Longer productive season and food base composition change/mismatch
Sea-ice thinning, reduction of the snow cover, earlier melt and later freeze-up of seasonal sea ice will increase the amount of sunlight reaching the marine ecosystem (Alou-Font et al. 2013, Aumack et al. 2014) and likely drive ecological tipping points in which primary producers flourish and outcompete dark-adapted species (Kortsch et al. 2012). In the areas where sea ice still persists, a longer productive season is predicted to benefit the ice-algae, phytoplankton and phytobenthos in both polar regions (Goldsmit et al. 2021). Higher competition will likely lead to widespread shifts from invertebrate- to algal-dominated states on coastal polar seabeds, reducing biodiversity and altering ecosystem functioning. Future extension of the productive season might drive some endemic shallow invertebrate communities into small refugia where sea-ice duration is maintained, impeding genetic connectivity. In many non-coastal areas, increased light will only allow phytoplankton blooms (Montes-Hugo et al. 2009). Thus, some invertebrate species may persist at depths where the light is too low for benthic algae irrespective of sea-ice.
Changes in sea-ice duration and timing of major bloom events are expected to negatively affect food base composition, particularly for secondary producers. Primarily herbivorous zooplankton which transfer energy-rich lipid compounds and essential fatty acids to higher trophic levels, strongly depend on the sea-ice algae blooms (Ershova et al. 2021). During spring, ice algal-derived carbon is essential for the maturation and reproduction of pelagic grazers before phytoplankton is available for their offspring (Leu et al. 2011). An earlier onset of the ice-associated and pelagic blooms is likely to create a mismatch in carbon source availability and grazer occurrence (Ji et al. 2013). Krill and Calanus copepods are a keystone species and a primary food resource for many fish, seabirds, and marine mammals (Cavan et al. 2019, Will et al. 2020). As many species are highly dependent on them, declines in abundance could have cascading effects throughout the food web. Large increases in gelatinous zooplankton biomass could imply an energetic turning point and trophodynamic restructuring.
In regions still covered by seasonal sea ice, more and fresher organic material is exported to the deep sea by the fast-settling ice-associated diatoms and surface-born microbial clades in comparison to slow-settling Phaeocystis aggregates in the ice-free regions (Savidge et al. 1995, Reigstad et al. 2008, Lalande et al. 2019, Fadeev et al. 2021). Weakening of the pelagic-benthic coupling through the decrease in vertical export efficiency and amount of sinking labile organic matter will have repercussions on arctic deep-sea ecosystems. Indeed, benthic biomass and diversity were found to be higher in the seasonally ice-covered areas (Pecuchet et al. 2022) with a shorter but with higher modularity food web. Modelling results indicate that climate-related changes in phytodetrital inputs can lead to important shifts in benthic biomass, community structure, and functional diversity, with loss of various common taxa (Lovvorn et al. 2016). Further warming and reduction in sea-ice coverage will most likely also continue to negatively affect arctic fish species whose food availability largely depends on the coupling between sea-ice cover and the benthic production (e.g. demersal). Most deposit-feeders (and, in turn, their consumers), however, rather depend on the total amount of settling microalgae, and not so much the timing.
Freshwater, nutrient, and sediment inputs. Stratification and vertical transport by buoyant plume
Sea ice and glacier ice melt release large volumes of freshwater. It can have stimulating effects on primary production early in the season by nutrient input and meltwater-induced stratification, which provides favourable conditions for phytoplankton growth and development (Detoni et al. 2015), but prevents the vertical export to the bottom, especially of small cells. Icebergs also leave a trail of trace nutrient enrichment as they gradually melt, enhancing phytoplankton blooms in their path. This phytoplankton bloom fertilisation increase bloom intensity, and increase the potential blue carbon capture, but whether this increases any storage or sequestration by benthic organisms depends on where and when this occurs (Death et al. 2014, Wadham et al. 2019). However, as outlet glaciers continue to retreat, many will lose their connection to the ocean. This decoupling of glaciers (and their meltwater) from the ocean will have important consequences for the timing and spatial distribution of nutrient delivery and the carbon cycle (Meire et al. 2015, Hopwood et al. 2018). Glaciers’ retreat on land will also affect seabirds as currently, submarine glacial plumes stun zooplankton by cold and osmotic shock and raise them up to the surface, thus making available to seabirds, in particular to surface-feeders over benthic-feeders and pelagic pursuit-divers.
Meltwater discharge also transports dissolved and particulate matter from land, leading to increased light attenuation and coastal waters darkening/shading. The total primary production may drop drastically due to intensive glacial melt (Hoffmann et al. 2019). Also, low light availability by glacier melt inputs with high turbidity restricts large phytoplankton productivity (diatoms). Whereas marine particulate organic matter provides rich food to higher trophic consumers, terrestrial POM is a low-quality food resource for marine organisms. Therefore, if glaciers continue to melt because of climate change, an impoverishment of the nutritional value of POM may be predictable impacting the food webs of polar coastal regions. Many zoobenthic species were also found to be sensitive to high sedimentation rates. Thus, if the increasing intensity of glacial processes will continue in the upcoming years, the diversity of the encrusting fauna, ascidians in the shallow sublittoral could dramatically decrease (Krzeminska and Kuklinski 2018).
Ice scouring
Iceberg scouring affects around 30% of world’s coastline and leads to significant mechanical disturbances (Conlan et al. 1998, Barnes and Souster 2011a) that skews population structure of benthic communities (Smale and Barnes 2008), contributes to the sediment resuspension and increment of inorganic particulate matter on the water column (Moon et al. 2015), and limits the potential immobilisation and, ultimately, sequestration of benthic blue carbon (Barnes et al. 2018). The probability of scouring effect is intertwined with sea ice dynamics, since coastal fast-ice losses and low coverage of seasonal sea ice can facilitate the movement of icebergs, thus producing high-frequency ice scour disturbance events (Zwerschke et al. 2022). After the space clearance, mobile and pioneer benthic organisms (via larval recruitment) are the first to recolonise the scour marks and their surroundings. Slow recolonisation and recovery post-disturbance which follows highly complex successional pathways (approximately 10 to 250 years), is a general tendency for polar benthic ecosystems due to slow growth and life cycles (Gutt and Starmans 2001, Zwerschke et al. 2022). The magnitude of iceberg-driven disturbance varies within and, most contrastingly, between polar regions and the peak effect is asynchronous. Abundance, disturbance area, and maximum age of ice scours are comparatively lower in the Arctic (Gutt et al. 1996). Icebergs drifting in the northern hemisphere are significantly smaller than those in the southern, and since the contributory glaciers and ice shelves have retreated more severely in the Arctic, massive disturbances from mega-icebergs should only be expected in the vicinity of the main ice shelves/glaciers in the Canadian high Arctic and Greenland. On the other hand, almost 90% of glacier fronts have retreated along the western Antarctic Peninsula since the early 1950’s (Cook et al. 2005). On ecological time scales, benthos will be increasingly affected in such areas, while ice scour disturbance will drastically drop in Antarctic shallows in the centuries to come; this asynchronous ice disturbance history in polar regions indicate that the Arctic may anticipate future Antarctic.
Human activity
The opening of polar seaways is facilitating opportunities among all types of vessels in areas that were previously covered by MYI, supporting the strong increasing statistical correlations observed between sea ice change and shipping trends in both polar regions (Constable et al. 2022). This will have an impact on ice-associated species by ice breakup, underwater noise, increase the risks of lethal and sublethal collisions of marine megafauna, and risks of invasive species. Loss of marine ice is also related to increasing human activity of other forms, e.g. fisheries and the use of other marine resources. Although the largest more sustainable fisheries are in polar waters, more periodical reassessment of the stocks will be needed particularly considering the imbalance that can be triggered by further changes of the sea ice, and the increase of demand of highly nutritious protein (Constable et al. 2022). Bottom trawling in polar shelves also makes the stored carbon therein more prone to be released back into the carbon cycle, thus blocking the long-term storage (Bax et al. 2021). However, an overarching management framework to protect benthic communities from bottom fisheries in high polar seas is the identification and designation of VMEs (Ashford et al. 2019). Oil, gas, and deep-sea mining exploration and exploitation in Arctic seas is dissonant with the Paris Agreement, and unlike Southern Ocean’s governance of The Antarctic Treaty System (ATS) through the Commission for the Conservation of Antarctic Marine Living Resources (CCAMLR), the northern polar region is more vulnerable to direct human interference.
Cumulative response to environmental stressors
Our ecosystem based risk assessment suggests that future marine ice loss is likely to pave the way for autotroph-dominated polar ecosystems with higher productivity and carbon-capture. The cumulative score of all the stressors shows that marine ice loss will be detrimental to secondary producers, i.e. zooplankton and zoobenthos, and particularly for primarily herbivorous copepods and krill. It is mostly driven by changes in the food base composition and timing after sea ice loss as well as sediment discharge with glacial meltwater that will also affect carbon storage and sequestration as well as regional biodiversity. These effects will likely suppress the effect of colonisation of pioneer species and biodiversity shifts in the newly open coastal zones becoming exposed in the near future. In the long term, there will be regions where ice-algal production will decline with the disappearance of the ice all together leading to new pelagic conditions, while in other areas sea ice will expand over previously ice-shelf covered waters. Such nonlinear responses complicate the prediction of future polar ecosystem dynamics. There is no doubt, however, that the consequences of the large abiotic changes in the polar regions are expected to be severe for ice-obligate species such as polar bears, pinnipeds and seabirds. Currently, they often find refuge at the front of marine terminating glaciers and on ice shelves, or slowly adapt to land habitats. Moreover, climate change-related biodiversity redistribution and range extensions open up new and large-scale opportunities for the recruitment of non-native, invasive species and human activity that have the potential to affect ecosystem functioning and ecosystem services.
The patterns of species loss or change in biomass will likely differ substantially in Arctic and Antarctic regions due to habitat differences related to the nature of the ice-cover, which is floating in the Arctic and land-associated in the Antarctic. Our analysis suggests that the nature of the changes that will occur with marine ice loss in polar regions may differ somewhat in magnitude, with the Arctic having lower cumulative scores and thus being more prone to species loss by the end of this century. Arguably, for all of the environmental stressors assessed here, the Arctic seems to be close to the tipping point of marine ice loss related shifts. In many environmental aspects and to a certain degree, the ongoing changes in the Arctic could be used to foresee possible ecosystem phenologies in Antarctic seas.
The current review shows that there is a great value in a synergistic approach held in compliance with recent state-of-the art assessment and predictions of the environmental stressors. Future reviews and meta-analyses would be facilitated if source studies focussing in species also provide some sort of functional classification of their study subjects that can be used as input for building score matrices or clustering type of responses based (pressure-response).