The vectoring of nutrients and organic matter via the movement of motile consumers has long been recognised as a mechanism that links ecosystems (Hutchinson, 1950). Allochthonous nutrient subsidies support many ecosystems, influencing productivity, functioning, diversity, biomass, and community dynamics, boosting food availability and quality, and enhancing ecosystem resilience (Polis et al., 1997; Lundberg and Moberg, 2003; Loreau and Holt, 2004; Benkwitt et al., 2019). Seabirds are important nutrient vectors, transporting nutrients from their pelagic ocean feeding grounds to coastlines and islands via marine prey items that drop on the colony, lost feathers, abandoned or destroyed eggs, and individuals that die on land. However, the greatest nutrient subsidy is in the form of nitrogen- and phosphorous-rich guano; nutrients that are often limiting resources across many ecosystems. At local scales, these levels of nutrient input dictate the structure and dynamics of communities to the extent that seabirds have been described as “keystone species” (Mulder et al., 2011b). At global scales, seabirds may link marine and terrestrial environments, contributing to inter-habitat connectivity and providing multiple ecosystem services (Otero et al., 2018; Signa et al., 2021).

Whilst the functional role of seabird nutrient inputs on terrestrial habitats has received a great deal of attention, empirical studies investigating the effects of bottom-up nutrients supplied by seabirds on coastal marine ecosystems remain comparatively limited (Kazama, 2019; Signa et al., 2021; Grant et al., 2022). Evidence does exist, though, of nutrient inputs from seabirds moving through systems, such as consistently higher nitrogen signals in primary producers and invertebrates that are in close proximity to seabird colonies (Jones, 2010; Kolb et al., 2010; Gagnon et al., 2013; Rankin and Jones, 2021). However, whilst it is important to understand the extent that seabird nutrient inputs propagate through ecosystems, it is also critical to establish whether these allochthonous nutrients impact biodiversity and productivity. These influences have received recent attention in tropical systems: seabird nutrients led to increases in the growth rates of coral, damselfish, and parrotfish, as well as the biomass of fish communities (Graham et al., 2018; Savage, 2019; Benkwitt et al., 2021a, Benkwitt et al., 2021b, Benkwitt et al., 2023). Some research on the bottom-up effects of seabirds on temperate environments has been performed, examining intertidal community structure, and seabird nutrients impacts on macroalgae. For example, on the Mercury Islands, New Zealand, macroalgal diversity was highest on islands with the longest history of seabird presence (Rankin and Jones, 2021), whilst in the Stockholm archipelago, there was a shift towards increased epiphytic algae and invertebrate biomass near seabird islands (Kolb et al., 2010). Further, in South Africa, higher levels of nutrients due to seabirds resulted in higher algal production (Bosman and Hockey, 1986) and growth rates (Bosman et al., 1986). Whilst research in both tropical and temperate environments exists, more is needed to understand the influence of seabird nutrients to recipient systems, with the United Kingdom (UK) amongst the countries where research is deficient.

Despite their importance, seabirds are one of the most threatened bird groups, with almost half of all seabird species threatened with extinction (IUCN, 2022), and over half of Britain and Ireland’s seabirds are in decline (JNCC, 2023). The threats to seabirds do not only impact their own populations, but could also cascade through ecosystems, threatening communities and food webs that are dependent upon seabirds’ nutrients (McCann et al., 2005; Rooney et al., 2006; Thoresen et al., 2017). Current anthropogenic threats to seabirds include commercial fishing, pollution, habitat degradation, human disturbance, climate change, severe weather events, and human-introduced invasive predators (Croxall et al., 2012). Indeed, invasive predator eradication programmes are becoming an increasingly important conservation technique, with approximately half of all successful European invasive alien vertebrate eradications on inhabited islands having occurred within the UK, including on Staple island of the Farne Islands, Northumberland (Stanbury et al., 2017).

Cold-water systems, such as those surrounding the UK, make up a large proportion of global oceans, yet the impact of seabird nutrient input in these environments remains understudied. Strong seasonal differences exist in these environments, and they are prone to stormy weather, which is predicted to increase under future climate predictions (IPCC, 2022). The UK is of international importance for breeding assemblages of seabirds and shorebirds, with many of its islands designated as Special Protection Areas under EU legislation (Stanbury et al., 2017). These breeding seabirds excrete vast quantities of nutrients along the UK’s coastlines (Riddick et al., 2012). Given the global threats to seabird populations, that cold-water systems make up a large proportion of global oceans, and the international importance of the UK’s seabirds, it is important to determine the influence of seabird nutrient input on UK nearshore marine systems for both regional and global understanding.

To test the impact of seabirds upon terrestrial and intertidal systems in the UK, islands with high densities of breeding seabirds (the Farne Islands) were compared to mainland sites on the adjacent Northumberland coast, where no seabirds breed. It was hypothesised that seabird nutrient input would cycle through the terrestrial and intertidal ecosystems of the Farne Islands, boosting intertidal biodiversity. We also predicted that seabird nutrient input would increase intertidal productivity through increased limpet size and biomass. Rocky shore intertidal systems are subject to regular disturbance and recolonisation by intertidal organisms, which are expected to increase under future climate change. Furthermore, experimentally adding nutrients to rocky intertidal habitats has been shown to increase recolonisation (Kraufvelin et al., 2006; Kraufvelin, 2007; Masterson et al., 2008). Therefore, following experimental clearances it was predicted seabird nutrient input would support a more rapid recolonisation of algae and barnacles recolonisation.

The findings of our study provide evidence that that seabirds are inputting pelagic nutrients to the terrestrial and intertidal ecosystems of temperate UK islands. Thus, seabirds provide a key cross-ecosystem link. However, there was limited evidence for seabird nutrients affecting intertidal community structure. There was some evidence that intertidal diversity was marginally greater on seabird islands (i.e., differences in intertidal fauna communities and functional group densities), while barnacle recolonisation rates were faster at low seabird density mainland sites. Disentangling the effects of seabird nutrients from other potential differences between the seabird islands and mainland locations will enable a better understanding of the role of seabird nutrient subsidies in temperate rocky intertidal systems.

Higher densities of seabirds were found on the Farne Islands, Northumberland, compared to the Northumberland mainland by four orders of magnitude. By foraging in pelagic feeding areas, seabirds were estimated to have vectored between three and four orders of magnitude more nitrogen inputs to the seabird islands compared to the mainland. The seabird density and nitrogen inputs revealed by our study are comparable with those observed in a tropical environment, where seabird density and nitrogen input were both three orders of magnitude greater on rat-free islands than those that had rats present on them (Graham et al., 2018). The higher density of seabirds on the Farne Islands is most likely due to these islands being predator free, in comparison to the mainland sites where predators including rats, cats and foxes would decimate seabird populations in accessible habitats.

The nutrient inputs from seabirds to the Farne Islands are used by terrestrial and intertidal organisms, evidenced by higher δ¹⁵N signals being found in soil, new-growth leaves, and lichen, as well as macroalgae, turf algae, barnacles, and limpets on seabird islands compared to low density seabird sites. The third highest δ¹⁵N values were found in limpets, which is likely explained by δ¹⁵N values increasing with trophic levels, and microphagous limpets grazing on ¹⁵N enriched early macroalgal stages and microalgal film (Jenkins and Hartnoll, 2001). This stepwise enrichment is consistent with other research in marine systems (Hobson and Welch, 1992; Zmudczyńska-Skarbek et al., 2015; Graham et al., 2018), although other studies have found mixed responses (Gagnon et al., 2013), or instances of primary consumers having higher ¹⁵N values than secondary consumers (Hong et al., 2019; Andrades et al., 2024). Significantly elevated nitrogen content was also found in soil, terrestrial plants, lichen, barnacles, and limpets, as well as marginally elevated turf algae nitrogen content. These findings support the hypothesis that seabird nutrient subsidies are cycling through the terrestrial and intertidal habitats. Previous studies in saltwater systems show that the proportion of seabird-derived nutrients decreases with increasing distance from seabird colonies (Jones, 2010; McCauley et al., 2012; Lorrain et al., 2017; Savage, 2019; Benkwitt et al., 2021a). Indeed, a similar pattern was generally observed in this study, with the samples closest to the seabird colonies (terrestrial plants and soil) having the greatest δ¹⁵N values on seabird islands, and macroalgae, epiphytic algae, and turf algae that inhabit intertidal areas further from the colony having lower values. By contrast, total nitrogen content was generally higher within intertidal samples than terrestrial ones, with all the intertidal samples except macroalgae having a higher %N value. This may indicate that nitrogen is also being input directly from the marine environment, such as from the North Sea. Whilst the difference in the patterns of δ¹⁵N and %N could be argued to contradict each other to an extent, the high δ¹⁵N values show that, despite other sources of nutrient input to the system, pelagic-derived seabird nutrients make up a dominant component of the Farne Island terrestrial and intertidal nutrient budgets.

Compared with tropical or polar regions, cold temperate regions are typically very productive and may provide high levels of background nutrients to the region (Sigman and Hain, 2012). Furthermore, two possible sources of nitrogen input exist in two areas close to the mainland study sites. First, there are reported marine nitrate toxicity issues in the River Tweed, around eight to ten kilometres north of the mainland study sites (Jarvie et al., 2002). Second, there is eutrophication at Lindisfarne National Nature Reserve, around 25 km north of the mainland study sites, where sources of nitrogen include diffuse agricultural runoff, misconnections in public and private sewage treatment works, and seasonal tourism pressures on dated sewage treatment facilities (NCC, 2020). Sewage δ¹⁵N signatures are comparable to those of animal waste and are almost always over 10‰, reaching up to around 20‰ (Heaton, 1986). In our study sites, mainland δ¹⁵N signals were below 10‰ in each sample type, providing evidence that sewage is not a major source of nitrogen input to these areas. Fertiliser δ¹⁵N signatures typically range from -2 to 2‰, though can reach values of 4‰ (Bateman and Kelly, 2007). A global study of soil δ¹⁵N trends estimated that UK δ¹⁵N values range between 3.5 and 6.2‰ (Amundson et al., 2003), and indeed, on the mainland study sites, δ¹⁵N signatures were 5.4 ± 0.3‰. Given that fertiliser δ¹⁵N values are lower than those of both typical soil and those observed on the mainland, δ¹⁵N cannot reliably be used to determine whether fertiliser is a source of nitrogen input to the mainland sites.

As well as being seabird breeding colonies, seals also inhabit the Farne Islands, predominantly grey seals (Halichoerus grypus), who use the islands as rookeries during their breeding season from October to January. Foraging distance in grey seals from the Farne Islands varies from short trips to known haul-out sites (mean distance 39.8 km) to long and distant journeys (2100 km; McConnell et al., 1999). Therefore, it is likely that seals are also providing an allochthonous source of nutrients to the Farne Islands. Furthermore, seal biomass exceeds that of seabird biomass on two of the three study islands (National Trust, 2022b). However, seabird abundance is approximately fifty times higher than seal abundance across the three islands (Webber et al., 2015; National Trust, 2022a, National Trust, 2022b). In terms of seal proximity to the sites on the seabird islands, it was confirmed by National Trust rangers working on the Farne Islands that none of the study sites were seal rookeries, and that the seabird colonies were closer to each study site than seal rookeries were. Some seabird species, like the gulls, also travel across the islands frequently, whereas the seals tend to haul out in one spot. Whilst the particularly high δ¹⁵N in the terrestrial samples suggest that it is the seabirds are the dominant cause of increased nitrogen loads to the terrestrial environments, it is clear that the seals are also providing a key source of nutrient input to the Farne Islands.

Nutrient input from seabirds can increase primary producer biomass, resulting in heightened algal herbivory and herbivore biomass, leading to increased resources for intertidal predators (Methratta, 2004). The similar densities of macroalgae, turf algae, sessile filter feeders, and other herbivores observed in this study indicate that the intertidal communities on seabird islands and the mainland are functionally similar. However, dogwhelks comprised 81% of the studied predators on seabird islands, yet were absent from mainland sites. Moreover, all sites were relatively sheltered and fairly similar in terms of wave exposure (on the Ballantine Index all mainland sites were considered slightly more exposed than those on the seabird islands: mainland sites = 5; Farne Islands = 6, 7 and 7). Given that typically dogwhelks are more common on exposed rocky shores than sheltered coasts (Ballantine, 1961), dogwhelks would have been expected to be more common on the mainland sites, but here, we found the opposite to be true. Thus, another cause of higher densities of intertidal predators could be that seabird nutrient subsidies result in bottom-up interactions, possibly cascading productivity up the food web to create greater predator densities (Young et al., 2011). While it is possible that the increased seabird nutrients on the Farne Islands are supporting this additional trophic level, on its own this evidence falls short of substantiating this effect.

When comparing species diversity and taxonomic relatedness, we found that the intertidal communities surrounding seabird islands and mainland sites were similar and that there was no observed difference in fauna species richness, untransformed community species diversity, or evenness for either algal or fauna species. A non-significant trend towards higher algal species richness was observed, though, similar to higher macroalgal diversity on islands with the longest history of seabird presence in New Zealand (Rankin and Jones, 2021). Overall, though, the diversity effects of seabird nutrients in the present study were small. Drivers of ecological community diversity are highly complex and it is possible confounding variables exist between the seabird islands and the mainland in this study, as well as high levels of nutrients already being present in the cold-water system. In addition, whilst the exposure levels between the mainland and the seabird islands were relatively similar (Ballantine Index of mainland sites = 5; Farne Islands = 6 and 7), those differences are large enough for some different community patterns to exist (Ballantine, 1961). Island biogeography may also have influenced differences in community patterns (MacArthur and Wilson, 1967). However, island sites were less than five kilometres from the mainland and were close to one another as well as the other surrounding islands of the Farne Islands archipelago, thereby reducing the likelihood of this impact (Figure 1). The effects of island biogeography are also more relevant to terrestrial species than marine or intertidal species, due to differences in their dispersal mechanisms, which formed the focus of this study (Carr et al., 2003; Hachich et al., 2015; Strathmann, 2015; Pinheiro et al., 2017).

Although no statistical difference was detected in limpet size, abundance or biomass between seabird islands and the mainland, the effect size between the mainland and seabird islands was high, with mean limpet size 20% larger on seabird islands. It is possible that greater sampling is needed to reduce variability and detect statistically significant differences. The relationship between body size and population density within communities is central to explaining community resource use (Pagel et al., 1991), and indeed, the quantity of algae grazed is more closely related to the biomass of limpets than their abundance (Branch, 1971; Bosman and Hockey, 1988). Shell growth in common limpets is rapid in their first two years, with growth rates between 0.4 and 2.0 mm/month. This rate then decreases over time, with a negative linear relationship existing between initial length and increment each year. By five years of age, they exhibit no measurable growth rate. Indeed, common limpet individuals that are five years and older are few, and form only negligible proportions of limpet populations (Blackmore, 1969). Therefore, limpets attaining a larger size may suggest that their growth rates are more rapid (Branch, 1974; Balaparameswara Rao, 1976). If indeed seabird nutrient input is increasing limpet growth rate on the Farne Islands, then this is consistent with previous findings of faster limpet growth in the presence of seabirds (Bosman and Hockey, 1988). Enhanced growth rates have been observed in other marine organisms such as corals and fishes due to the bottom-up effects of allochthonous nutrients (Graham et al., 2018; Savage, 2019; Benkwitt et al., 2021b, Benkwitt et al., 2023). Another possibility is that limpet growth rates is related to population density in both populations. The lower population densities and higher body size on the seabird islands, and vice versa on the mainland, could be explained by density-dependent growth rates in response to the limited resources of space and food. This pattern has been observed in the closely related species, Patella depressa, and is further supported by the similar biomass between the seabird islands and the mainland (Boaventura et al., 2003).

Contrary to predictions, recolonisation rates were faster on the mainland than seabird islands for both barnacles and limpets, which was inconsistent with other studies (Kraufvelin et al., 2006; Kraufvelin, 2007; Masterson et al., 2008). Whilst we sought to minimise the impact of confounding variables in the study, the mainland and seabird islands are inherently physically and biologically different, and it is likely that seabird nutrient input is not the dominant driver of demographic differences between the communities. For example, lower recolonisation rates are expected on islands with increasing isolation (MacArthur and Wilson, 1967), but all of our study islands were close to the mainland (<5 km) so it is unlikely that the Theory of Island Biogeography alone explains the higher rates of recolonisation observed on the mainland. Barnacle abundances increase with heightened wave exposure levels, with the most sheltered shores being largely absent of barnacles (Ballantine, 1961). Indeed, the three sites on the seabird islands were more sheltered from the prevailing UK west to south-westerly winds compared to the mainland. Furthermore, biological barnacle settlement cues will also impact community differences, these including the presence of algae species and diatoms, the abundance/fineness of detrital matter, microheterogeneity, food supply, and chlorophyll a concentration (Le Tourneux and Bourget, 1988; Hills and Thomason, 1998; Burrows et al., 2010). Faster limpet recolonisation was also observed on mainland sites than on seabird islands, with limpet abundance at week eight driving the difference. It is likely that limpets were more abundant on the recently disturbed plots as this disturbance drove increases in the abundance of microphagous limpets’ preferred food type within the plots: a microalgal film largely consisting of diatoms and cyanobacteria (Jenkins and Hartnoll, 2001). It is also possible that the observed increase in barnacles recolonisation was in part driven by the increased limpet recolonisation, as has been observed in other empirical research (Benedetti-Cecchi, 2000).

To conclude, strong evidence was found of seabird nutrients moving through UK terrestrial and intertidal ecosystems, through elevated δ¹⁵N signals and total nitrogen content. This study suggests a clear link by seabirds from pelagic to terrestrial and intertidal habitats. Intertidal diversity differences between the Farne Islands and the low seabird abundance mainland were small, though higher numbers of intertidal dogwhelks and weak evidence of greater algal species richness did offer some support of community differences on the seabird islands. However, cold-water systems typically have high levels of background nutrients, and anthropogenic nutrient inputs from the mainland could not be quantified, so community differences between the islands and the mainland cannot definitively be attributed to seabird nutrients. This research shows the importance of seabirds in connecting ecosystems, but further work is necessary to disentangle the influence of seabird nutrient inputs compared to other factors that structure temperate rocky intertidal ecosystems.

The raw data supporting the conclusions of this article will be made available by the authors, without undue reservation.

The animal study was approved by Lancaster University Animal Welfare and Ethics Review Board. The study was conducted in accordance with the local legislation and institutional requirements.

SH: Writing – review & editing, Writing – original draft, Visualization, Software, Resources, Project administration, Methodology, Investigation, Formal analysis, Data curation, Conceptualization. CB: Writing – review & editing, Supervision, Resources, Methodology, Investigation, Funding acquisition. RD: Writing – review & editing, Visualization, Supervision, Resources, Methodology, Investigation, Funding acquisition. NG: Writing – review & editing, Validation, Supervision, Resources, Project administration, Methodology, Funding acquisition, Conceptualization.