Wintering gulls as key vectors for dispersal of seeds, nutrients and contaminants in Andalusia

Abstract

Ecosystems services provided by waterbirds are diverse, and processes such as seed dispersal and biovectoting are among the most important. Due to global change, those services may become “disservices” if the seeds dispersed by waterbirds are inconvenient (e.g. weeds), or the biovectoring of nutrients and contaminants to wetlands that are already vulnerable to global change. Previous studies have shown the role of waterbirds on ecosystem services, but the occurrence of new GPS tracking technologies allow to make more realistic quantification of such services by detailed movements. In this PhD thesis, I focus in the role of the lesser black-backed gull Larus fuscus on seed dispersal and biovectoting at Andalusian scale. The availability of wintering detailed GPS movements, along with the species capacity to exploit several habitats, make the lesser black-backed gull a good model species. In Chapter 1, the combination of GPS movements with network analyses allowed the creation of a connectivity network between the most exploited habitats by the lesser black-backed gull. In this way, I could determine where and in which frequency the different connections between most important habitats were occurring. Therefore, those connections were involved in the seed dispersal and biovectoring processes. Landfills and ricefields are the most important habitats to maintain the connectivity generated by gulls, and the connections in which an aquatic systems act as a final source are especially relevant. Most of the conections occur within a 60 km range, so different functional units within Andalusia may be created. Within such distance, biovectoring and dispersal processes are more likely to occur. Chapter 2 is focused solely in Doñana ricefields and it compares the dispersal potential of two generalist species that feed within the ricefields in the same way: lesser black-backed gull and white stork. Faeces and pellet analyses and germination experiments under laboratory conditions showed that the community of plants dispersed is extense, covering a range of 21 taxa, including weeds. However, there was no great variation in the community of plants dispersed by the two bird species, despite the differences in body mass. After determining the main weed species dispersed by gulls, the objective of Chapter 3 is to combine GPS gull movements with seed retention times (obtained theoretically and expertimentally with captive gulls) to generate spatial seed dispersal models taking Doñana ricefields as a starting point. Due to the vast ricefield area, high percent of the seeds were deposited within the ricefields, so gulls may have an important dispersal role of weed population homogeneization. Other seeds were deposited outside of the ricefield area, with dispersal distances up to 150 km. Taking into account waterbirds for plant dispersal distances has important implications in species connectivity between suitable habitats as well as weed dispersal towards other agricultural environments. In Chapter 4, the objective is to quantify the external nutrient loading to Fuente de Piedra Lake (Málaga, Andalusia) by the lesser black-backed gull. Through the combination of GPS data with nutrient analyses of total nitrogen and phosphorus in gull faeces and pellets, the quantity of nutrients loaded to the lake were estimated. GPS data were used to calculate the time that gulls spent in the Lake during several years, but also to correct gull countings for such gulls that departured earlier and are not included in census data. Furthermore, four landfills in the surroundings were the main foraging sites for gulls in the region, and thus, the main source of such external nutrients. Finally, Chapter 5 applied faecal analyses to determine the role of gulls as bioindicators of environemental exposure to heavy metals. There were important variations in heavy metals concentrations in faeces along sites, which would be related to the pollution exposure of the area. Futhermore, the previous nutrient model from Chapter 4 was adapted to quantify the external input of heavy metals into Fuente de Piedra Lake. Long term heavy metal deposition may be of importance to the lake dynamics.