Abstract
With many plant-pollinator interactions undergoing change as species’
distributions shift, we require a better understanding of how the
addition of new interacting partners, such as antagonists, can affect
plant reproduction. One such group of antagonistic floral visitors,
nectar robbers, can deplete plants of nectar rewards without
contributing to pollination. The addition of nectar robbing to the
floral visitor assemblage could therefore have costs to the plant´s
reproductive output. We focus on a recent plant colonist,Digitalis purpurea , a plant that in its native range is rarely
robbed, but experiences intense nectar robbing in areas it has been
introduced to. Here, we test the costs to reproduction following
experimental nectar robbing. To identify any changes in the behaviour of
the principal pollinators in response to nectar robbing, we measured
visitation rates, visit duration, proportion of flowers visited and rate
of rejection of inflorescences. To find the effects of robbing on
fitness, we used proxies for female and male components of reproductive
output, by measuring the seeds produced per fruit and the pollen export
respectively. Nectar robbing significantly reduced the rate of
visitation and lengths of visits by bumblebees. Additionally, bumblebees
visited a lower proportion of flowers on an inflorescence that had
robbed flowers. We found that flowers in the robbed treatment produced
significantly fewer seeds per fruit on average but did not export fewer
pollen grains. Our finding that robbing leads to reduced seed production
could be due to fewer and shorter visits to flowers leading to less
effective pollination. We discuss the potential consequences of new
pollinator environments, such as exposure to nectar robbing, for plant
reproduction.
Keywords: bumblebee, nectar robbing, Digitalis purpurea ,
antagonism, pollination.
Introduction
Many plant-pollinator interactions are undergoing change due to multiple
anthropogenic influences (González-Varo et al . 2013, Goulsonet al ., 2015). At present we have a poor understanding of how
alterations in the pollinator environment affect plant reproductive
success. One interaction with consequences for plants is nectar robbing,
where plants have their mutualism with pollinators bypassed by floral
visitors (robbers) that consume nectar rewards without pollinating. In
the same way as pollination mutualisms, this antagonistic interaction
between plants and nectar robbers can also be altered as a consequence
of changes in abundance or distributions of the plants or floral
visitors (Irwin & Maloof 2002; Traveset et al ., 1998). In turn,
changes in the incidence of nectar robbing have the potential to affect
a plant´s reproductive success.
During nectar robbing, a floral visitor bites a hole in the corolla
(“primary robbing”) or utilises an existing hole previously created by
another robber (“secondary robbing”) to feed from nectar, which often
results in no contact with the stigmas or anthers and hence no
contribution to pollination (Inouye 1983; Rojas-Nossa et al .,
2016). Some previous studies found limited or no negative fitness
consequences of robbing for the plant (Richman et al ., 2018;
Stout et al ., 2000) with some examples of robbing increasing
plant fitness through increasing pollen flow and dispersal distance
(Higashi et al ., 1988; Maloof & Inouye 2000) and increasing the
frequency of visitation from legitimate pollinators (Stout et
al ., 2000). However, other studies have reported detrimental effects on
at least one component of the plant’s reproductive success (Adleret al ., 2016; Burkle et al ., 2007; Lara & Ornelas 2001;
Castro et al. , 2008; Irwin & Brody 1999). Negative effects of
robbers include damage to the reproductive organs, a reduction of the
attractiveness of the floral display, and exhaustion of the nectar
reward, all of which could potentially alter the foraging behaviour of
legitimate pollinators that are required for plant reproductive success
(Irwin et al ., 2010). The extent to which robbers affect plant
fitness could depend on the frequency, the amount of damage done, and
how much the behaviours of the legitimate pollinators are affected
(Adler et al., 2016; Irwin et al ., 2010). Additionally, if
extra resources are allocated towards nectar production in the plant in
response to robbing, this could have a detrimental effect on the number
of seeds and/or fruits produced (Navarro 2001; Pyke 1991; Southwick
1984). In this study we focus on the effects of robbing on a plant that,
after range expansion, experiences a high level of nectar robbing that
is not present in the native range.
The common foxglove, bumblebee-pollinated Digitalis purpurea,expanded its range from native European woodland to areas including
tropical mountains in Central and South America following anthropogenic
introductions (Mackin et al. 2021). As a consequence, the plant now
experiences geographically variable rates of nectar robbing. To our
knowledge, there are few records of nectar robbing in native D.
purpurea populations across Europe, associated with robbing specialistBombus wurflenii (Reinig & Rasmont 1988), whereas in American
populations we found that the plants are robbed at a high rate. For
example, in preliminary observations in 110 plants across two non-native
populations in Colombia, we recorded that 288 out of 677 (42.5%)
post-anthesis flowers had been robbed at least once (pers. obs.). In
these populations, the bumblebees Bombus hortulanus and B.
rubicundus, and additionally some species of hummingbird and flower
piercers (specialised robbers in the genus Diglossa ), frequently
feed on nectar from D. purpurea by robbing the flowers. In Costa
Rica, the high-altitude bumblebee B. ephippiatus is the main
pollinator of the plant but also a frequent robber. Frequently,
individual bumblebees use a mixed foraging strategy on D.
purpurea , robbing and visiting legitimately on the same foraging bout.
Here we test the costs of novel levels of nectar robbing on D.
purpurea by experimentally robbing plants and measuring the effect on
pollinator behaviour and on the plant’s reproductive success. We expect
nectar robbing will reduce the volume of nectar or alter the rate of
nectar production, causing bees to visit flowers at a lower rate
(Parachnowitsch et al 2018). Additionally, it is possible that robbing
could also reduce energetic resources available for fruit production,
regardless of effects on bee visitation (Navarro 2001). We measured both
male and female components of reproduction for plants and quantified the
visitation rates by naturally foraging bumblebees to each treatment. We
also recorded other measures of bumblebee visitation patterns on
inflorescences, including time spent visiting flowers and proportion of
an inflorescence’s flowers that were visited in a foraging bout.
Materials and Methods