Extraterrestrial nature reserves (ETNRs)

Abstract

If human population growth is not controlled, natural areas must be sacrificed. An alternative is to create more habitat, terraforming Mars. However, this requires establishment of essential, ecosystem services on a planet currently unamenable to Terran species. Shorter term, assembling Terran-type ecosystems within contained environments is conceivable if mutually supportive species complements are determined. Accepting this, an assemblage of organisms that might form an early, forest environment is proposed, with rationale for its selection. A case is made for developing a contained facsimile, old growth forest on Mars, providing an oasis, proffering vital ecosystem functions (a forest bubble). It would serve as an extraterrestrial nature reserve (ETNR), psychological refuge and utilitarian botanic garden, supporting species of value to colonists for secondary metabolites (vitamins, flavours, perfumes, medicines, colours and mood enhancers). The design presented includes organisms that might tolerate local environmental variance and be assembled into a novel, bioregenerative forest ecosystem. This would differ from Earthly forests due to potential impact of local abiotic parameters on ecosystem functions, but it is argued that biotic support for space travel and colonization requires such developments. Consideration of the necessary species complement of an ETNR supports a view that it is not humanity alone that is reaching out to space, it is life, with all its diverse capabilities for colonization and establishment. Humans cannot, and will not, explore space alone because they did not evolve in isolation, being shaped over aeons by other species. Space will be travelled by a mutually supportive system of Terran organisms amongst which humans fit, exchanging metabolites and products of photosynthesis as they have always done.

Introduction

The United Nations (UN) urges protection and restoration of global ecosystems (UN, 2015). However, the human population, predicted to exceed nine billion in 2050 (UN, 2017), requires feeding and c. 37% of the world's land area is used for agriculture (pasture and arable) (FAO, 2017). Some predictions indicate this will increase by 2050 (Öborn et al., 2011); in less than a century, globally significant wilderness may not exist (Watson et al., 2016).

Various estimates put Earth's carrying capacity for Homo sapiens at or below eight billion (UN, 2012), so choices are necessary. If human population growth is not controlled, semi-natural areas must be sacrificed to food production and urbanization, engendering ecosystem collapse and environmentally forced population reduction. An alternative is to create more living space, e.g., habitats orbiting Earth or on the Lunar or Martian surface. The concept of extraterrestrial (ET) nature reserves (ETNRs) arises, as envisaged in Douglas Trumbull's 1972 film ‘Silent Running’.

Such work is portended by the European Space Agency's (ESA's) MELiSSA project, envisaging closed, bioregenerative life-support for human space missions (e.g., Lasseur et al., 2010), and the contained ecosystems of Biosphere 2 (cf., Nelson, 2018). Plants will be critical to human survival outside Earth (Poulet et al., 2016), as life-support systems, providing food, oxygen (O₂) and water purification (Wolff et al., 2014). Therefore, space exploration requires establishing ecosystem services in ET environments, for pragmatic, selfish and altruistic reasons.

Terraforming Mars is considered (e.g., Sagan, 1973, McKay, 1982, McKay et al., 1991, Birch, 1992, McKay and Marinova, 2001, Beech, 2009, Jakosky and Edwards, 2018, Pazar, 2018). Accepting long time scales, this offers security for Earth life threatened by astronomical or anthropogenic catastrophe. Eventually, Terran-type ecosystems (TTEs) might be assembled on a modified planet's surface, though contained communities may be achievable sooner and as a necessary step.

Mars is therefore proposed as a location for a contained TTE (CTTE). Assuming containment can provide tolerable conditions, Mars offers gravity, atmosphere, sufficient sunlight for photosynthesis (Table 1) and water (Rummel et al., 2014), while proximity to Earth allows management.

Table 1. Relative characteristics of Earth and Mars

¹Perihelion: closest point to sun during an elliptical orbit, therefore maximum incident UV flux.

A Martian CTTE would self-justify, offering refuge, retreat and ecosystem services for astronauts concerned with exploration, resource harvesting or colonization. It would provide wonder, inspiration, purpose and a psychological stepping-stone for more ambitious projects. Humanity's exploration of space will need a network of such oases, supporting terraforming resources and offering biotic refreshment. Forest ecosystems would be apposite.

Forest environments have health benefits (stress reduction, recovery from fatigue, rehabilitation) (Karjalainen et al., 2009). Ancient forests have emotional, spiritual and cultural significance (e.g., Lowman and Sinu, 2017). Threatened by human impact (Frank et al., 2009; Laurance, 2015), they are archetypes, visions of arboreal majesty and biotic complexity (e.g., Wirth et al., 2009). Such TTEs would offer relief from ET sterility.

Facsimile old growth forest could be established over a century on Earth (Smith, 2018), but present-day Mars' surface is hostile to Earth-adapted life, with high radiation levels (Nixon et al., 2013), thin atmosphere and other stressors (e.g., Table 1). TTEs would need shielding.

This account assumes a semi-autonomous, contained environment could be created on Mars, large, strong and shielded enough to support a forest, hold positive atmospheric pressure, protect it from meteorites (e.g., Daubar et al., 2019) and exclude harmful radiation (a ‘forest bubble’, McKay 2022 personal communication 20th August) (Fig. 1).

Fig. 1. Conceptual drawing of 20 ha footprint ‘forest bubble’ (location may constrain shape).

Challenges of constructing, large Martian ‘worldhouses’ have been discussed (Taylor, 1992; 1998). Acknowledging small containments' limitations (Taylor, 1998), and flaws in minimum habitat areas (e.g., van der Hoek et al., 2015), a hemispherical, environmentally controlled, enclosure of c. 0.25 km radius is envisaged (20 ha footprint). Many semi-natural Earth woods are smaller (e.g., Peterken, 1993) and Biosphere 2's designed rainforest occupies c. 0.2 ha (Nelson, 2018).

Outlining of constraints (cf. Table 1) leads to justification of a concomitant forest design with novel species complement (cf. Table 2). Discussion of ethics follows.

Table 2. Potential integrants for contained Martian TTE

Non-human vertebrates omitted as ability to engage in natural behaviours is not ensured.

Environmental constraints

Ionizing particle radiation (IR)

Unlike Earth, Mars lacks a significant global magnetic field to exclude incoming charged particle radiation (e.g., Atri, 2016), its surface subject to solar energetic particles and galactic cosmic radiation that damage living tissues (e.g., Nelson, 2013, 2016).

Artificially generated magnetic (e.g., Townsend, 2005; Battiston et al., 2012; Bamford et al., 2014; Ambroglini et al., 2016), or electrostatic (Tripathi et al., 2008; Joshi et al., 2013) fields, improved passive shielding, local crustal magnetic fields (e.g., Alves and Baptista, 2004) and/or thick layers of Mars regolith (e.g., Röstel et al., 2020) might provide solutions.

Dohm et al. (2011) consider cavern refuges. One existing skylight accesses a void at least 37 m deep (Cushing, 2012). Such places might protect CTTEs from IR. Prisms could refract sunlight into underground chambers (cf. Luxfer prisms, Neumann, 1995), IR bypassing them.

Sunlight (including UV)

Martian day length resembles Earth days, and CTTEs can exploit local sunlight (Table 1). Subterranean situations might use mirror/fibre optic delivery, augmented by light emitting diodes (Nakamura et al., 2013). Electric light requires maintenance but sunlight collection is vulnerable to dust storms (e.g., Fernández, 1998; Martínez et al., 2017). Self-cleaning, light-harvesting surfaces are needed. Superomniphobic materials (Sun and Böhringer, 2019) augmented by electrodynamic technology (Mazumder et al., 2016) or ‘plasma brooms’ (Ticoş et al., 2017) might provide solutions but surface micro-/nano-coatings are short-lived (Sun and Böhringer, 2019). Vertical light-harvesting surfaces might shed dust, while light-transmitting plants (e.g., Duckett and Ligrone, 2006) might inspire materials.

Besides diffuse illumination, forest understoreys experience sunflecks (short periods of direct irradiance when sun penetrates canopies) (e.g., Pallardy, 2008). Their nature depends on canopy physiognomy, solar declination and solar time (Chazdon and Pearcy, 1991). Sunflecks are energy sources, significant to small organisms (including seedlings) and potential stressors (e.g., Leakey et al., 2004). Photosynthesis during them may provide 30–60% of daily C gain (Chazdon, 1988), plants' responses involving UVB photoreceptors (Moriconi et al., 2018).

Static, unidirectional lighting in windless CTTEs would not engender sunflecks' dynamic chiaroscuros (cf. Way and Pearcy, 2012), so lighting manipulation, exposure to natural solar ecliptic and/or leaf-animating wind is needed.

Mars' harsh surface UV flux (Table 1) is sterilizing due to thin atmosphere and lack of significant ozone (Cockell et al., 2000; Kminek et al., 2010). UV has positive (Juzeniene and Moan, 2012) and negative (e.g., Lee et al., 2013, Rettberg et al., 2004) effects on organisms. Fortunately glass/plastic combinations can exclude harmful wavelengths whilst transmitting beneficial UV and visible light (e.g. (Duarte et al., 2009; Tuchinda et al., 2006), so flux in CTTEs can be controlled.

Some UV is necessary for vitamin D synthesis and other mechanisms in animals (e.g., Juzeniene and Moan, 2012; Wilson et al., 2012; Baines et al., 2016) and necessary irradiances may be determined (Cockell and Andrady, 1999). Many, non-human animals have vision in the UV spectrum (e.g., Bennett and Cuthill, 1994, Cronin and Bok, 2016) including honeybees (Apis mellifera) (Reverté et al., 2016). Some pollinators use UVA for navigation (Cockell and Andrady, 1999). Human wellbeing and ecosystem function will therefore require modulation, not total exclusion, of Mars' UV flux.

Magnetic fields (MFs)

Life evolved within Earth's geomagnetic field (GMF) (Maffei, 2014). Magnetic guidance mechanisms exist in some microorganisms, and many animals (Frankel, 1984) and MF changes might impact plant growth and development (Wolff et al., 2014). So, some behaviours might be compromised on Mars due to the lack of a GMF. This may affect CTTEs.

Climate, temperature and pressure

Mars' mean surface temperature is −63°C. While periods above freezing occur, surface atmospheric pressure is so low (Table 1), any water ice that melts usually sublimes to vapour (Lewis, 2003). The limit for higher plant tissue growth may be 5°C, little occurring at 6–7°C (Körner, 2008). ‘Biologic zero’ relates to soil temperatures when microorganisms and or plants become inactive, sometimes considered 5°C (Rabenhorst, 2005). CTTEs must therefore maintain elevated internal air temperature and pressure.

Mars' equator might offer a thermal advantage over other Arean locations for a CTTE. However, other factors apply. Haberle et al. (2001) discussed regions where ground temperature and surface pressure can be favourable for the existence of liquid water, including the Hellas basin. The base of this 9 km deep impact crater (Ali and Shieh, 2014) potentially experiences 12.4 mbar surface pressure during the northern summer (Haberle et al., 2001). Such locations might facilitate contained pressure differentials.

CTTEs will require heating and dispersal of excess heat. Inevitably, heat will be lost to the external environment over time. No insulation is perfect, seals must allow ingress/egress, and dust storms (cf. Fernández, 1998) will reduce solar benefits but ‘intelligent’ computer-controlled structures (Taylor, 1998) might maintain suitable environments. Biosphere 2's complex control systems indicate engineering challenges and power needs (Nelson, 2018).

Solar power has potential (e.g., Delgado-Bonal et al., 2016; Vincente-Retorcillo et al., 2018) but due to dust storms (Fernández, 1998), hybrid power generation, with rechargeable batteries and/or nuclear thermoelectric technology (e.g., LaMonica, 2012; NASA, 2019a), may be needed. Geothermal options might exist (Morgan, 2009; Sori and Bramson, 2019).

Seasons

Biomes change seasonally, so CTTEs require seasons. Temporality determines critical developmental stages, individual physiologies and interspecific relationships, while timing of abiotic events influences global nutrient fluxes (Forrest and Miller-Rushing, 2010). Photoperiod and winter chilling are involved in temperate plants' phenology (Richardson et al., 2013). Development of many insects is seasonally synchronized, enabling tolerance of adversity (Danks, 2007). Phenological cycles are fundamental to ecosystem function (e.g., Stucky et al., 2018) and climate changes can desynchronize critical interactions (Thackeray et al., 2016). Seasons also imbue characteristics critical to psychological restoration, e.g., autumn colour, winter silence, spring flowers and summer leafiness.

Mars has four seasons, approximately twice duration of Earth's (e.g., ESA, 2019). These vary in length due to its elliptical orbit, spring in the northern hemisphere (autumn in the southern) being the longest (NASA, 2019b). Whether Earth organisms can adapt to Mars seasons, even in containment is unknown.

Conditions on Earth have not selected for tolerance of seasons of such asymmetry and length (Taylor, 1998). So, CTTEs need artificially controlled seasons (diverging from semi-autonomy) or to be assembled from species tolerant of seasonal aberrance, the latter if relying heavily on passive sunlight delivery.

Lunar cycle

Most Earth organisms have circadian clocks, endogenous, molecular timing systems, allowing anticipation of Earth's 24-h light-dark cycle and maintenance of behavioural cycles (Bollinger and Schibler, 2014). Similarity of day length of Mars and Earth (Table 1) suggests adaptation may occur in CTTEs, processes according with the same temporal cues (zeitgebers).

However, the lunar cycle is also relevant. Earth's moon, Luna, is a zeitgeber for many ecological processes, some pertaining to monthly or half monthly cycles, others to shorter periods (e.g., Raible et al., 2017). Examples include animals (e.g., Raible et al., 2017; Sinclair, 1977; Dixon et al., 2006) and plants (e.g., Barlow, 2015; Ben-Attia et al., 2016). Lunisolar tidal force may also influence plant growth (Barlow and Fisahn, 2012).

Mars' two moons, Phobos and Deimos, have maximum radii of c. 13.5 km (NASA, 2019c) and c. 7.5 km (NASA, 2019d) respectively, small compared to Luna's radius of c. 1737 km (cf. Williams, 2021c). Rao (2015) speculates there are parts of Mars from which the moons are never visible due to orbital proximity and Mars' curvature.

Evidence for organisms' responses to Earth's lunar cycle varies from well substantiated to speculation but inevitably Terran species, translocated to Mars, would experience a different lunar influence, the effects hard to predict.

Soil

TTEs require suitable organic substrate. Freighting constraints require local development. Mars has a basaltic upper crust, with variable abundances of other materials (Ehlmann and Edwards, 2014). Basalt-derived soils with volcanic ash are good agricultural soils (e.g., Olowolafe, 2002). Crushed basalt can increase soil pH, while its dissolution releases beneficial nutrients, including phosphorus (P) (Shamshuddin and Che Fauziah, 2010).

Martian substrate probably contains nutrients to sustain plant growth (e.g., Jordan, 2015). ‘Mars regolith simulant’, supports angiosperms (Wamelink et al., 2014) and, with added organic matter, earthworms (Wamelink et al., 2022).

Plants require 16 essential elements, C, hydrogen, O_2, nitrogen (N), P, potassium, calcium, magnesium, sulphur, iron, zinc, manganese, copper, boron, molybdenum and chlorine (Uchida, 2000). These are all reported from Mars or Mars meteorites (Jordan, 2015). Cobalt and nickel (e.g., Brown et al., 1987, López and Magnitskiy, 2011) are also relevant, being involved in biological N-fixation. Nickel has been detected in Martian substrate (Gellert et al., 2006; Yen et al., 2006) and cobalt in putative Martian meteorites (Lodders, 1998).

Plant growth requires reactive N, predominantly nitrate (NO₃⁻); 40–60 ppm NO₃⁻ advised for vegetable crops (Cantisano, 2000). Evidence suggests up to c. 1100 ppm of NO₃⁻ in Mars' sedimentary deposits (Stern et al., 2015).

Phosphates are essential for Earth life (Tirsch and Airo, 2014). Evidence indicates Mars is 5–10 times more phosphate rich than Earth, mineralogical studies (Adcock et al., 2013) suggesting biological accessibility.

So Martian regolith may contain necessary nutrients for a CTTE, while low organic C, water holding capacity and cation accessibility might be improved by microbiological weathering (Cockell, 2011).

Cyanobacteria are proposed for in situ resource processing (Verseux et al., 2016). Photosynthetic, N-fixing Nostoc, will grow on Martian regolith simulant (Arai et al., 2008) and early successional cyanobacterial communities improve soil moisture retention (Danin et al. (1998).

Toxicity

Martian substrate contains perchlorates (ClO₄⁻) at concentrations much higher than typically found on Earth (Davila et al., 2013). These affect thyroid function (e.g., Srinivasan and Viraraghavan, 2009) and some plant growth experiments with regolith simulant assume remediation (Gibbens, 2017). Other oxidants present at Mars' surface include hydrogen peroxide and iron oxides (e.g., Lasne et al., 2016). Mars has over twice as much iron in its outer layers as Earth (Peplow, 2004) and, though an essential plant nutrient, it can accumulate to become toxic (Connolly and Guerinot, 2002). In combination, iron oxides, hydrogen peroxide, perchlorates and Mars' UV flux, are highly deleterious to living cells (Wadsworth and Cockell, 2017). Extreme salinity is another potential stressor (e.g., Ramírez et al., 2017).

Tolerance of such parameters will be desirable in ETNRs, though CTTEs allow remediation. Many perchlorate-reducing bacteria exist (e.g., Coates and Achenbach, 2004) and bacterial enzymes have potential to detoxify hydrogen peroxide (Nóbrega and Pauleta, 2019). Perchlorate is also highly soluble in water (Davila et al., 2013), allowing biotic and/or abiotic decontamination.

Water

Present Mars is a cold desert (McKay, 2010). However, the freshwater content of Mars' permanent north polar ice cap is c. 100 times that of the Laurentian Great Lakes (Rummel et al., 2014). Liquid water may even exist beneath the southern polar ice (Orosei et al., 2018) and ‘recurring slope lineae’ may be active surface brine flows (e.g., Ojha et al., 2015).

Evidence indicates sufficient water reserves for CTTEs (toxin removal possible). Conifer needles collect cloud drops (Unsworth and Wilshaw, 1989) suggesting delivery options. Atmospheric temperature gradients with dew points (Lu and Ho, 2019) and microstalactite ceiling materials (condensation foci) merit exploration for artificial rain.

Di-oxygen (O₂) and di-nitrogen (N₂)

Mars' atmosphere is CO₂ rich with little O₂ or N₂ compared to Earth (cf. Table 1). O₂ is essential for aerobic TTEs, while relatively inert N₂ is useful in bulking atmospheric pressure. Reactive N is present in proteins and nucleic acids, so sufficient atmospheric N₂ must be available for biological N-fixation (McKay and Marinova, 2001) and cycling. CTTEs on Mars therefore require increased atmospheric O₂ and N₂.

Fortunately, Mars' resources include oxygen bound in perchlorate, carbonate (Bridges et al., 2019) and nitrate (the latter providing fixed N) that might be harvested. Davila et al. (2013) propose enzymic release of O₂ from perchlorate and N₂ might be liberated by bacterial denitrification (e.g., Hart et al., 2000). Technologies are also developing for mining Martian atmosphere (Finn et al., 1996; Sridhar et al., 2000) and CTTEs do not necessitate duplication of Earth's mean atmospheric pressure and composition; atmospheric pressure varies altitudinally and species' tolerances vary.

Klingler et al. (1989) showed some bacteria capable of N-fixation from partial pressures of N₂ down to 5 mbar (25 times current Mars levels). Some plants can utilize O₂ levels well below, and tolerate CO₂ levels above, current Earth values. Photosynthesis can be enhanced at O₂ concentrations below ambient (e.g., Downes and Hesketh, 1967), due to reduction in photorespiration (e.g., Hagemann et al., 2016). Some show higher photosynthetic rates under elevated CO₂ (e.g., Ainsworth and Rogers, 2007), benefitting from the ‘CO₂ fertilization effect’ (Zheng et al., 2018).

Green plant photosynthesis might generate elevated O₂ levels in a CTTE. Fogg (1995) considered root respiration demand could limit plant growth on Mars until atmospheric O₂ was raised to 20–100 mbar (>3000 times current levels) but levels in containment could be primed.

Modification of contained Martian atmosphere is therefore conceivable and may be less demanding than anticipated. As initial O₂ levels rise, and biological N-cycle initiates, photosynthetic eukaryotes may mediate further atmospheric modification, ultimately achieving conditions tolerable by invertebrates.

Gravity

Terran life evolved within Earth's gravitational field (1 g) and CTTE success depends on development and function under Mars' lower gravity (Table 1).

Light and gravity modulate plant development (Vandenbrink et al., 2014). Experiments indicate 0.3 g (< Mars) sufficient to trigger gravitropic responses, but that meristematic competence can be lost under lunar-like (0.17 g) gravity (Manzano et al., 2018). Nevertheless, plants will grow and photosynthesize even in microgravity (e.g., Monje et al., 2005; Wolverton and Kiss, 2009). Though some biochemical (Cowles et al., 1988) and anatomical (Hoson et al., 2003) changes may occur, results conflict (Stanković, 2001).

From such evidence, it is conceivable that some plants (and fungi, cf. Kern, 1999) will tolerate Mars' gravity. However, forest function is also influenced. Leaf and propagule fall, leaping, flight, deadwood collapse, raindrop impact and drainage of water contribute dynamism. On Mars, things weigh 38% their Earth weight, potentially benefitting trees etiolated by low light, or lacking wind-induced reaction wood (Groover, 2016) (cf. Biosphere 2, Nelson, 2018).

Many organisms reproduce and disperse by airborne propagules. If these develop normally, greater dispersal capacity under lower gravity may not be problematic, provided they can disperse. This may require vectoring to avoid intergenerational competition, so, CTTEs need wind. However, lower gravity means lighter propagules and thermal gradients might be exploited to generate air currents.

Some plants and fungi exploit ‘splash cups’ from which propagules are dispersed by raindrop impact (Brodie, 1951). Such structures evolved in response to rain falling under 1 g, their functionality on Mars unknown but testable.

Potential to leap, climb or fly on Mars with less effort will influence TTE function and some animals may benefit from positive energy budgets. Some insects learn to fly in microgravity (Nelson and Peterson, 1982; Vandenberg et al., 1985), so potential exists. Capacity of most animals to adapt is unknown but 0.38 g is not zero g.

Forest design

Species complement dictates forest appearance, physiognomy and functioning. Limited by the abiotic environment sustained, this will perforce include an unusual assemblage of species (integrants), tolerant of prevailing conditions, comprising a novel ecosystem. On Earth, species niches are limited by competition and availability. Local environmental parameters in CTTEs will lead to new fitnesses, species occupying different roles where niche requirements are provided.

It would be counterproductive to plan replication of a specific forest biome. Earth's forests owe their assemblages to environmental and evolutionary pressures that will differ to those in Martian CTTEs. No single forest food web has been fully mapped, canopies themselves potentially comprising over 100 000 trophic links (Nakamura et al., 2017), challenging duplication. Lack of GMF, reduced sunlight, aberrant seasons, variant lunar cycle, reduced gravity and pedological peculiarities will engender novel ecosystem function.

Significant seasonal differences make it unlikely the same palettes of synchronized mutualisms, which define Earth's forests, could be established on Mars, though dormancy traits might prove useful (e.g., Taylor, 1998) and potentially some species would adapt. If so, a forest might be established but it would only consist of those organisms than can adapt. Design must therefore include planned redundancy, allowing for unknowns.

Mitsch and Jørgensen (2003) indicate that if enough organisms and propagules are delivered, local conditions will select out the best-adapted assemblage. In Odum's (1983) terminology, ecosystems self-organize from the available (Smith, 2018 discusses) and designers must allow ‘self-organization’ since active assembly of complex species networks would demand unattained prescience.

ETNR designers should consider species as ecological cogs that might be assembled into functional ecosystems. Replication of Earth forests is currently unfeasible but development of new ecosystems, functioning in unexpected ways, is conceivable. Mars' forests would not resemble or function exactly like Earth's forests but could still deliver wonder; autumn at 0.38 g offering dreamlike leaf fall.

Early ETNRs may be relatively oligospecific, freighting considerations, even for seeds, restricting initial complement. Selection must acknowledge survivability and ecosystem function, while expedience requires instrumental value, species producing wood, fibre and important secondary metabolites (vitamins, flavours, perfumes, colours, mood enhancers). Species diversity must be built incrementally, over time, by assisted colonization, monitoring, adjustment and replenishment.

The proposed forest is intended as an expansion of Earth's ecosystem, a utilitarian botanic garden and restorative sanctuary. Arboreal communities that can be ‘entered’, offering a sense of ‘escape’ support the latter, several tree taxa proposed as canopy. The design incorporates some organisms considered problematic on Earth but exhibiting potential ET adaptability or terraforming value. Heretical recombination is incited, selecting species from various forest biomes to exploit useful traits, fulfilling essential roles. On Earth, ecosystems rely on co-existence for services including N-fixation and mineral breakdown but in a CTTE all needs must be met either artificially or by integrants. Experimentation will be necessary, knowledge accruing, anticipating subsequent ecosystem modification.

Varietal forests adapted to extreme ambient parameters offer templates of resilience. High-altitude forests tolerate low atmospheric pressures and temperatures. Early successional forests exploit soils low in nutrients and sometimes high in heavy metals. Attempted duplication of a specific high-altitude ecosystem has merit in that species complement can be determined, co-evolution satisfied and incompatibilities minimized. However, this does not allow selective assembly, failing to acknowledge precedent diasporic Earth forest (Smith, 2018) and unique Martian exigencies.

Species complement

Mars' forest complement is designed with reference to local constraints, instrumental value and survivability (Fig. 2). The assemblage is broadly justified below and detailed in Table 2.

Fig. 2. Selection factors for Mars’ forest species complement based on local constraints, instrumental value and survivability.

Trees of high altitudes provide the foundation. Earth has elevational limits beyond which trees cannot grow (Körner, 2012). About 100 species worldwide form trees at the climatic treeline, reducing to c. 20 (Pinaceae and Betulaceae) at the arctic equivalent, (Körner, 2012). Miehe et al. (2003, 2007) discuss high-altitude Tibetan forests, including the sacred Reting Forest, where Juniperus tibetica (Cupressaceae) grows in an open shrub layer of J. pingii var. wilsonii, Potentilla fruticosa, Lonicera spp. and Caragana spp. Two junipers, Juniperus convallium (3500–4570 m a.s.l) and J. tibetica (4100–4850 m a.s.l.), are widespread in southern Tibet (Miehe et al., 2003). They form small forests, sometimes amongst Picea and in open stands of Cupressus torulosa var. gigantea (Farjon, 2010). J. tibetica forms the highest northern hemisphere treeline (e.g., Miehe et al., 2007), some stands being pilgrimage sites (Miehe et al., 2003) or religious landmarks (Miehe et al., 2008).

Climatic limits of cold timberlines relate to isotherms of 10°C for warmest month's mean temperature (Daubenmire, 1954). Similar mean growing-season temperatures of c. 6.7°C, at climatic, high-elevation treelines worldwide indicate temperature control (Hoch and Körner, 2005). Rainfall data suggest the drought limit of juniper trees correlates with annual precipitation of 200–250 mm and Miehe et al. (2003) argue the high mountain deserts of southern Tibet could be reforested without irrigation, even lacking a high groundwater table.

Mars' soils will shape the contained forest, so, early successional colonizers, common to high altitude environments, e.g., pines (Pinus spp.) and birches (Betula spp.) are included. Substrate pH constrains species choice. Contained soils may be modified but minimal intervention is expedient. Phoenix Mars Lander measured an alkaline pH of 7.7 ± 0.5 for substrate (Hecht et al., 2009) whereas Opportunity rover found evidence of slightly acidic to circum-neutral pH (Arvidson et al., 2014), so soils of varying pH could be developed allowing species diversity.

Juniperus tibetica grows on rocky soils derived from siliceous and calcareous materials, experiencing extremes of solar radiation and frost (Farjon, 2010) but CTTEs might also incorporate canopies from high altitude heaths, i.e., Erica arborea or E. trimera (e.g., Beentje, 2006). Ericaceous plants' tolerance of acidic, metalliferous soils (Bradley et al., 1982) may be useful. N-fixing plants will also be essential. The N-fixing shrub Caragana versicolor, tolerant of Himalayan cold aridity (Kumar et al., 2016), is proposed for clearings and margins.

Creation of ETNRs as biotic insurance against planetary disaster requires human redundancy, currently unachievable. A biocentric ideal would be a self-monitoring, self-repairing containment able to support a living ecosystem even following human abandonment. However, this would represent a pyrrhic legacy if the containment survived an event, its ecosystem extirpated. CTTEs should include some species resilient to sub-catastrophic containment failure, providing opportunity for re-establishment in the manner of Terran ecosystems post adversity.

Plants categorized in USDA Hardiness Zones 1a to 7b (USDA, 2012) offer potential for survival of unplanned temperature drops, these covering species able to survive average, annual winter temperatures of −51.1 to −12.2°C. Choice of range is subjective but −12.2°C must offer ‘some’ insurance against partial containment failure (increasing levels below that). This provides significant opportunity for the development of boreal forest types, dominated by conifers but with associated broad-leaved plants.

Long-lived seeds, able to withstand deep cold, O₂ starvation, fire and/or decompression and to subsequently germinate, would contribute to recovery from temporary containment failure. Ability to resprout from hypogeal structures (e.g., James, 1984) would also be useful. A mixture of species, resilient to periods of reduced atmospheric pressure, darkness, desiccation or other extremes, facilitates biotic insurance, i.e., however severe life support failure is, something will survive.

Individual longevity is desirable in certain integrants. Initiating ancient tree development would provide visionary opportunity. Intangible though such issues are, ancient trees may ultimately engender historical connection between CTTE designers, facilitators and future visitors, potential for legacy, a harnesser of political and financial support.

Junipers are long-lived. J. communis can live c. 200 years (Ward, 1982), J. pingii var. wilsonii >300 years (Liang et al., 2012). Dendrochronological analysis provides ages >2000 years for J. occidentalis (Rocky Mountain Tree-Ring Research, undated) and >2230 and 1265 years for J. przewalskii and J. tibetica respectively (Liu et al., 2019). By vegetative reproduction, some trees persist as clonal organisms for centuries. A Populus tremuloides clone in North America, extending over 43.6 hectares (DeWoody et al., 2008), is potentially of great age (Rogers and McAvoy, 2018).

Fecundity is important in future proofing. Rapidly reproducing plant species can repopulate in the event of non-critical losses. Less fecund species may be vulnerable to vicissitudes of population decline, entering extinction debt in the event of partial catastrophe. Rapidly propagating species (e.g., Betula spp.) should therefore be included.

ET ecosystem design philosophy is nascent. Just as some plants develop winter hardiness, surviving freezing (e.g., Vitasse et al., 2014), unexpected phenotypic adaptations to other stressors might be expressed in ETNRs. Such unpredictable phenomena make the difference between success and failure, so adaptability is important and ‘invasiveness’ on Earth may be a valuable trait in ET environments.

To facilitate ecosystem construction, some understorey species are selected with consideration to natural occurrence alongside canopy integrants. Necessity to avoid allelopathic incompatibility motivates this, though, notably, some invertebrates require multiple plant species for life cycle completion. Such niche requirements could be difficult to fulfil but experiment may reveal unknown tolerances and suitabilities, while selection of some naturally co-occurring plant species may assist.

Use of extremophiles from Mars-like, high-altitude deserts on Earth is not emphasized. Though replication of cold desert ecosystems might be easier to achieve on Mars than forests and would offer biotic insurance, it is arguable whether similar psychological and inspirational benefits would ensue. Recreation of non-forest ecosystems on lifeless planets is laudable, a responsibility inherent to humanity's burgeoning space-faring ability, but the creation of wonder and incitement to pilgrimage require a physically imposing plant community; this demands trees.

Acknowledging the issues above, Table 2 presents a forest-like assemblage incorporating biodiversity, resilience and functionality. Human psychological restoration requires interest and distraction, variation in colour, hue and form, species variety and opportunity for haptic exploration, so these are integrated. Ecosystem services that integrants could supply, relating to human life support and life quality are listed.

ET forest (ETF) survival requires future proofing against system failures. Political support dwindles less if such failure is only partial. Trial and error will shape the species palette, but that presented plans for unmitigated success and limited catastrophe. Since all photosynthetic plants provide O₂ and absorb CO₂, this major ecosystem service is common throughout. Designers must note that dioecy sometimes demands two genders.

Decomposition

Decomposition must occur in CTTEs. Without breakdown of dead biological material, nutrients become sequestrated, atmospheric CO₂ depleted and ecosystem cycling ceases (e.g., Chapin et al., 2002). Table 2, therefore includes decomposers.

Organic litter fall is crucial in biogeochemical cycling (e.g., Krishna and Mohan, 2017). Arthropod communities mediate its degradation (e.g., Berg et al., 1998). Litter and deadwood are also C sources for forest soil microbes (e.g., Lladó et al., 2017). Kjøller and Struwe (1992) discuss microfungi's key role in degrading diverse complex molecules. Bacteria are also important, especially in the soil N-cycle (e.g., Takai, 2019).

CTTE designers must provide suitable biota able to carry out decomposition processes (disassembly, fragmentation, trituration, digestion, solution and N-cycle steps from ammonification to denitrification). The risk of N-cycle dysfunction requires monitoring and proactive correction technology may be necessary, diverging from ideals of human redundancy. With Earth's functional biogeochemical cycles, creation of a forest ecosystem is readily conceivable but incorporation of such support into CTTEs poses challenges.

Ethics

Whether an ETNR is ecologically effective depends on scale. Optimal size of Earth nature reserves is debated (e.g., Diamond and May, 1981; Higgs, 1981) and ETNRs demand similar scrutiny. Size of CTTEs may be limited by engineering constraints, but ‘minimum-area requirements’ and ‘minimum viable population’ sizes will be relevant, as per the SLOSS debate (i.e., whether a single large reserve will conserve more species than several small, e.g., Tjørve, 2010).

To provide biotic insurance ETNRs require assemblages from all Kingdoms of living things including animals. This raises ethical issues, ecosystem dysfunction potentially leading to suffering through system failures, unsuitable design or intolerances.

In facilitating psychological recovery of space workers, animals would be beneficial. Woodland without birdsong or butterflies is a poor TTE. Such lack may exacerbate homesickness. However, when creating habitats on Earth, many animals can elect to inhabit or leave by their own volition. This choice is denied in a CTTE. Introduction of species unable to engage in natural behaviours should be avoided, consideration of the ‘five freedoms’ (Farm Animal Welfare Council, 1993; Webster, 2016) will be necessary and human management may be essential.

Consequences of contact between biospheres is also a consideration, as reflected in the UN's Outer Space Treaty of 1967 (United Nations, 1967) and the International Council for Science's Committee on Space Research (COSPAR) Planetary Protection Policy (COSPAR, 2002 (amended 2011), Rummel et al., 2014). Creation of contained biospheres reduces risk of ecosystem contamination but, since no containment is perfect, protection of Mars' ‘Special Regions’ (Rummel et al., 2014) influences location choice.

Conclusions

Creating a contained ETF is more complex than establishing woodland plants in a protected environment. Even gardens rely on natural nutrient cycling, soil disturbance and irrigation. CTTEs should be almost self-sustaining with propagule dispersal vectors, internal weather and replication of the myriad changes that forests exploit. The designers' task is daunting but, if survival of Earth life is to be ensured, challenges must be overcome.

Humanity does not know if life exists elsewhere in the universe. Mars may support native organisms, but even if it does, Earthly life may be endemic to Earth. Perhaps, life only exists on Earth. In either of which cases, Homo sapiens as the local sentient, technologically empowered species, has responsibility.

From a biocentric perspective, world leaders should be concerned about the future of life in the Universe and humanity's role in its protection and promulgation. On a planet of limited habitability, this is a significant duty. The survival of life, in any form, is the ultimate biocentric priority.

The global ecosystem changes and its conservation requires imagination. Evidence indicates that a contained ETF could be established on Mars. A partially human-redundant protection system would be needed but, like the juniper forests of Tibet, the forest's existence would incite pilgrimage, emboldening efforts for space travel. It is easier enter a desert, knowing it contains an oasis.

This paper does not consider economics. Sending humans into environments without ecosystem services adds to space travel's cost (e.g., Glenn Smith and Spudis, 2015). ESA's MELiSSA project indicates that humans should not think of travelling alone but with a supporting biosphere. We travel through space every moment, sustained by Earth's biodiversity. Our planet carries a self-supporting, bioregenerative ecosystem that, accepting Lovelock's (1979) Gaia hypothesis, modifies and sustains its own life supporting qualities. So, spacecraft should be reimagined as symbiotic communities.

The sailing ships of past explorers were not sterile. They carried animals for food and as living cargo (e.g., Blancou and Parsonson, 2007), for companionship (Mäenpää, 2016) and as pests (e.g., Atkinson, 1973). Sometimes, animals were released or escaped onto foreign shores where some thrived or became problematic (e.g., Campbell and Donlan, 2005; Harper and Bunbury, 2015), examples of accidental, incidental and deliberate dispersal of Terran species. Goats were once purposely liberated on remote islands by mariners, as a self-renewing food resource (Dunbar, 1984). Such attitudes may become necessary during space exploration, creating oases on barren but habitable planets. Spacecraft will carry multiple species complements, contributing life support for long journeys and on arrival at lifeless destinations.

ETNR design will be inspired by human dependency on ecosystem services, even in purely utilitarian fashion, because, despite technology, that dependence cannot be shed. We need plants as chemical factories, producing secondary metabolites with greater ease and more autonomy than industry. Ultimately, humans must take Earth's ecosystem with them, acting as the medium through which it colonizes the planetary archipelago of space. We will not travel alone because we did not evolve in isolation. Homo sapiens was shaped, over aeons, by other species and will travel with a mutually supportive system of Terran organisms amongst which we fit, exchanging metabolites as we have evolved to do. It is not humanity that is reaching out from Earth, it is life, with all its diverse capabilities for colonization, humanity the ineluctable vector.