 2011).
Agroforestry systems are a land use that might aid in
enhancing connectivity between natural and semi-natural
areas (Perfecto and Vandermeer, 2008). In the tropics,
agroforestry may perhaps be one of the most important land
management systems for pollinator conservation, because
the majority of trees are animal pollinated and pollinators
therefore may rely more on floral resources from trees
compared to herbaceous wild flowers (Bawa, 1990).
49
2. DRIVERS OF CHANGE OF POLLINATORS,
POLLINATION NETWORKS AND POLLINATION
Fire can have significant, negative impact on plant
reproductive success and is associated with statistically
significant lower fruit set (McKechnie and Sargent, 2013).
The effects of agroforestry practices on the diversity of
pollinators and pollination have been studied principally for
two tropical crops, coffee and cacao, and show overall the