COLLOQUIUM

Synthetic threads through the web of life

PAPER
Mary E. Powera,1
a
Department of Integrative Biology, University of California, Berkeley, CA 94720

Edited by Daniel F. Voytas, University of Minnesota, Saint Paul, MN, and approved March 25, 2021 (received for review July 7, 2020)

CRISPR-Cas gene editing tools have brought us to an era of syn- and “How could we detect change, and early warnings of change
thetic biology that will change the world. Excitement over the going awry?” Promising methods once thought to be in the do-
breakthroughs these tools have enabled in biology and medicine main of either molecular biology or of field ecology are now in-
is balanced, justifiably, by concern over how their applications creasingly used in both and will help us collaborate to meet these
might go wrong in open environments. We do not know how challenges. Next, I discuss values. To the credit of scientists and
genomic processes (including regulatory and epigenetic pro- others at the forefront of the CRISPR revolution, broad societal
cesses), evolutionary change, ecosystem interactions, and other values have been very much a part of the discussion from its onset
higher order processes will affect traits, fitness, and impacts of (1–7). Last, I describe resources, efforts, and training that could
edited organisms in nature. However, anticipating the spread, help those who release edited creatures into open environments to
change, and impacts of edited traits or organisms in heteroge- meet their responsibility to track and steward their impacts.
neous, changing environments is particularly important with
“gene drives on the horizon.” To anticipate how “synthetic Linking Genes, Traits, Organisms, and Ecosystems:
threads” will affect the web of life on Earth, scientists must con- Challenges to Prediction
front complex system interactions across many levels of biological
organization. Currently, we lack plans, infrastructure, and funding The systems we isolate mentally are not only included as part of larger
for field science and scientists to track new synthetic organ- ones, but they also overlap, interlock, and interact with each other.
isms, with or without gene drives, as they move through open
environments. Arthur Tansley, 1935 (8)

ECOLOGY
Genes, including edited genes, may affect traits of organisms that
|
CRISPR gene drives | ecological impacts | scale linkages | in turn alter their performances and, more rarely, their impacts
interaction strength
on ecosystems. Reciprocal feedbacks run countercurrent, from
environment to ecological interactions to gene expression. Few if

C RISPR-Cas gene editing tools have brought us into an era of any pathways linking cause and effect across these scales of bi-
Downloaded from https://www.pnas.org by 205.178.183.22 on December 25, 2023 from IP address 205.178.183.22.

synthetic biology that will change the world (1, 2). Having ological organization are straightforward. At every level, strongly
already enabled breakthroughs in basic research and medicine, nonlinear controls, variable time lags, and context dependence
CRISPR tools promise more in agriculture, public health, and “change the ground rules for existence” (9) and the impacts of
conservation (1–3). For applications outside the laboratory or traits, sometimes abruptly (10). Add to this system-level com-
hospital, we need to anticipate how gene-edited organisms and plexity the rich diversity of components: genes, organisms, and
their traits will spread, change, and affect other biota and eco- landscapes with idiosyncratic evolutionary and natural histories,
systems. As I write this in 2020, we in the United States find and it is clear why ecology, the culmination of all these nested
ourselves crippled in our efforts to contain and address a new interactions, can justifiably be called the most complex system
human virus because we are blinded to its spread by a lack of science has ever tried to understand (11).
field testing and monitoring. Will we do better as (or before) we
unleash our own novel organisms into Earth’s biosphere? Predicting Traits from Genes
How will humans and nature be changed as threads of new,
It was long believed the sequence of genes in a genome was all that
synthetic biology weave their way through our lives and our
was needed to understand that organism’s biology. Recently, scien-
world? What must we do to anticipate, track, and steward these tists have realized there’s another level of control: the epigenome.
changes? Ecologists, particularly field ecologists, must engage
with these issues before, during, and for prolonged periods after Joseph Ecker, 2020 (12)
releases, if we are to realize CRISPR technology’s benefits and
reduce its potential for harm. Geneticists once hoped to predict phenotypes of organisms di-
These questions require consideration: 1) If an edited organ- rectly from their genetic codes, but this has not proven possible
ism is viable in the laboratory, what will the altered gene(s) do to for most traits of ecological significance. As they probed gene
the organism in an open environment, in the near term and over regulation networks, molecular and developmental geneticists
time? 2) If an organism with an edited genome can reproduce in
an open environment, how will its progeny change and spread
through populations? The development of gene drives in par- This paper results from the NAS Colloquium of the National Academy of Sciences, “Life
ticular compels attention to this question (3–7). 3) How will 2.0: The Promise and Challenge of a CRISPR Path to a Sustainable Planet,” held December
10–11, 2019, at the Arnold and Mabel Beckman Center of the National Academies of
these new organisms perform in different ecological contexts? 4) Sciences and Engineering in Irvine, CA. NAS colloquia began in 1991 and have been
How will they affect their ecosystems? 5) How will we, or future published in PNAS since 1995. The complete program and video recordings of presenta-
humans, feel about these organisms? Will we like their direct and tions are available on the NAS website at http://www.nasonline.org/CRISPR. The collection
of colloquium papers in PNAS can be found at https://www.pnas.org/page/collection/
indirect effects?
crispr-sustainable-planet.
Below, I review nested scales of interacting biological systems
Author contributions: M.E.P. wrote the paper.
that determine fates and impacts of organisms modified by gene
The author declares no competing interest.
editing in open environments. Second, I address issues raised by
applications of gene drives that can spread synthetic alleles This article is a PNAS Direct Submission.

through wild populations. Third, I explore the impact of nitrogen Published under the PNAS license.

fixation by a river diatom under drought versus normal flow re- 1

Email: [email protected].
gimes. The questions become: “What could possibly go wrong?” Published April 30, 2021.

PNAS 2021 Vol. 118 No. 22 e2004833118 https://doi.org/10.1073/pnas.2004833118 | 1 of 8

 uncovered increasing complexity. Gene expression clearly changes female flies completely infertile. Starr and Cline (21) found that
during developmental ontogeny, but how it changes can be under parasitism by the bacterium Wolbachia allowed parasitized Sxl
strong environmental control, both at cellular and ecosystem mutant female flies to lay viable eggs. The authors pointed out
scales. How given gene sequences, including edited sequences, will that such interactions (which would be impossible to anticipate)
actually manifest as traits in living organisms depends on a variety may shift the impact of species interactions from negative to
of genomic features (transposable elements, copy number varia- positive, and even to obligate interactions of hosts with former
tions, chromosomal aneuploidy); epigenetic effects (methyl groups parasites (21). Subsequent work showed that a bacterial endo-
on cytosine [DNA] bases, chromatic modifications, small RNAs); symbiont of aphids made aphids resistant to parasitoid wasps if
as well as primary and secondary pleiotropy, homeobox regulation, the bacteria were infected by phage with genes encoding Cyto-
phenotypic plasticity, and other phenomena (13–15). Epigenetic lethal Distending Toxins (22, 23). Whiteman, students, and col-
elements that alter gene expression interact with each other in leagues (24) showed that one of these genes, cdtB, was encoded
complex, nonlinear ways, for example, during meiosis and mitosis in the eukaryotic genomes of two host aphids (Myzus spp.), likely
in plants (16). Environmentally induced changes in gene tran- arriving via horizontal gene transfer. They also found that this
scription can have lifelong or even intergenerational phenotypic “domestication” of the prokaryotic cdtB gene had occurred in
consequences (14, 17, 18). several drosophilid lineages, possibly via horizontal gene transfer
Epigenetics are particularly exciting, and humbling, for the mediated by mites, brachoviruses, or direct integration of the
quest to predict phenotype from genotype. Empirical and ex- phage into the aphid and fly genomes (24). For our discussion
perimental study of the phenotypic, ecological, and evolutionary here, two points are critical. First, the change in gene sequencing
consequences of epigenetics for nonmodel organisms in natural and traits was triggered by species interactions of phage, bacte-
environments has just begun, and is particularly challenging for ria, and insects. This vividly make Tansley’s (8) point, that the
the many plants that have large, complex genomes or are poly- ecosystems “we isolate mentally are not only included as part of
ploid (17). Ecological epigeneticists call for increased transfer of larger ones, but they also overlap, interlock, and interact with
knowledge and methods from model species research to ge- each other.” Second, these tortuous evolutionary steps and
nomes of evolutionarily divergent species, and for more mech- “worm holes” link genes to ecology via multilevel upscaling and
anistic, experimental studies in complex natural environments downscaling processes. They are quite challenging to explain,
(17, 18). As Richards et al. (17) point out, to anticipate impacts let alone predict, as captured by the title of Jonathan Losos’ (25)
of these genomic-level processes in diverse, nonmodel organisms book, Improbable Destinies: Fate, Chance, and the Future of
in real (open) environments, we would need to know the extent Evolution.
and sources of epigenetic variation in natural populations, and
whether such variation could alter ecological interactions and Gene Drives
have evolutionary consequences. In wild populations facing uncertain futures, maintaining grist
Downloaded from https://www.pnas.org by 205.178.183.22 on December 25, 2023 from IP address 205.178.183.22.

Complex interactions of genomic elements, the general lack of for future evolution argues against tinkering with life in ways that
detailed genomic information for most organisms, and feedbacks diminish wild genetic diversity. However, gene drives that over-
from complex ecological interactions make the consequences of write alternative “wild-type” alleles do exactly that as they spread
gene editing hard to predict in whole organisms. However, if we through populations. At least drives using a genetic scalpel like
could reliably map edited genes to traits of organisms released CRISPR-Cas9 rather than the obsidian knives of earlier gene-
into natural environments, how well could we then anticipate altering technologies overwrite much shorter (typically <1 kb)
their ecological impacts, particularly under different environmental stretches of the adjacent genomic sequence.
conditions? Context dependence at all scales, but particularly at Gene drives are genetic elements that bias transmission of
this largest scale, is a formidable challenge for predicting the fates alleles in sexual species upward from the expected Mendelian
and impacts of edited organisms in nature. probability of 50%. Drives have evolved in nature [e.g., Wolba-
chia, transposons, and other “selfish” genetic elements (3, 4)].
Predicting Performance from Genes and Traits Wolbachia gene drives were proposed for biocontrol of pests (26)
Selective environments shift, even in highly controlled agricul- and have been deployed in open environments in Brazil, French
tural, hospital, or laboratory ecosystems. Perhaps the most con- Polynesia, southeastern Australia (27), and the Central Valley of
trolled environment in the history of life on Earth has still given California (28) for biocontrol of mosquito vectors of human
rise to evolutionary surprises. Richard Lenski and his students disease. With CRISPR-Cas techniques, genetic cargos can be
and colleagues have tracked genetic changes in 12 initially precisely edited, then driven with much higher efficiency (>99%
identical populations of Escherichia coli through >70,000 gen- copying fidelity and transmission rates) than achieved by Wol-
erations as of April 2020 when the experiment was paused for bachia drives (3, 29). CRISPR-Cas drives to render female
COVID (19). Every day, 1% of each population was transferred Anopholes gambiae (which vectors malaria in Africa) sterile have
to a flask of fresh identical medium. At 500-generation intervals, been extensively studied in enclosures at Imperial College,
a subsample of each population was cryopreserved. London, and are being considered for release in Africa in the
Despite this extraordinary control over their genotype and future (30). Gene drive applications to suppress mosquitos and
environment, the 12 E. coli clones delivered many evolutionary insects in general, however, are currently challenged by rapid
surprises. One is particularly cautionary for the use of gene drives evolution of host resistance (31, 32). Various solutions are under
to spread edited genes through wild populations. Two E. coli clones investigation, including targeting a doublesex gene in which a
had mutations that would eventually take over the population, but mutation that blocked the drive would also render Anopholes
they were initially were outcompeted by two other mutated clones gambiae females sterile. The chance of evolved resistance can be
that later went extinct. Clones that won over the longer term had reduced if the genes at more than one locus were edited (30–32).
greater potential for further adaptation due to a mutation that Another approach to counter evolved host resistance is to use
altered chromosomal supercoiling and affected gene transcrip- drives that might enhance, rather than decrease, the fitness of
tion, enhancing mutation rates (20). Initially conferring a fit- the edited host. Anthony James and colleagues (33) have used
ness disadvantage, over the longer term, the topoisomerase CRISPR to edit Anopholes stephensi, which vectors malaria in
mutation that altered chromosomal coiling allowed these India. They introduced two independent transgene alleles into
clones to generate descendants with new, beneficial mutations. the mosquito hosts. Each codes for antibodies that attack dif-
Another dramatic “reversal of fortune” was documented in ferent life stages of Plasmodium parasite: one that infects the
Drosophila with an Sxl mutation that normally renders eggs of host’s gut and the other, host salivary glands. Again, these two

2 of 8 | PNAS Power
https://doi.org/10.1073/pnas.2004833118 Synthetic threads through the web of life
 COLLOQUIUM
gain-of-function alleles make the evolution of Plasmodium re- limitation) or because of competitors or natural enemies. Im-

PAPER
sistance less probable. In addition, resistance to Plasmodium may pacts of consumers, particularly predators and pathogens, are
improve mosquito host fitness and increase the probability that difficult to observe in nature, so we tend to underestimate their
gene-driven Anopholes stephensi could persist and spread after importance. We often assume that if “the world is green,” it is
introduction to the Indian subcontinent. Such a release awaits because conditions and resources support lush plant growth. If
considerable public and regulatory discussion, but meanwhile, the world looks more barren, conditions for plants are not hos-
the performance of edited A. stephensi is being carefully studied pitable. More often than is generally appreciated without ex-
in increasingly realistic but contained environments—insectary periment or prolonged observation, consumers are exerting
cages in Irvine, California, where (as in the trials in London), the cryptic, strong direct and indirect effects on ecological com-
edited mosquitoes could not persist if they escaped (7). munities: limiting some species, and indirectly releasing oth-
Could we recall organisms with gene drives after they have ers: their prey’s prey (42).
been released? Creative ideas for controlling unwanted impacts Paine (43–45) showed the explanatory, and potentially pre-
or halt the spread of gene-driven organisms released into nature dictive power of identifying “trophic cascades”—strong chains of
have been proposed (2, 29), and some tested in the laboratory direct and indirect interactions that link plants through con-
(34). These include immunization drives (e.g., to protect species sumers to predators, and send ripples through ecosystems if the
in their native range that are targeted as exotics elsewhere); abundances or performances of “strong interactor” species
recall drives that could be sent through a population to seek and change. Strong interactors are “foundational” or “dominant”
destroy edited, dispersed organisms; chemical tethers already species if they maintain sufficient biomass to physically structure
used successfully in laboratory (34) trials, to make edited or- their ecosystems, and “keystone species” if they are uncommon
ganisms reliant on chemicals that they cannot obtain in the wild. but nevertheless capable of strongly affecting ecosystem struc-
We currently cannot assess how well immunization or recall ture, sometimes by suppressing or releasing dominants (43, 46).
drives might work over time in open natural environments. Nor If, as Paine and others have postulated, relatively few strong
do we have sufficient funding, organization, and personnel to interactions trigger cascades of indirect effects that reverberate
track edited genes driven through open environments. It is un- throughout communities, the dynamics of many co-occurring
clear how we would learn whether genes in mosquitos that confer populations may be entrained to the fates of a few (43), mak-
resistance to a particular parasite may leave the vector more ing ecology more predictable.

ECOLOGY
vulnerable to hosting another pathogen (4). The low prevalence Context dependence, however, remains a formidable challenge
of infection of mosquitoes in most areas of malaria endemism to ecological prediction (46–48). Species interaction strengths
(35) might reduce this concern (why are not uninfected mos- change across space and time as conditions and resources alter
quitoes hosting other pathogens?), unless gene editing itself performance of web members. Adding or subtracting players can
improved the suitability of engineered mosquitoes as vectors for
change everything. For example, the trophic cascade through
Downloaded from https://www.pnas.org by 205.178.183.22 on December 25, 2023 from IP address 205.178.183.22.

other pathogens, a change for which, at this time, there seems to

which sea otters protect kelp forests from overgrazing throughout
be no obvious mechanism. If, however, genetic mutation or non-
much of the northeastern Pacific (49) collapsed when killer whales
genetic mechanisms triggered such changes in gene-driven edited
began to prey on, and locally extirpate sea otters (50), possibly as a
organisms over time, we would at this point be blind to them.
consequence of the depletion of great whales, once preferred prey
Other general concerns apply to gene drives, independent of
of killer whales, followed by population collapses of other, fatter
our intended use of them to suppress, enhance, or simply change
populations of wild plants and animals. In light of Lenski’s long- marine mammals (51). Although impacts of an organism cannot
term evolution experiment, it is concerning that CRISPR-Cas9 be predicted from traits, there should be rules that let us map
gene drives will efficiently overwrite homologous alternative al- changes in interactions and interaction strength over space and
leles on paired chromosomes as they spread. Loss of allelic di- time. Developmental geneticists have discovered such rules as they
versity might deprive descendant populations of their tickets painstakingly unraveled complex, idiosyncratic control paths un-
through an unpredictable future (20, 25). Finally, context de- derlying tissue-engendered effects on gene expression (14, 15).
pendence thwarts prediction over larger scales. Novel traits that Despite the challenges of working over much larger scales, ecol-
seem beneficial (to biota or society) in some settings may be ogists need to advance analogous efforts in landscapes, seascapes,
harmful in others, and vice versa. These considerations under- and fresh waters, our domains of investigation.
score the need for careful evaluation, preparation, and consensus
Hidden Processes, Cryptic Players
before organisms with gene drives are released (3–6), as well as
chemical tethers, genetic time bombs, recall drives, or other In his remarkable book, The Serengeti Rules, Sean Carroll (52)
methods to cope with unpleasant surprises after release. pointed out the parallels between trophic cascades in ecosystems,
and regulation networks governing gene expression in cells.
Predicting Impacts from Traits Genetic circuits—molecular chains of command—regulate how,
Ecology is classically defined as the study of factors affecting the when, and if genes will be expressed, repressed, or induced. The
distribution and abundance of organisms (36). Most observers, Lac operon defied understanding until Monod and Jacob real-
including many ecologists, assume that populations are primarily ized that another cryptic player was involved, a protein that re-
limited by environmental conditions and resources. Organisms pressed the beta-galactosidase gene until its substrate, lactose,
should occur where they can tolerate physicochemical conditions repressed the repressor (52). Carroll pointed out that this control
and where resource supplies meet their requirements (37). To path functioned in the same way as a three-level trophic cascade,
predict responses of species to greenhouse warming, for exam- in which predators indirectly protect plants by suppressing her-
ple, a popular approach is to analyze their climate envelopes (the bivores (42). As we discover and unravel the often indirect, idi-
ranges of temperature, moisture, sometimes other conditions osyncratic control paths in gene regulation networks, metabolisms,
where they are currently found) and assume that they will survive or ecosystems, we improve our chances of anticipating, under-
in or move to regions where such conditions occur in the future standing, and managing the surprises. Another hopeful sign is that
(38–40). from genes to ecosystems, multiple manifestations first perceived
These views ignore species interactions, so are in a sense as mutually contradictory can be conceptually unified when the
analogous to “bean bag” genetics (41). Resource supplies and underlying mechanism is revealed (14, 15, 53). This often involves
conditions clearly matter, but organisms are often not where the discovery of controls over phenomena that operate at higher
they could be, either because they have not arrived (dispersal levels of organization.

Power PNAS | 3 of 8
Synthetic threads through the web of life https://doi.org/10.1073/pnas.2004833118
 Mutations in single genes can lead to multiple manifestation of closest free-living relatives are cyanobacteria in the genus Cya-
brain disorders through pleiotropy, copy number variation, single- nothece (67, 68). While other nitrogen-fixing cyanobacteria syn-
nucleotide substitutions, and epigenetic dysregulation, while dif- thesize nitrogen-bearing toxins, Epithemia’s endosymbionts synthesize
ferent genes can lead to convergent phenotypes by disrupting all 23 essential amino acids required by animals (68), despite ge-
common neurodevelopmental pathways (14, 15). Analogously, bird, nome reduction that eliminated both photosystems (68, 69). In
lizard, and plant ecologists have argued for the primacy of com- unpolluted lakes and rivers where they often abound, Epithemia-
petition as a force structuring ecological communities (e.g., ref. 54), rich algal assemblages are consumed voraciously and preferen-
while this view is more rare among insect ecologists (e.g., ref. 55). tially by algivorous insects (70) and tadpoles (71) and, when
These views could be (somewhat simplistically) reconciled if three- exported from rivers to estuaries, by amphipods and isopods
level food chains are common, so that birds or lizards at the top and that are important fish and shorebird prey (72). As diatom-
plants at bottom of the food chain are often resource limited, cyanobacterial holobionts, Epithemia support much higher rates
whereas insects (at the second trophic level) are predator limited, of growth and emergence in aquatic insects (70) and tadpoles (71)
and therefore not in competition for limiting resources (53). than do other detrital or algal diets.
Making a diatom into an even more complete “superfood” like
Context-Dependent Impacts of Traits: Nitrogen Fixation Epithemia would seem a worthy goal for gene editing, had not
Bioavailable nitrogen is broadly limiting to biota across terres- evolution already done it. During drought, however, Epithemia
trial, marine, and many freshwater environments. Nitrogen fix- blooms that fuel salmon-bearing river food chains during higher
ation (reduction of N2 gas to ammonia) appears to have evolved river flows can trigger indirect effects that threaten public
only once, before Earth’s atmosphere was oxygenated, and is still health (73).
restricted to only a few archaea and bacteria (56). A number of Under the Mediterranean seasonality of Northwestern Cal-
these have entered into symbiotic partnerships with eukaryotes, ifornia, rainy winters are followed by summer drought. Winter
from diatoms (57) to vascular plants (56), in which microbial and summer flows govern algal phenology, and how river food
nitrogen fixers trade bioavailable nitrogen for reduced carbon webs assemble during the biologically active summer season (48,
and energy from their hosts. Since the 1970s, a major goal for 73, 74). In spring, as river flow subsides, clears, and warms, fil-
gene editing in agriculture has been to extend the host range of amentous green macroalgae (Cladophora glomerata) can grow
these symbioses to nonlegume crops, particularly cereals (58). several meters long, increasing surface habitat for attached dia-
Another approach would be to transfer prokaryotic nif genes for toms and other epiphytes by five to six orders of magnitude (70).
nitrogenase biosynthesis and function directly into the plant By midsummer, Cladophora streamers turn from green to rusty
genome (59). Broadly replacing synthetic nitrogen fertilizer with red as they are smothered by Epithemia. They nourish prey that
biological fixation in agriculture would have clear economic and support rearing salmonids, not only in the river, but also in the
environmental benefits, saving enormous energy costs from in- estuary. Diatom–Cladophora assemblages that drift or are ex-
Downloaded from https://www.pnas.org by 205.178.183.22 on December 25, 2023 from IP address 205.178.183.22.

dustrial Haber–Bosch nitrate–ammonia production, and reduc- perimentally introduced into the estuary are grazed away within
ing eutrophying runoff from agricultural fields to surface and minutes by swarms of amphipods and isopods, who greatly prefer
coastal waters. What could possibly go wrong? them to the local soft green seaweeds (Ulva and Enteromorpha)
Environmental and social concerns echo those raised by en- generally assumed to be important food sources for coastal marine
vironmentalists and the public since the original introduction of food webs (72). These observations suggest that drift algae from
GMO (genetically modified organism) crops (60). These con- rivers can be important but cryptic energy and nutrient source to
cerns include the spread of nitrogen-fixing genes into wild rela- estuarine and coastal food webs, yet another example showing that
tives, particularly via pollen for wind-pollinated cereals. Release strong top-down limitation can conceal strong links in food chains,
from nitrogen limitation could make wild graminoids into ag- until experiments or changing circumstances reveal them.
gressive weeds. On the other hand, with increased nitrogen Under severe summer drought, the fate of Epithemia-rich as-
content, crops and of nontarget recipients of edited genes could semblages and the nitrogen they fix changes. These prolifera-
suffer more herbivory, due to their increased nutritional value tions stop fueling salmon-bearing food chains, and instead
(61). If root exudates or litter from edited crops or transformed support blooms of potentially neurotoxic cyanobacteria (73, 75).
wild plants enriched soil nitrogen, they could facilitate weed If flows in sunlit river mainstems drop below critical levels [which
invasions and loss of native flora adapted to low nutrient con- can occur even after rainy winters due to human water extraction
ditions. When nonnative Myrica invaded and established as (76)], pools warm and stagnate, and nutritious Cladophora–
Hawaii’s first nitrogen-fixing shrub, it changed the “ground rules Epithemia assemblages are overgrown by heat-tolerant, poten-
for existence” for native plant species, and many went extinct (9). tially neurotoxic cyanobacteria dominated by Anabaena spp. (75,
Native plant and arthropod diversity has been lost from patches 77). As fronts of Anabaena spread through Cladophora–Epithemia
where nitrogen fixers (vetches and lupine) enriched soils and assemblages, they smother red-brown host assemblages under
facilitated weed invasion in coastal (62) and inland (63) Cal- blue-green to black cloaks. Anabaena get access to sunlight and
ifornia grasslands. In the Catalina Mountains of Arizona, inva- likely also nutrients or energy from their dying hosts. While
sive buffelgrass brought wildfire [still burning as of June 2020 Anabaena fixes nitrogen, it also takes up carbon heterotrophically,
(64)] that kills ancient saguaro forests. which extends its ability to fix nitrogen into hours of darkness (78).
Impacts of nitrogen fixation in river ecosystems can also Over the last decade, neurotoxic cyanobacteria have killed more
change from beneficial to adverse (for humans), as hydrologic than a dozen dogs in the Eel, Russian, and other Northern Cal-
context shifts the fate of the fixed nitrogen. In clear, sunlit rivers, ifornia Rivers (73, 75). To the south, 20 sea otter deaths off
attached algae fuel food webs (65). Diatoms, often dominants in Monterrey Bay were linked to microcystins (hepatotoxins) pro-
riverine algal biofilms, are among the most nutritious of all pri- duced by other cyanobacteria in the agriculturally enriched rivers
mary producers. With the exception of one toxic genus, diatoms (79). Whether nitrogen-fixing diatoms or other nutrient sources
are digestible, synthesize beneficial (to animals) secondary com- deliver nourishment or toxins to food webs in rivers and linked
pounds like carotenoids, and are rich in lipids, including polyun- upland and coastal ecosystems depends on winter and summer
saturated fatty acids that animals require for membrane and nerve hydrology, climate, and increasingly, human choices (73, 76).
health, but cannot synthesize themselves (66). Freshwater at-
tached diatoms in the family Epithemiaceae are particularly nu- Tinkering with Genes, Organisms, and Ecosystems: Values
tritious. Epithemia spp. contain what appears to be Earth’s Charles Mann (80) contrasts two schools of 20th-century think-
youngest endosymbiont: nitrogen-fixing “spheroid bodies” whose ers and scientists: “wizards”—engineers, inventors, or tinkerers

4 of 8 | PNAS Power
https://doi.org/10.1073/pnas.2004833118 Synthetic threads through the web of life
 COLLOQUIUM
who use technology to fix problems—and “prophets” who revere fire, flood, and drought regimes (95). Indigenous people’s views

PAPER
nature: organisms and landscapes beyond human domination. of using synthetic biology to restore species (e.g., American
The wizard “sees people as endlessly inventive ... wily managers chestnuts or salmonids) or ecosystems in their lands appear,
and thinkers and doers that can expand endlessly. [Prophets] see however, to be negative or mixed at present (93).
us as fundamentally embedded in ... something larger, and we
shouldn’t wreck that larger thing” (81). Twenty-first century Constant Vigilance, Beyond the Laboratory: The Importance
CRISPR scientists would at first seem to be in the techno-optimist of Tracking Synthetic Threads through Real Environments
wizard camp. A more nuanced view is raised by their hopes of
[S]cientists want to transfer, enhance, or silence genes to make mi-
applying CRISPR editing and gene drives for conservation of wild
crobes work for us without having to hassle with natural selection. We
species and restoration of seminatural ecosystems (2). Gene will be the creators of microbial metabolism, and will design microbes
editing, gene drives, and other modern genetic tools give conser- to do our bidding. We have the power to do so, but that power does
vationists new, and in some cases, perhaps our only hope of un- not appear to come with an understanding of the potential tremen-
doing some of the damage humans have done to wild species and dous consequences for microbial evolution, let alone our role in al-
seminatural ecosystems. Interspecies somatic cell nuclear transfer tering the future trajectory of the planet.
may enhance genetic diversity in species like black-footed ferrets
or gray wolves that humans have reduced to low population sizes Paul Falkowski (56)
(82). Coral biologists are exploring gene editing to make corals, or
Symbiodinium, their dinoflagellate endosymbionts, less stressed in Just by being vigilant—by being out there—one comes to recognize
the current and near-future oceans that our greenhouse gases have change. Why is the world we know so well changing?
warmed and acidified (83). Gene drives are widely discussed as
ways with fewer off-target impacts (than, for example poisoning) Robert T. Paine (96)
for eliminating pests, disease vectors, or predators we have intro-
duced to islands, where they now threaten native species, including Dr. Frankenstein’s crime was not that he invented a creature through
endemics (3, 4, 84). Their ethical use requires attempts to predict some combination of hubris and high technology, but rather that he
the full range of outcomes (my focus in this paper); analysis of risks, abandoned the creature to itself. [LaTour then cites from Mary
benefits, and opportunity costs; public engagement and acceptance; Shelley’s Frankenstein: “Remember, I am thy creature,” the monster

ECOLOGY
and oversight—what Sandler (85) has classified as criteria for “an protests, “I ought to be thy Adam; but I am rather the fallen angel,
instrumentalist ethical perspective,” in which the technologies or whom thou drivest from joy for no misdeed ... I was benevolent and
tools are “neither good nor bad, but neutral.” good; misery made me a fiend. Make me happy, and I shall again be
Sandler argues that, while important, the instrumentalist eth- virtuous.”]
ical perspective is incomplete without another “form-of-life
Downloaded from https://www.pnas.org by 205.178.183.22 on December 25, 2023 from IP address 205.178.183.22.

perspective” that considers how gene editing will restructure our Bruno LaTour (97)
activities and our relationships with life. Gene editing is not only
In these early days of gene editing of nonmodel systems, many
an efficient, perhaps sometimes necessary means to certain
synthetic biologists focus on whether their new creations will be
ends—it will alter how we and future humans will feel about
viable when not coddled in the laboratory. However, it is not too
nature as it is increasingly populated by gene-edited creatures,
early to ask “what happens if they survive, establish, and spread?
even if these serve as our agents for species conservation or
What happens if they are ‘driven’ through wild populations?”
ecosystem restoration. For some of us, it is very appealing to
Forward prediction of their futures, and how they will affect
“resurrect” species that could remake past landscapes and eco-
systems that we mourn, from the American chestnut-dominated ours, will be challenged and often thwarted by scale-spanning
forests of central and eastern North America (86–88) to the complexities and contingencies, including those I have reviewed
mammoth steppes of Beringia (89, 90). If we restrict human here. The more we seek predictive mechanistic understanding of
activities and free enough space and time so that these species processes that drive change and link genes to ecosystems, how-
can evolve and reconfigure ecological systems on their own, we ever, the more nimble we will be in postdicting—explaining sur-
could possibly maintain aspects of wildness and wilderness im- prises. For either prediction or postdiction, 21st-century tools
portant to the prophets (80) among us today. If, in an even more should help us realize some of the promise of synthetic biology
crowded, human-dominated world, we do not so restrict ou- and reduce its potential for harm. Methods once thought to be in
rselves, we will have to endlessly tinker with Earth’s biota and the domain of either molecular biology or of field ecology are now
ecosystems to fix problems and maintain life-supporting pa- are increasingly available in both (17, 98). Spillover of methods
rameters within acceptable ranges (91) (the Band-Aid on a across fields should enhance the interchange among molecular
Band-Aid problem). geneticists, cell biologists, epigeneticists, ecologists, evolutionary
How will current and future humans feel about gene-edited biologists, and Earth scientists needed to meet the challenge and
creatures, even as our agents of species conservation or ecosys- responsibility of CRISPR-Cas use in the world. LaGrangian ob-
tem restoration? In his near-future science fiction novel about servations of free-living individuals, which have greatly enlight-
rampant synthetic biology and its potential, via competitive ened field biologists, are now enabling cell biologists to learn
biowarfare among agrobusiness companies, to impoverish the about the functioning and fates of single cells as they spread and
human food supply, Paolo Bacigalupi (92) depicts the cultural change across space and time within their living microenviron-
revulsion some humans feel toward synthetic creatures as per- ments (99, 100). Quantitative stable isotope probing lets ecologists
sisting. Human perceptions and values change, however. Amer- and ecosystem scientists identify key microbial taxa and track the
ican conservationists of European descent once revered North elemental and molecular exchanges they mediate in nature (99,
America’s “pristine wilderness” but have recently begun to ap- 101). Geneticists use “knockin” and “knockout” experiments to
preciate the degree to which old growth forests and the diverse investigate how genes affect phenotypes; ecologists use enclosure/
mosaics of plants and game that European colonists found in exclosures, and other manipulative experiments to reveal how
were, in fact, created and maintained by careful tending and fire particular species affect food webs. Ecologists (43–50, 102)
management by Native Americans (93–95). Wildlands managers searched for trophic cascades set off by strong interactors (foun-
and conservation biologists increasingly realize that we need to dation or keystone species), in hopes of explaining or predicting
return to such management to sustain western forests and their indirect “knockon” effects entrained to these. Similarly, molecu-
biota and life-supporting functions through changing regional lar biologists search for critical “gene programs,” those gene

Power PNAS | 5 of 8
Synthetic threads through the web of life https://doi.org/10.1073/pnas.2004833118
 regulatory pathways that entrain many others and determine that no Wolbachia-infected females be released. In a robotic
higher order phenotypic characters, like sex (103). At all scales of larval rearing system, Crawford and colleagues infected larvae of
biological organization, context controls interaction strengths and both sexes, but then separated female pupae with automated
outcomes, with the effect of “pattern on process” in landscape size-based sex sieving, then checked by an industrial image
ecology (104) scaling down to topologically associated domains of analysis and a machine learning classifier (with some human
gene expression in spatial genomics (105). Remote sensing ad- checks). Importantly for the argument here, the released Wol-
vances in Earth systems science enable us to track changes in the bachia males and the entire Aedes aegypti population were
distributions, abundances, and physiological states of biota, as well monitored over large areas. Carefully mapped release sites and
as environmental conditions, from aircraft or space (106). Broad subsequent trap monitoring let investigators check the success of
calls for “convergence” from funding agencies and organizations their biocontrol program, and monitor for mishaps, like the ac-
demand both intellectual humility and synthetic imagination from cidental release of Wolbachia-infected females.
biologists and Earth scientists to breach silo walls and bridge deep The strategic, extensive monitoring in this study shows that
historic divides among our subdisciplines. responsible tracking of released edited organisms is feasible, with
Such teamwork will be crucial for building detection networks sufficient support. The authors’ technological innovations (28)
to track edited organisms through open environments. These could facilitate other efforts. These monitoring programs will
should be designed to focus sampling effort on “hot spots or hot need coordinated networks of field biologists—trained, funded,
moments” where performances, selective forces, or impacts of and organized on the ground. It would be ideal if teams were led
edited organisms are most likely to change. We could map by local scientists and employed local youth, like Dan Janzen’s
shifting environmental controls, as well as the spread or con- parataxonomists in Costa Rica (113). Ideally, teams would include
spicuous change in edited organisms (e.g., trees or megafauna) members trained in genomics and epigenomics, ecophysiology,
using repeated photogrammetry from drones, aircraft, or space- and the relevant “ologies,” so they could anticipate behavior and
based observation platforms (106). Organisms whose CRISPR- fates of species with different evolutionary and natural histories.
Cas altered DNA has no externally visible phenotypic expression Team members educated in community and ecosystem ecology
would require genetic monitoring, either of trapped organisms or and remote sensing methods could design sampling across ap-
of environmental DNA (eDNA). propriate space–time intervals to determine whether traits and
The last effort is growing more feasible due to scientific and impacts of these organisms remain as intended.
technological innovations using CRISPR-Cas technology not Field monitoring programs to track our creatures, our
only for editing cells or organisms, but also for tracking their “hopeful monsters,” could not only provide meaningful work for
eDNA. New CRISPR-Cas12a [e.g., DETECTR (107)] and people in impoverished regions of the world (including the
CRISPR-Cas13 [SHERLOCK (108)] platforms, developed for United States) but should enlist people with crucial Traditional
clinical diagnoses, have detection sensitivity down to attomolar Ecological Knowledge to advise stewardship. Releases that can
Downloaded from https://www.pnas.org by 205.178.183.22 on December 25, 2023 from IP address 205.178.183.22.

concentrations, and can discriminate nucleotide strands that be spatially contained should occur only if all indigenous and
differ in just a few base pairs. Following preamplification of other local people in the arena have consented. More uncon-
target DNA with recombinase polymerase amplification, a guide tained releases that could obliterate a targeted species from the
RNA directs a CRISPR-Cas12a nuclease to the targeted site. Earth should require more widespread agreement (6), although
Once the enzyme cleaves that site, CRISPR-Cas12a cleaves how to attain this from ∼8 billion human individuals is an open
other single-stranded DNA indiscriminately (“transcleavage”). If question. Consent or consensus may or may not develop, even
a single-stranded DNA fluorophore-quencher molecule has been after prolonged engagement and intense listening, but the pro-
added, Cas12a transcleavage activity will release the fluorophore cess would educate both scientists and stakeholders in the perils
from its quencher, triggering fluorescence that can then be and promise of using edited organisms in nature. It also might
measured in the sample. These detection systems are potentially help more of us think about where nature (including human
inexpensive and adaptable for field applications: lyophilized for nature) will find a place in our increasingly engineered future.
cold-chain independence, then used with paper spotting (108) or When could we relax our watchful stewardship of our “hopeful
smartphone‐enabled detection devices (109); or a recently de- monsters”—our edited biotic agents, and the feedbacks they
veloped handheld fluorescent monitor developed for bacterial
trigger? Perhaps never, although the more we seek predictive
detection (110, 111).
understanding of processes and linkages from functional geno-
Williams et al. (110, 112) appear to be the first to apply this
mics to species interactions and ecosystem responses, the better
technology for species detection in nature. Sampling eDNA in
we will become at our two crucial responsibilities at the dawn of
four Irish streams, they could detect and differentiate closely re-
the CRISPR-altered Anthropocene: first, designing efficient schemes
lated Atlantic salmon (Salmo salar) and brown trout (Salmo
to track the spread, change, and impacts of edited genes across
trutta). Sensitive CRISPR‐Cas12a detection that greatly enhances
landscapes, and second, analyzing why things go awry, perhaps
eDNA “early warning” detection of valued, rare, or potentially
even soon enough to correct problems, and survive them.
invasive exotic species (109, 110, 112) could also track edited or-
ganisms spreading through open environments via gene drives. Data Availability. There are no data underlying this work.
In one of the most exhaustive gene drive monitoring programs
to date, Crawford et al. (28) have tracked Aedes aegypti mos- ACKNOWLEDGMENTS. Thanks very much to Barbara Meyer for inviting and
quitos (vectors for dengue, chikungunya, Zika, and yellow fever) encouraging me to participate in this colloquium; Tom Cline, Christina
with Wolbachia gene drives over an area almost 400 ha (including Richards, and Noah Whiteman for invaluable comments and genetic counsel;
3,000 homes) in Central California. If Wolbachia-infected males Aaron Pomeranz for guidance through the literature on CRISPR-Cas 12a and
CRISPR-Cas 13 detection systems; Bill Dietrich and Terry Chapin for Earth
mate with uninfected females, zygotes die. If infected females science comments; Kevin Esvelt and an anonymous reviewer for very helpful
mate with infected or uninfected males, however, zygotes are suggestions; and National Science Foundation Award CZP EAR-1331940 (Eel
viable. For this sterile insect biocontrol to work, it is imperative River Critical Zone Observatory) for support.

1. J. A. Doudna, S. H. Sternberg, A Crack in Creation (Houghton Mifflin Harcourt, 2017). 4. G. E. Kaebnick et al., Precaution and governance of emerging technologies. Science
2. G. Church, E. Regis, Regenesis (Basic Books, 2012). 354, 710–711 (2016).
3. National Academies of Sciences, Engineering, and Medicine, Gene Drives on the 5. E. Heitman, K. Sawyer, J. P. Collins, Gene drives on the horizon: Issues for biosafety.
Horizon: Advancing Science, Navigating Uncertainty, and Aligning Research with Applied Biosafety International 21, 173–176 (2016).
Public Values (The National Academies Press, 2016). 6. K. A. Oye et al., Biotechnology. Regulating gene drives. Science 345, 626–628 (2014).

6 of 8 | PNAS Power
https://doi.org/10.1073/pnas.2004833118 Synthetic threads through the web of life
 COLLOQUIUM
7. J. Kahn, The gene drive dilemma: We can alter entire species, but should we? New 47. B. A. Menge, E. Berlow, C. Blanchette, S. Navarrete, S. Yamada, The keystone species

PAPER
York Times Magazine, 8 January 2020. https://www.nytimes.com/2020/01/08/maga- concept: Variation in interaction strength in a rocky intertidal habitat. Ecol. Monogr.
zine/gene-drive-mosquitoes.html. Accessed 19 January 2020. 64, 249–286 (1994).
8. A. G. Tansley, The use and abuse of vegetational concepts and terms. Ecology 16, 48. M. E. Power, M. S. Parker, W. E. Dietrich, Seasonal reassembly of a river food web:
284–307 (1935). Floods, droughts, and impacts of fish. Ecol. Monogr. 78, 263–282 (2008).
9. P. M. Vitousek, L. R. Walker, L. D. Whiteaker, D. Mueller-Dombois, P. A. Matson, 49. J. A. Estes, J. F. Palmisano, Sea otters: Their role in structuring nearshore commu-
Biological invasion by Myrica faya alters ecosystem development in Hawaii. Science nities. Science 185, 1058–1060 (1974).
238, 802–804 (1987). 50. J. A. Estes, M. T. Tinker, T. M. Williams, D. F. Doak, Killer whale predation on sea
10. M. Scheffer, S. Carpenter, J. A. Foley, C. Folke, B. Walker, Catastrophic shifts in otters linking oceanic and nearshore ecosystems. Science 282, 473–476 (1998).
ecosystems. Nature 413, 591–596 (2001). 51. A. M. Springer et al., Sequential megafaunal collapse in the North Pacific Ocean: An
11. M. E. Power et al., The role of experiments in ecology. Science 270, 561 (1995). ongoing legacy of industrial whaling? Proc. Natl. Acad. Sci. U.S.A. 100, 12223–12228
12. Salk Institute for Biological Studies, Joseph Ecker. https://www.salk.edu/scientist/joseph- (2003).
ecker/. Accessed 31 March 2020. 52. S. B. Carroll, The Serengeti Rules (Princeton University Press, 2016).
13. D.-S. Lee et al., Simultaneous profiling of 3D genome structure and DNA methylation 53. M. E. Power, Top-down and bottom-up forces in food webs: Do plants have primacy?
in single human cells. Nat. Methods 16, 999–1006 (2019). Ecology 73, 737–746 (1992).
14. E. Charney, Behavior genetics and postgenomics. Behav. Brain Sci. 35, 331–358 54. T. W. Schoener, Field experiments on interspecific competition. Am. Nat. 122,
(2012). 240–285 (1983).
15. C. K. Deutsch, W. J. McIlvane, Non-mendelian etiologic factors in neuropsychiatric 55. D. R. Strong, “Density-vague ecology and liberal population regulation in insects” in
illness: Pleiotropy, epigenetics, and convergence. Behav. Brain Sci. 35, 363–364 A New Ecology, P. W. Price, C. N. Slobodchikoff, W. S. Gaud, Eds. (Wiley, 1984), pp.
(2012). 313–327.
16. T. Kawakatsu, J. R. Ecker, Diversity and dynamics of DNA methylation: Epigenomic 56. P. G. Falkowski, Life’s Engines (Princeton University Press, 2016).
resources and tools for crop breeding. Breed. Sci. 69, 191–204 (2019). 57. H. R. DeYoe, R. L. Lowe, J. C. Marks, Effects of nitrogen and phosphorus on the
17. C. L. Richards et al., Ecological plant epigenetics: Evidence from model and non- endosymbiont load of Rhopalodia gibba and Epithemia turgida (Bacillariophyceae).
model species, and the way forward. Ecol. Lett. 20, 1576–1590 (2017). J. Phycol. 28, 773–777 (1980).
18. O. Bossdorf, C. L. Richards, M. Pigliucci, Epigenetics for ecologists. Ecol. Lett. 11, 58. S. Burén, L. M. Rubio, State of the art in eukaryotic nitrogenase engineering. FEMS
106–115 (2008). Microbiol. Lett. 365, 1–9 (2018).
19. Wikipedia, E. coli long-term evolution experiment. https://en.wikipedia.org/wiki/E._ 59. M. Charpentier, G. Oldroyd, How close are we to nitrogen-fixing cereals? Curr. Opin.
coli_long-term_evolution_experiment. Accessed 20 January 2020. Plant Biol. 13, 556–564 (2010).
20. R. J. Woods et al., Second-order selection for evolvability in a large Escherichia coli 60. K. V. Pixley et al., Genome editing, gene drives, and synthetic biology: Will they
population. Science 331, 1433–1436 (2011). contribute to disease-resistant crops, and who will benefit? Annu. Rev. Phytopathol.
21. D. J. Starr, T. W. Cline, A host parasite interaction rescues Drosophila oogenesis 57, 165–188 (2019).
defects. Nature 418, 76–79 (2002). 61. W. Mattson, Herbivory in relation to plant nitrogen content. Annu. Rev. Ecol. Syst.
22. K. M. Oliver, P. H. Degnan, M. S. Hunter, N. A. Moran, Bacteriophages encode factors 11, 119–161 (1980).

ECOLOGY
required for protection in a symbiotic mutualism. Science 325, 992–994 (2009). 62. J. A. Maron, R. L. Jefferies, Restoring enriched grasslands: Effects of mowing on
23. K. M. Oliver, P. H. Degnan, G. R. Burke, N. A. Moran, Facultative symbionts in aphids species richness, productivity, and nitrogen retention. Ecol. Appl. 11, 1088–1100
and the horizontal transfer of ecologically important traits. Annu. Rev. Entomol. 55, (2001).
247–266 (2010). 63. K. B. Suttle, M. A. Thomsen, M. E. Power, Species interactions reverse grassland re-
24. K. I. Verster et al., Horizontal transfer of bacterial cytolethal distending toxin B sponses to changing climate. Science 315, 640–642 (2007).
genes to insects. Mol. Biol. Evol. 36, 2105–2110 (2019). 64. M. Puleo, Wildfires dot the Southwest as dry, windy conditions prove disastrous.
Downloaded from https://www.pnas.org by 205.178.183.22 on December 25, 2023 from IP address 205.178.183.22.

25. J. Losos, Improbable Destinies: Fate, Chance, and The Future of Evolution (Riverhead AccuWeather, 15 June 2020. https://www.accuweather.com/en/severe-weather/wildfires-
Books, Random House, 2017). dot-the-southwest-as-dry-and-windy-conditions-prove-disastrous/759478. Accessed 1 July
26. A. Burt, Site-specific selfish genes as tools for the control and genetic engineering of 2020.
natural populations. Proc. Biol. Sci. 270, 921–928 (2003). 65. Y. Vadeboncoeur, M. E. Power, Attached algae as the cryptic base of inverted trophic
27. L. Alphey, Genetic control of mosquitoes. Annu. Rev. Entomol. 59, 205–224 (2014). pyramids in freshwaters. Annu. Rev. Ecol. Evol. Syst. 48, 255–279 (2017).
28. J. E. Crawford et al., Efficient production of male Wolbachia-infected Aedes aegypti 66. M. T. Brett, D. C. Muller-Navarra, The role of highly unsaturated fatty acids in aquatic
mosquitoes enables large-scale suppression of wild populations. Nat. Biotechnol. 38, foodweb processes. Freshw. Biol. 38, 483–499 (1997).
1–15 (2020). 67. J. Prechtl, C. Kneip, P. Lockhart, K. Wenderoth, U. G. Maier, Intracellular spheroid
29. K. M. Esvelt, A. L. Smidler, F. Catteruccia, G. M. Church, Concerning RNA-guided gene bodies of Rhopalodia gibba have nitrogen-fixing apparatus of cyanobacterial origin.
drives for the alteration of wild populations. eLife 3, 20131071 (2014). Mol. Biol. Evol. 21, 1477–1481 (2004).
30. K. Kyrou et al., A CRISPR-Cas9 gene drive targeting doublesex causes complete 68. T. Nakayama et al., Complete genome of a nonphotosynthetic cyanobacterium in a
population suppression in caged Anopheles gambiae mosquitoes. Nat. Biotechnol. diatom reveals recent adaptations to an intracellular lifestyle. Proc. Natl. Acad. Sci.
36, 1062–1066 (2018). U.S.A. 111, 11407–11412 (2014).
31. J. Champer et al., Reducing resistance allele formation in CRISPR gene drive. Proc. 69. C. Kneip, C. Voss, P. J. Lockhart, U. G. Maier, The cyanobacterial endosymbiont of the
Natl. Acad. Sci. U.S.A. 115, 5522–5527 (2018). unicellular algae Rhopalodia gibba shows reductive genome evolution. BMC Evol.
32. J. Champer, A. Buchman, O. S. Akbari, Cheating evolution: Engineering gene drives Biol. 8, 30 (2008).
to manipulate the fate of wild populations. Nat. Rev. Genet. 17, 146–159 (2016). 70. M. E. Power et al., Algal mats and insect emergence in rivers under Mediterranean
33. V. M. Gantz et al., Highly efficient Cas9-mediated gene drive for population modi- climates: Towards photogrammetric surveillance. Freshw. Biol. 54, 2101–2115 (2008).
fication of the malaria vector mosquito Anopheles stephensi. Proc. Natl. Acad. Sci. 71. S. J. Kupferberg, J. C. Marks, M. E. Power, Effects of variation in natural algal and
U.S.A. 112, E6736–E6743 (2015). detrital diets on larval anuran (Hyla regilla) life-history traits. Copeia 1994, 446–457
34. J. E. DiCarlo, A. Chavez, S. L. Dietz, K. M. Esvelt, G. M. Church, Safeguarding CRISPR- (1994).
Cas9 gene drives in yeast. Nat. Biotechnol. 33, 1250–1255 (2015). 72. C. Ng, “The transport of chemicals and biota into coastal rivers and marine ecosys-
35. M. S. Alam et al., Prevalence of anopheline species and their Plasmodium infection tems,” PhD dissertation, University of California, Berkeley, CA (2012).
status in epidemic-prone border areas of Bangladesh. Malar. J. 9, 15 (2010). 73. M. E. Power, K. Bouma-Gregson, P. Higgins, S. M. Carlson, The thirsty Eel: Summer
36. H. Andrewartha, L. Birch, The Distribution and Abundance of Animals (University of and winter flow thresholds that tilt the Eel River of Northwestern California from
Chicago Press, 1954). salmon-supporting to cyanobacterially degraded states. Copeia 2015, 200–211 (2015).
37. G. E. Hutchinson, Concluding remarks. Cold Spring Harb. Symp. Quant. Biol. 22, 74. J. B. Sculley, R. L. Lowe, C. A. Nittrouer, T. M. Drexler, M. E. Power, Eighty years of
415–427 (1957). food-web response to interannual variation in discharge recorded in river diatom
38. J. Harte, Maximum Entropy and Ecology: A Theory of Abundance, Distribution, and frustules from an ocean sediment core. Proc. Natl. Acad. Sci. U.S.A. 114, 10155–10159
Energetics (Oxford University Press, 2011). (2017).
39. S. J. Phillips, R. P. Anderson, R. E. Schapire, Maximum entropy modeling of species 75. K. Bouma-Gregson, R. M. Kudela, M. E. Power, Widespread anatoxin-a detection in
geographic distributions. Ecol. Modell. 190, 231–259 (2006). benthic cyanobacterial mats throughout a river network. PLoS One 13, e0197669 (2018).
40. S. J. Phillips, M. Dudik, Modeling of species distributions with MaxEnt: New exten- 76. J. K. Carah et al., High time for conservation: Adding the environment to the debate
sions and a comprehensive evaluation. Ecography 31, 161–175 (2008). on marijuana liberalization. Bioscience 65, 822–829 (2015).
41. J. B. S. Haldane, A defense of beanbag genetics. Perspect. Biol. Med. 7, 343–359 77. K. Bouma-Gregson, M. E. Power, M. Bormans, Rise and fall of toxic benthic fresh-
(1964). water cyanobacteria (Anabaena spp.) in the Eel River: Buoyancy and dispersal.
42. N. G. Hairston, F. E. Smith, L. B. Slobodkin, Community structure, population control, Harmful Algae 66, 79–87 (2017).
and competition. Am. Nat. 94, 421–425 (1960). 78. J. Sahu, S. P. Adhikary, Heterotrophic growth and nitrogen fixation in the fila-
43. R. T. Paine, A note on trophic complexity and community stability. Am. Nat. 103, mentous blue-green alga Anabaena sp. Z. Allg. Mikrobiol. 21, 669–676 (1981).
91–93 (1969). 79. M. A. Miller et al., Evidence for a novel marine harmful algal bloom: Cyanotoxin
44. R. T. Paine, Food webs: Linkage, interaction strength and community infrastructure. (microcystin) transfer from land to sea otters. PLoS One 5, e12576 (2010).
J. Anim. Ecol. 49, 666–685 (1980). 80. C. Mann, The Wizard and the Prophet (Vintage, 2019).
45. R. T. Paine, Food-web analysis through field measurement of per capita interaction 81. A. Anthony, Interview: Charles Mann: “The relationship between population and
strength. Nature 355, 73–75 (1992). consumption is not straightforward.” The Guardian, 10 June 2018. https://www.
46. M. E. Power et al., Challenges in the quest for keystones. Bioscience 46, 609–620 theguardian.com/environment/2018/jun/10/charles-mann-book-wizard-prophet-interview.
(1996). Accessed 19 February 2020.

Power PNAS | 7 of 8
Synthetic threads through the web of life https://doi.org/10.1073/pnas.2004833118
 82. S. M. Wisely, O. A. Ryder, R. M. Santymire, J. F. Engelhardt, B. J. Novak, A road map 98. B. A. Hungate et al., Quantitative microbial ecology through stable isotope probing.
for 21st century genetic restoration: Gene pool enrichment of the black-footed Appl. Environ. Microbiol. 81, 7570–7581 (2015).
ferret. J. Hered. 106, 581–592 (2015). 99. D. Kotliar et al., Identifying gene expression programs of cell-type identity and
83. M. J. H. van Oppen, J. K. Oliver, H. M. Putnam, R. D. Gates, Building coral reef re- cellular activity with single-cell RNA-seq. eLife 8, 507 (2019).
silience through assisted evolution. Proc. Natl. Acad. Sci. U.S.A. 112, 2307–2313 100. J. Eberwine, J. Y. Sul, T. Bartfai, J. Kim, The promise of single-cell sequencing. Nat.
(2015).
Methods 11, 25–27 (2014).
84. B. L. Webber, S. Raghu, O. R. Edwards, Opinion: Is CRISPR-based gene drive a bio- 101. B. Koch et al., Estimating taxon-specific population dynamics in diverse microbial
control silver bullet or global conservation threat? Proc. Natl. Acad. Sci. U.S.A. 112,
communities. Ecosphere 9, e02090–e15 (2018).
10565–10567 (2015).
102. J. A. Estes et al., Trophic downgrading of planet Earth. Science 333, 301–306 (2011).
85. R. Sandler, The ethics of genetic engineering and gene drives in conservation.
103. B. J. Meyer, Sex and death: From cell fate specification to dynamic control of
Conserv. Biol. 34, 378–385 (2020).
X-chromosome structure and gene expression. Mol. Biol. Cell 29, 2616–2621 (2018).
86. National Academies of Sciences, Engineering, and Medicine, Forest Health and
104. M. G. Turner, Landscape ecology: The effect of pattern on process. Annu. Rev. Ecol.
Biotechnology: Possibilities and Considerations (The National Academies Press, 2019).
87. W. A. Powell, A. E. Newhouse, V. Coffey, Developing blight tolerant American Syst. 20, 171–197 (1989).
chestnut trees. Cold Spring Harb. Perspect. Biol. 11, a034587 (2019). 105. J. Dekker, E. Heard, Structural and functional diversity of topologically associating
88. G. Popkin, Can genetic engineering bring back the American chestnut? New York domains. FEBS Lett. 589, 2877–2884 (2015).
Times Magazine, 30 April 2020. https://www.nytimes.com/2020/04/30/magazine/ 106. National Academies of Sciences, Engineering, and Medicine, A Vision for NSF Earth
american-chestnut.html. Accessed 10 May 2020. Sciences 2020–2030: Earth in Time (The National Academies Press, 2020).
89. S. A. Zimov, N. S. Zimov, A. N. Tikhonov, F. S. Chapin, III, Mammoth steppe: A high- 107. J.S. Chen et al., CRISPR-Cas12a target binding unleashes indiscriminate single-
productivity phenomenon. Quat. Sci. Rev. 57, 26–45 (2012). stranded DNase activity. Science 360, 436–439 (2018).
90. R. Anderson, Welcome to Pleistocene Park. The Atlantic, April 2017, pp. 1–34. 108. J. S. Gootenberg et al., Nucleic acid detection with CRISPR-Cas13a/C2c2. Science 356,
91. J. Rockström et al., A safe operating space for humanity. Nature 461, 472–475 (2009). 438–442 (2017).
92. P. Bacigalupi, Wind Up Girl (Nightshade Books, San Francisco, CA, 2009). 109. M. Phelps, Increasing eDNA capabilities with CRISPR technology for real-time mon-
93. M. K. Anderson, Tending the Wild (University of Californa Press, Berkeley, CA, 2006). itoring of ecosystem biodiversity. Mol. Ecol. Resour. 19, 1103–1105 (2019).
94. K. Lightfoot, O. Parrish, California Indians and Their Environment (University of
110. M. A. Williams et al., The application of CRISPR-Cas for single species identification
Californa Press, Berkeley, CA, 2009).
from environmental DNA. Mol. Ecol. Resour. 19, 1106–1114 (2019).
95. N. Gilles, Wildfires are essential: The forest service embraces a tribal tradition (2017).
111. B. Heery et al., ColiSense, today’s sample today: A rapid on-site detection of
https://www.yesmagazine.org/issue/science/2017/04/03/wildfires-are-essential-the-forest-
service-embraces-a-tribal-tradition/. Accessed 15 June 2020. β-D-glucuronidase activity in surface water as a surrogate for E. coli. Talanta 148,
96. R. T. Paine, Serengeti Rules (Passion Pictures). https://www.youtube.com/watch? 75–83 (2016).
v=rB72Hy5AGuA. Accessed 15 June 2020. 112. M. A. Williams et al., Comparing CRISPR‐Cas and qPCR eDNA assays for the detection
97. B. LaTour, Love your monsters: Why we must care for our technologies as we do our of Atlantic salmon (Salmo salar L.). Environ. DNA 3, 297–304 (2021).
children. The Breakthrough Journal, 14 February 2012. https://thebreakthrough.org/ 113. D. H. Janzen, How to save tropical biodiversity. Am. Entomol. (Lanham Md.) 37,
journal/issue-2/love-your-monsters. Accessed 15 June 2020. 159–171 (1991).
Downloaded from https://www.pnas.org by 205.178.183.22 on December 25, 2023 from IP address 205.178.183.22.

8 of 8 | PNAS Power
https://doi.org/10.1073/pnas.2004833118 Synthetic threads through the web of life