第1章
The Termite's Secret World: A Journey Into Earth's Hidden Engineers
Have you ever paused to consider that for every human on Earth, there are approximately ten termites quietly reshaping our world? These tiny creatures-often dismissed as mere pests-outweigh humanity and silently engineer ecosystems across the planet. In Lisa Margonelli's "Underbug," we discover that termites represent nothing less than "the poster bug for the twenty-first century," guiding us toward profound insights about ecology, technology, and even our own consciousness. This book has become something of a cult favorite among scientists and technologists, with figures like Elon Musk reportedly keeping a copy on his nightstand. Its influence extends beyond biology into fields like robotics, architecture, and climate science, offering revolutionary perspectives on how simple organisms create complex systems. As climate change accelerates and we search for sustainable solutions, termites-those unassuming architects of negative space-may hold keys to our collective future.
第2章
The Quest for Termite Knowledge: From Arizona to Namibia
My journey into the secret world of termites began in July 2008 on a "Termite Safari" in Arizona. Escaping my depressing research on oil problems, I joined microbial geneticists hunting termite gut microbes that might transform wood into "grassoline." Our guide was Rudi Scheffrahn, a passionate termite expert who could sense termites nearby-his knees literally twitching when colonies were present. He shared fascinating termite facts: queens live 25+ years, some soldiers' mandibles close at 120 mph, and their mounds can reach 30 feet high. Despite 2,800 termite species worldwide, only 28 are actually pests.
In 2007, termites had lost their taxonomic identity, reclassified as "eusocial cockroaches" based on genetic evidence. They evolved from cockroaches 155-250 million years ago when some ingested wood-digesting microbes. Their communal living developed from sharing "woodshake"-a slurry of feces, microbes, and wood chips. As I hunted termites in the Coronado National Forest near the Mexican border, I pondered what constitutes "one" termite-is it the individual, the individual with its gut microbes, or the entire colony functioning as a superorganism?
Despite their ecological importance, termites remain unloved compared to bees and ants. While parents dress children in bee costumes and ants star in movies, termites appear only as crude cartoons on exterminators' vans. Nearly half of all termite research focuses on extermination. They cause billions in property damage annually, eating anything pulpy-even consuming $220,000 in an Indian bank. Yet termites are essential to Earth's ecosystems; without them and their kin, tropical ecosystems would collapse and rivers would silt up.
By the end of our trip, we'd collected 8,000 termites to be frozen, labeled, and genetically sequenced. Standing in the lab that night felt like witnessing the intersection of natural history and an unnatural future, as scientists worldwide try to understand biology's rules to reinvent nature for human use. Little did I know my termite obsession would lead me to spend eight years reading research papers and visiting scientists from California to Namibia, discovering curious creatures and revolutionary ideas along the way.
第3章
Termite Mounds: Nature's Architectural Marvels
Flying to Namibia from the United States takes almost a day. From above, Namibia's landscape appears as a finely patterned carpet. Driving north from Windhoek toward Otjiwarango, I witnessed my first Macrotermes mound-a fifteen-foot dirt spire that looked as if it had erupted from below ground. Soon hundreds appeared, all seemingly pointing northward like dirt fingers or monks in robes, their colors shifting with the surrounding soil from Creamsicle orange to bone to salmon to brown.
Termite colonies begin dramatically on rainy evenings when winged termites (alates) emerge from existing mounds. These fertile termites, unlike their eyeless, wingless colony-mates, have eyes and four translucent wings. Their nuptial flights are brief-most become food for predators. Survivors land, discard their wings with remarkable speed, and begin founding new colonies. The female emits pheromones to attract males. When a male arrives, they scuttle off together to dig a burrow where they'll mate. These termite parents bite off the ends of their antennae, possibly reducing sensory overload to better cope with raising millions of offspring-somewhat like how sleep-deprived human parents manage toddler tantrums.
The queen soon begins laying eggs, inspecting each one "long and carefully." Over time, her abdomen swells enormously, becoming "physogastric"-resembling a soft sausage bound with string-while she lays an egg every three seconds. The king remains by her side while workers tend to her needs.
The termite mound functions as what Eugene Marais called a "composite animal"-a superorganism uniting millions of termites with the dirt structure itself. The hard-packed exterior forms a "skin," internal passageways act as "lungs," the symbiotic fungus serves as a "stomach," and foraging tunnels function as a "mouth." Workers resemble blood cells, soldiers form an immune system, and the queen becomes a captive ovary. This concept of superorganisms, originated by entomologist William Wheeler, temporarily fell out of scientific favor after World War II but has since experienced a revival.
Scott Turner, my guide in Namibia, had spent nearly a decade studying these remarkable structures. In his battered VW van, we drove past hundreds of termite mounds showing signs of his experiments-drilled, cut in half, invaded with homemade scopes, or sprayed with water. At an open shed, I saw something white glowing under the roof-a plaster cast Scott had made by filling a termite mound with plaster of paris and washing away the mud. This eerie white sculpture revealed the mound's internal structure-an inverse record of termite engineering showing large sinuous tunnels surrounded by smaller ones.
第4章
Inside the Superorganism: Termites' Hidden World
One morning, Scott stood in the kitchen mixing fluorescein dye in a Coke bottle until it glowed yellowish green like a Disney movie prop. The next day, a backhoe arrived to split open his dye-injected mound. As the blade came down with a hollow whomp, half the mound fell away, releasing the familiar smell of socks and bread.
The mound's interior revealed a ruined hierarchy-dirt shards, fungus combs, and sculpted mud tumbling down while termites organized themselves from gauzy nets into streams and then rivers. The elaborate scheme of tunnels, chambers, and fungus was jaw-dropping-like the rock drawings made by ancient San people who saw termite mounds as gateways to the spirit world.
The top contained hollow vertical tunnels with smooth interiors that braided together. Termites build by piling dirt and then removing it strategically, using their antennae to feel for smoothness. They're not architects with plans but millions of Martha Stewarts constantly remodeling according to an innate aesthetic. When the mound was breached, worker termites rushed to repair it, each placing dirt balls that quickly accumulated into frilly walls while dark-headed soldiers with scissoring mandibles defended the breach.
Below the mound lived the fungus, digesting grass in hundreds of little chambers. The fungus comb resembled graham cracker pie crust arranged in brain-like folds to maximize surface area, with tiny white balloon-like blooms. This symbiotic relationship between Macrotermes and Termitomyces fungus is so tight it's hard to tell where one ends and the other begins.
Eugene used a pickax to reveal the royal chamber containing the king and queen. While the king was merely large and dark, the queen was monstrous-as big as a finger with a fluid-filled, pulsing body like a toothpaste tube being squeezed. Despite being called a "queen," she's more slave than ruler, captive to her body, children, and the mound she helped create.
Throughout history, humans have projected their social structures onto termites. European naturalists saw monarchies in insect colonies, with Henry Smeathman describing termite mounds to the Royal Society in 1781 as "most closely imitating mankind in provident industry and regular government." Social insects have been used to justify everything from aristocracy and racism to socialism and feminism. The danger in these anthropomorphic views is that they obscure the true nature of these creatures, whose behavior emerges not from central regulation but from simple environmental cues that make individuals work for the whole.
第5章
The Emergence of Collective Intelligence
The traditional view of ant colonies as factories with assembly-line workers has been challenged by scientists like Deborah Gordon, who observed tremendous variation in ant tasks. Rather than having fixed assignments, ants change behavior based on environmental cues and interactions with each other. Gordon suggests we think of ants not as factory workers but as "the firing patterns of neurons in the brain," where simple environmental information guides individuals to work for the whole without central regulation.
Scott began doubting the theory of stigmergy-where termites coordinate building through pheromone-marked dirt balls that guide subsequent workers. While this concept inspired computer science applications like swarm intelligence and agent-based modeling, Scott found it didn't fully explain termite behavior. It couldn't account for why termites deconstructed certain structures or why some tunnels were polished while others remained rough.
Scott sees parallels between mounds and bodies through homeostasis-both create environments suitable for survival. He poses profound questions: "Why do we say the termite is alive but not the mound?" The mound extends the termite organism, creating external memory and boundaries. This challenges gene-centered biology, suggesting that organisms actively shape their environment rather than merely being vessels for genes.
Under black light, termites glow chartreuse and violet as they transfer dyed water mouth-to-mouth in an almost ritualistic exchange. This social interaction involves more than mere water transfer-they groom antennae, stroke flanks, and appear to grovel. While scientists might view termites as neuron-like components, their behavior suggests something more vibrant and social. The joy of termite interaction may be fundamental to their organization, echoing William Wheeler's 1920 observation that "Our ancestors did not start society because they thought they loved one another, but they loved one another because they were so sweet, and society supervened as a necessary and unforeseen by-product."
Termite behavior changes dramatically based on their numbers. Twenty termites in a dish wander aimlessly, but forty termites begin circling together, running faster and faster in a thundering herd. A single termite running counter to the group can even redirect the entire crowd-showing how termite behavior emerges from their interactions rather than individual programming.
第6章
From Termites to Robots: The Engineering Challenge
In spring 2012, I returned to Namibia as Scott's lab hosted Harvard roboticist Radhika Nagpal and her team. The lab transformed into a high-tech space with laser scanners, cameras, and Plexiglas obstacle courses for termites. Radhika, animated and enthusiastic in orange fleece and polka-dotted boots, had just received tenure at Harvard after a career spanning computer science, biology and robotics.
Radhika was less interested in the flashy RoboBee than in understanding the fundamental principles of swarm behavior-what she called the "global-to-local" problem. She observed how human behaviors propagate, like tourists suddenly all taking photos when one begins. The challenge was understanding how global responses emerge from local cues-the same principle behind termite construction, where thousands of individuals with simple rulesets create complex structures without central planning.
The roboticists approached termites as engineering black boxes-systems with unknown internal workings but measurable inputs and outputs. This approach fundamentally differed from Scott's, treating termites not as biological entities but as mechanical systems. When the termites refused to cooperate-working frenetically only when not being recorded-it revealed the challenge of applying engineering principles to biological systems that don't follow predictable patterns.
The cultural clash between biology and engineering became evident as Berry Pinshow complained the roboticists weren't "thinking like termites." While I sorted termites with a paintbrush, entering the meditative state of "bug vision" where minute behaviors became fascinating, the researchers remained detached from the creatures themselves. Radhika confessed that biology's unpredictability had once unsettled her engineering mindset-she'd begun making "bargains" in biology labs and using "blond baby's hair" to cut frog eggs, entering a world where experiments couldn't always be replicated and researchers spoke of "robust frogs."
The roboticists faced endless frustration in their quest to understand self-organizing systems, living in what Justin described as "Waiting for Carnot"-a reference to Beckett's absurdist play and the physicist who established thermodynamics. Complex systems researchers await a unifying theory that could transform technology, warfare, and disaster response. The scientists question their fundamental assumptions about termite behavior, wondering if what appears random might actually be memory or personality.
第7章
The Microbial Universe Inside Termites
When termites eat this book, they'll process tiny particles through an elaborate gut system-first grinding paper in the gizzard, then soaking it in alkaline fluid before passing it through an enteric valve into the hindgut's microbial wonderland. There, 1,400 different bacterial species and exotic protists like the giant whirling Trichonympha break down cellulose into usable sugars. These microbes meticulously arrange themselves in neighborhoods where they can process each other's waste products, with oxygen-friendly species at the edges and anaerobes in the center.
Phil Hugenholtz approached science with an almost psychedelic blend of reason and data, what he called "playing jazz." His work inverted conventional thinking-perhaps the termite is just a delivery vehicle for the gut microbes, not the other way around. This led to bigger questions about who's really in charge, even suggesting our own gut microbes might control our desires and behavior. For Phil, termites were a "Rosetta stone" for understanding biological organization.
At the Joint Genome Institute in California, a digital sign tracked the lab's phenomenal output of DNA sequencing-"the firehose" of genomic data that doubled every seven months. While termites were no longer candidates for biofuel production, they remained inspirational mascots proving cellulose could be broken down. The team continued investigating termite guts through painstaking benchwork, with researchers like fearless Anna and meticulous Falk dissecting frozen termites to extract microliter-sized samples of gut paste containing millions of microorganisms.
When the Deepwater Horizon disaster erupted in April 2010, it made a compelling case for biofuels as clean alternatives to petroleum. Unlike genetically modified crops, synthetic biology had largely avoided controversy, with scientists like Jay Keasling positioning the field as solving global problems. Keasling had engineered E. coli to produce artemisinin precursors, potentially creating affordable malaria treatments for millions. Meanwhile, Craig Venter announced the creation of "Synthia," a cell with synthetic DNA, declaring it "the first self-replicating species whose parent is a computer."
In May 2010, Phil invited me to observe the team's "jazz" sessions-their process for interpreting termite gut data. In a darkened room, researchers merged into a collective intelligence, navigating vast databases of termite gut microbe genes. The process revealed the limitations of metagenomics: of the 470,000 protein-coding genes in the dataset, fewer than 40% had predicted functions.
第8章
The Physics of Life: Modeling Biological Systems
After Phil left for Australia, physicist Hector Garcia Martin fundamentally disagreed with the biologists' observational approach. Where Phil conducted free-associative "jazz sessions" to understand termite gut functions, Hector wanted to make microbial ecology a predictive science based on metabolism-what he called "burning very slowly." He showed a flowchart of termite gut wood breakdown resembling the Tokyo subway system, with hundreds of interconnected reactions.
Hector left Bilbao in Spain's Basque region to study condensed matter physics at the University of Illinois, hoping to "change the world." He saw potential in applying physics and math to biological problems, particularly microbial ecology where both evolution and ecological changes happen rapidly. Unlike studying forests that take lifetimes to mature, microbial systems offered a perfect laboratory for modeling complex systems holistically.
The termite research revealed a deeper multidisciplinary argument about what constitutes life itself. Scientists disagreed about whether genes, individuals, superorganisms, or metabolisms were the relevant units for understanding biology. Hector found biologists frustratingly resistant to prediction, creating a provocative slide comparing a precisely engineered Alfa Romeo with the unpredictable "black box" of sewage treatment plants.
I reflected on how technological solutions to pressing problems often create unforeseen consequences. I compared the potential of synthetic biology to Fritz Haber's invention of nitrogen fertilizer in 1909, which enabled population growth from 1.9 billion to 7.5 billion people but also contributed to environmental problems. The irony wasn't lost on me that we once labored to turn natural gas into fertilizer to grow crops, and now we're working to turn plants into replacements for fossil fuels-suggesting we focus too much on inventions rather than the systems of power surrounding them.
Three years into my termite obsession, I discovered termites had been eating through the beams behind my bed, causing structural damage to my apartment. Rather than being horrified, I found this interesting-evidence of how deeply I'd immersed myself in the termite world. The incident made me realize these insect "vacations" had become more than just an escape from oil and work; they gave me purpose and joy.
第9章
The Rise of Autonomous Swarm Technology
Despite her earlier intentions to abandon termite research, Kirstin kept returning to the unimpressive videos from Namibia. After two years of studying the footage from those simple petri dish experiments, she made a surprising discovery-the termites weren't acting like robots at all. Instead, she found evidence of individual personalities, declaring with scientific certainty: "I'm sure there are key individuals who trigger everything."
Kirstin and her undergraduates spent over 200 hours manually tracking termites through video frames, identifying their behaviors like digging, carrying mud balls, and movements. When manual tracking proved too cumbersome, they developed specialized software using neural networks and Bayesian statistics that could predict termite movements and actions more accurately than humans. This breakthrough meant that after 150 million years of working as an obscure mass, termites could finally be tracked as individuals.
Kirstin's data revealed something surprising: most termites would be fired if they were factory workers. In one dish, only five out of twenty-five termites were building. In another, just two termites did the actual building while nineteen simply ran around. Termites weren't the downtrodden drones of a totalitarian assembly line but more like residents of a Danish socialist village-each contributing in its own way.
Watching the termites revealed distinct personalities. Termite One barely moved but was "clearly smelling the dirt." Termite Three actively gathered and carefully placed dirt balls. Termite Four did absolutely nothing. Others ran around aimlessly or moved balls placed by others. This observation forced a reappraisal: "The informed individuals have a purpose. They have an opinion."
When Kirstin, Justin, and Radhika's TERMES robots made the cover of Science in February 2014, they received significant attention. Media coverage veered toward apocalyptic narratives, with headlines asking "Has the robot apocalypse arrived?" and videos mixing TERMES footage with horror films. The roboticists found these Frankenstein narratives frustrating and uninteresting, preferring discussions about complex systems over simplistic good-versus-evil stories.
The militarization of termite-inspired technology reveals uncomfortable parallels between human technological development and termite superorganisms. The military's fascination with insect-inspired technology has evolved into concepts for "single-serving death bots"-miniature drones carrying shaped charges capable of puncturing human skulls. Defense strategists envision using 3D printers to produce billions of these drones at a dollar apiece, creating "smart clouds" that could "flood" areas to find enemies.
第10章
Termites as Ecosystem Engineers
After years studying termites as bugs, robots, and superorganisms, I journeyed to Australia to understand how termites' evolutionary fate intertwines with our world. In 2011, Phil (now leading an ecogenomics lab at the University of Queensland) invited me on a termite safari to Darwin, Australia, where we hoped to find Mastotermes darwiniensis-the "mastodon of termites"-and its legendary gut protist Mixotricha paradoxa, a "composite animal" with five separate genomes collaborating as one.
Within minutes of arriving in Darwin, we discovered Mastotermes darwiniensis infesting Uncle David's mahogany tree. These termites, larger than any I had seen before, were "the size and color of a grain of cooked white rice" with yellowish heads and translucent bodies. Once found globally (even gnawing dinosaur bones), climate change had made them extinct everywhere except Australia. In northern Australia, these "Mastos" had achieved folk antihero status, respected for their opposition to human settlement, with locals claiming they could be heard chewing through walls at night.
In Fly Creek's eucalyptus forest, we collected eight different termite species with the help of local experts. Each mound revealed unique architectural wonders-from small conical eruptions housing tiny Microcerotermes with snapping mandibles, to knee-high Buddha-shaped mounds containing Coptotermes (which had rafted from Asia 13 million years ago), to the massive "magnetic" Amitermes mounds perfectly aligned to regulate temperature, and finally to cathedral-like structures over twenty feet tall built by minuscule Nasutitermes triodiae. Most surprising was learning that 80% of local eucalyptus trees were hollow, eaten by termites from within, changing how the forest burns in this fire-shaped ecosystem.
Unable to truly vacation, I continued obsessing over termites in an Australian hotel room. Reading scientific papers, I discovered studies showing termites had rehabilitated over eleven square miles of land near a bauxite mine over twenty-six years. Another study revealed farmers could increase wheat yields by 36 percent by allowing termites and ants to work the soil, potentially eliminating expensive nitrogen fertilizers. As climate change transforms landscapes, termites could become crucial ecosystem engineers, helping grow more food on less land while restoring damaged environments.
第11章
Patterns in the Landscape: Termites as Earth's Hidden Architects
After returning from Australia, I wondered how termites rehabilitated Dieter's land beyond simply fertilizing soil and recycling dead grasses. Searching for researchers studying termites' landscape impacts led me to mathematician Corina Tarnita and ecologist Rob Pringle at Princeton. Years earlier, Rob had discovered that termite mounds in Kenya created hotspots of biodiversity, with significantly more geckos, plants, and nutrients than surrounding areas. The termites were somehow "pulling the strings from below" to organize entire landscapes into patterns of fertility.
During fieldwork in Kenya, Corina had an epiphany while standing on a Range Rover overlooking burned grassland. She spotted two interacting patterns: the polka-dot pattern of termite mounds and a leopard-spot pattern in the vegetation between them. This second pattern resembled a Turing pattern-a mathematical concept explaining how natural systems self-organize into spots, stripes and other formations.
Working with other scientists, Corina built mathematical models combining termite mound organization with scale-dependent feedback patterns in grass growth. When combined, these models produced images resembling African patterned cloth that matched satellite images of actual termite landscapes. The models revealed how local and global patterns interacted across scales, explaining the Persian rug-like patterns I had observed from planes in Namibia and Australia.
The most significant discovery came when Corina adjusted rainfall parameters in the model: grass associated with termite mounds could survive on remarkably little water. While previous models predicted dry landscapes would catastrophically collapse into deserts when rainfall decreased, landscapes with termite mounds showed much greater resilience. With termites present, grass would first reappear on mounds in distinctive patterns, making termites crucial buffers against desertification across the world's drylands-about 40 percent of Earth's land surface.
Building the model proved extraordinarily challenging as it forced the team to create rules for everything, leaving no blank spaces in their understanding. In 2017, the team published groundbreaking work modeling how burrowing animals like termites, ants and rodents interact with grasses to create vast patterns across tens of thousands of square miles worldwide. These patterns revealed a planet reshaped not by a mastermind but by trillions of mini-minds working in concert, creating order visible only from above.
The mound reveals a nested hierarchy of worlds: from the termite's gut processing grass, to termites digging and grooming, to termites and fungus communicating through chemistry, to plants and geckos on the surface, to giraffes obliviously munching leaves, to the regular carpet of fertility visible from above, and finally to a planet with an atmosphere. Like giraffes, humans remain largely ignorant of these interconnected worlds we cannot see, anthropomorphizing complex relationships into simplistic concepts.
Returning to Namibia in 2014, I reflected on how my obsession with termite mounds had transformed my worldview. The mounds had become everything that mattered-the meaning of life, keys to the future, parables about science and narrative. I saw them simultaneously as the dirt's impression of the solar system and as building blocks of terrestrial fertility. The mounds represented both the complexity of dumbness and the dumbness of complexity, a hopeful beacon in a rapidly changing biosphere.
