Avian seed dispersal out of the forests: A view through the lens of Pleistocene landscapes

Abstract

1 INTRODUCTION

The way we tackle and interpret how ecological communities respond to anthropogenic impacts largely depends on our baseline, that is, on the scenarios we envisage prior to human intervention (Pausas & Bond, 2019; Vera, 2010). During the 19th and 20th centuries, the baseline for the temperate lowlands of Europe was a continuous primeval forest dominated by broadleaved deciduous trees (reviewed by Vera, 2000). This hypothesis was mainly founded on the observation that fields and meadows can spontaneously develop into forests after the abandonment of agriculture and livestock farming. The baseline of the primeval forest was challenged by Frans Vera (2000) in his influential book Grazing Ecology and Forest History where the author presented diverse and compelling evidence supporting an alternative hypothesis: grazing and browsing by extinct megaherbivores should have created and maintained wood pastures, which must have been a widespread landscape type in post-glacial temperate Europe (Vera, 2000). The book contains perspectives from succession ecology, plant regeneration, palynology, paleoecology, history and even linguistics (Vera, 2000). Recently, a large-scale palynological study (Pearce et al., 2023) has provided further support to Vera's hypothesis by estimating that light woodland (shade-intolerant taxa) and open vegetation covered more than half of European landscapes during the Last Interglacial period (129,000–116,000 years ago). In parallel, another study has estimated a loss of ~95% of community-wide biomass of European megafauna (wild terrestrial mammals ≥10 kg) since the Last Interglacial along with the loss of the largest species of megaherbivores, including elephants and rhinos (Davoli et al., 2024). In addition, a further palynological study suggests that the vegetation composition in the Last Interglacial is better explained by the role of megaherbivores than by fire regimes (Pearce et al., 2024).

Hence, current evidence supports that the temperate zone of Europe was not a closed continuous forest, but a more heterogeneous biome with open vegetation (mainly grasslands), light woodlands and forests (Pearce et al., 2023) that held a diverse community of large herbivorous mammals (Davoli et al., 2024; Svenning et al., 2024). This is congruent with the fact that many temperate woody plants are light-demanding species that are often restricted to forest edges because they fail to regenerate in the shaded forest interiors (Vera, 2000). Moreover, many light-demanding plants are thorny (e.g. Berberis vulgaris, Crataegus spp., Prunus spinosa, Rosa spp., Rubus spp. and Ulex europaeus) or have prickly leaves (e.g. Ilex aquifolium and Juniperus spp.). These are defensive traits against herbivorous mammals that make these species common in grazed and browsed landscapes (Vera, 2000). Importantly, the thorny shrubs facilitate the establishment of non-thorny species, including light-demanding oaks (Quercus spp.), which benefit from the anti-herbivore protection of their nurse plants (Bakker et al., 2004; García & Obeso, 2003; Martínez & García, 2017; Olff et al., 1999; Vera, 2000).

Notably, most of the light-demanding woody species in temperate Europe are dispersed by birds and mammals via two different mechanisms: endozoochory and synzoochory (van der Pijl, 1982). On the one hand, tens of species produce fleshy fruits (e.g. genera Berberis, Cornus, Crataegus, Euonymus, Frangula, Hedera, Ilex, Ligustrum, Malus, Rhamnus, Rosa, Rubus, Prunus, Pyrus, Sambucus, Sorbus, Viburnum), fleshy cones (e.g. genus Juniperus) or arylated seeds (e.g. Taxus baccata), which are consumed by frugivorous animals, mostly birds (González-Varo et al., 2023; Rumeu et al., 2020). Frugivores transport viable seeds internally until they eject them via defecation or regurgitation (i.e. endozoochory; Jordano, 2014). On the other hand, the acorns of oaks (genus Quercus) and the nuts of hazel (Corylus avellana) foster dispersal by seed-caching animals, mostly corvids and mice (Kollmann & Schill, 1996) that, for different reasons, do not consume all the scatter-hoarded seeds (i.e. synzoochory; Gómez et al., 2019). Synzoochory by mice generally occurs within very short distances (<20 m; Kollmann & Schill, 1996), thus, only corvids play an important role at the landscape scale by dispersing acorns and nuts over hundreds of meters (Pesendorfer et al., 2016).

Besides wind dispersal of pioneer trees (e.g. Alnus, Betula, Populus, Salix), animal-mediated seed dispersal sets up a starting template for woody vegetation dynamics and community assembly. Therefore, knowledge on how birds disperse seeds at the landscape scale is essential to understand the functioning of woodlands and to guide their restoration (Carlo & Morales, 2016; González-Varo et al., 2023; Martínez & García, 2017; Pesendorfer et al., 2016). Here, I reflect on the current knowledge on avian seed dispersal of woody plants in the fragmented anthropogenic landscapes of Europe from a Pleistocene perspective. In other words, I interpret the present-day patterns of seed dispersal in and outside forest patches by considering that the baseline were heterogenous landscapes with light woodlands and open vegetation inhabited by megaherbivores. The aim of this exercise is to discuss linkages between past and present landscapes, seeking an historical understanding of the high spatial complementarity by which different bird species are known to disperse seeds in and out of European forests. I also argue on a likely and important difference regarding the landscape patterns: the scale of woodland openness and sharpness of habitat boundaries in the Pleistocene must have been very different to that of anthropogenic deforestation due to modern agriculture. Finally, I briefly discuss the generality to other biogeographical regions of the main ideas addressed in this article.