6. Spatial and temporal monitoring to understand declines in and threats to insular arthropods
Evidence is accumulating globally that terrestrial insect abundance and biomass are in decline across multiple regions, habitats and taxa (Hallmann et al., 2017; Wagner, Grames, Forister, Berenbaum, & Stopak, 2021), although the historical scarcity of highly-standardised long-term arthropod monitoring programs leaves uncertainty of the nature, scope, and taxonomic and geographic variation of the problem. The establishment of such programs has been limited by the time and specialised expertise required to process samples across whole arthropod communities. It is difficult to evaluate which of the potential drivers of long-term declines are most responsible, and thus most important to address, without extensive data with minimal biases (Cardoso & Leather 2019; van Klink et al., 2020, 2022). Thus, HTS combined with methods to measure abundance provide a route to the practical and comparable long-term monitoring of communities globally.
Islands should be particular priorities for monitoring given their relative biodiversity value, high human impact, and utility as harbingers of more general global change (Fernández-Palacios et al. 2021). However, they are underrepresented in global initiatives for biodiversity monitoring and biodiversity indicator frameworks, prompting calls for coordinated surveying and monitoring of island biotas (Borges et al., 2018). Two facets of island biota provide cause for concern in the context of potential decline in abundance and biomass. First, general decreases in arthropod abundance are likely to exacerbate extinction risk at the level of individual species, due to already geographically limited range sizes (Manes et al., 2021; Veron, Mouchet, et al., 2019). Second, island biotas are inherently at risk from species invasion and decreased local abundance of native species could increase this threat (Bellard et al., 2017; Russell & Kueffer, 2019; Borges et al., 2020). Understanding island arthropod decline in the (broadly common) absence of historical data (but see e.g. Colom, Traveset, Carreras, & Stefanescu, 2021; Theng et al., 2020) can, to some extent, be addressed by sampling across gradients for suspected drivers of decline, such as climate (Ferreira et al., 2016) and disturbance (Cardoso, Rigal, Fattorini, Terzopoulou, & Borges, 2013). Given the critical focus on abundance and biomass, multiplex barcoding can be used, and total abundance partitioned to individual species for the identification of more nuanced abundance changes across species. Alternatively, PCR free metagenomic approaches allow estimating the abundance of different arthropod species (Ji et al. 2020), and wocDNA metabarcoding has been demonstrated to reveal relative abundances of species, when coupled with specimen counts (Lim et al. 2021). Although still in early development, CNNs and DL (see above) hold promise for photographing and archiving samples prior to wocDNA metabarcoding, for future abundance estimation (Arribas et al., 2022).
The high throughput and efficiency of HTS barcoding approaches represent a viable long-term solution for monitoring and documenting change within island arthropod communities, and in the context of an iGON, these can be integrated within existing frameworks (e.g. Borges et al., 2018). Suggestions for a coordinated approach to inventory and temporal monitoring, through spatially extensive inventory with a subset of sites subject to temporal sampling (Arribas et al., 2021a), can provide needed baseline data for conservation planning. Range size is frequently used in conservation planning, within which species with small ranges and often declining abundances are given higher priority. Indeed, range restriction and population trends are integral to the International Union for Conservation of Nature (IUCN) criteria to identify and classify species threatened with global extinction. While downstream bottlenecks of the red-listing process are now being addressed by semi-automated systems (Cazalis et al. 2022), upstream the common data deficiency for arthropods within islands limits effective red listing by the IUCN. Strategically designed spatial and temporal HTS barcode sampling networks can provide species records at scales appropriate for both IUCN needs, and the needs of local conservation and management agencies and stakeholders. Such scales go as far as the microhabitat level, as threat might be influenced by it. As an example, classical sampling of Madeiran spiders suggests ground associated species are at greater risk of local extinction than those from canopy microhabitats (Cardoso, Crespo, Silva, Borges, & Boieiro, 2017; Crespo, Silva, EnguĂ­danos, Cardoso, & Arnedo, 2021). This parallels more general inferences for greater climate change impacts for forest floor arthropod species compared to canopy species in Puerto Rico (Lister & Garcia, 2018). More involved implementations of stratigraphically structured sampling and HTS barcoding with abundance data (i.e. multiplex barcoding or wocDNA barcoding with artificial intelligence for image recognition) also has the potential to simultaneously contribute local species records, basic niche information (stratigraphic distribution), and local abundance.