Six Thousand Crosses

First Cycle: November

She arrived at the station in the dark. The bus from Obregon had broken down twice on the highway and the driver had fixed it both times with a wrench and a length of wire and the second time the passengers had stood along the shoulder in the desert wind while he worked beneath the chassis by the light of his phone. The station was a compound of concrete buildings behind a chain-link fence at the edge of the irrigated land. Beyond the fence the Sonoran desert went out flat to the mountains.

Her name was Lena Arriaga and she was forty-three years old and she had been breeding wheat for nineteen years. She carried a duffel bag and a laptop case and a cardboard box of seed packets labeled in her own handwriting, each packet sealed with tape and dated. The guard at the gate checked her credentials and waved her through. A dog watched her from beneath a pickup truck. She found her assigned room and set down her bags and stood at the window.

The fields were dark. Twelve hundred hectares of experimental plots stretching south toward the irrigation canal, the wheat just planted, the soil still showing between the rows. In three months those rows would be dense with tillers and she would begin the crosses. Six thousand minimum. Each one done by hand.

She had been at CIMMYT Obregon before, years ago, as a graduate student. She remembered the heat. The stools. The paper bags. The way your back seized up after four hours of leaning over wheat plants. The work had not changed. The pathogen had.

The briefing was at seven in the morning in a room with fluorescent lights and a projector that took five minutes to warm up. Dr. Kwame Asante ran the station. He was a Ghanaian geneticist in his sixties who had been breeding rust-resistant wheat since before Lena was born. He stood at the front of the room with his hands behind his back and spoke without notes.

The situation, he said. He clicked to a satellite image. The red overlay showed the progression of the new race — designated TTKTT-7, though everyone called it Seven — across East Africa, through the Arabian Peninsula, into Pakistan and northwestern India. The red was dense and continuous across the Punjab. In six months it had crossed from Kenya to Rajasthan.

Asante clicked again. Yield losses. Kenya: forty-two percent. Ethiopia: fifty-eight. Pakistan: the numbers were still coming in but the FAO preliminary was thirty to thirty-five percent of the winter wheat crop.

He clicked again. A graph showing the genetic distance between Seven and its predecessor strains. Seven had acquired virulence genes that defeated every major resistance gene deployed in commercial wheat. Sr31, Sr38, Sr24 — all broken. The gene stack that CIMMYT had spent two decades assembling was compromised.

We have eighteen months, Asante said. Two planting cycles here, one summer cycle at Toluca. Our mandate is to identify new resistance sources and cross them into adapted backgrounds. He paused. We have done this before. We will do it again.

He said this without conviction or its absence. He said it the way you say a fact.

The first weeks were preparation. Lena and four technicians — Rosario, Pilar, Ana, and a young man named Tomas who had just finished his degree in Hermosillo — organized the crossing blocks. Two hundred parent lines planted in paired rows. Susceptible testers planted in borders and alleys to serve as spreader rows when the time came to inoculate.

Lena walked the rows each morning. The wheat was in early tillering. Pale green shoots pushing out of the cracked soil, the irrigation water darkening the furrows, the Sonoran sun already hot at eight oclock. She carried a clipboard and a hand lens and she checked each plot for uniformity, for vigor, for the first signs of anything wrong.

Nothing was wrong yet. You prepared for a catastrophe that had not arrived, in a field where the catastrophe could not yet reach, using tools that the catastrophe had already rendered obsolete. You did this because the alternative was to do nothing.

At night she opened the genome browser on her laptop. The annotated wheat reference genome — twenty-one chromosomes, sixteen billion base pairs, five times the size of the human genome — scrolled across her screen in colored blocks. She was looking for resistance candidates. Genes from wild wheat relatives, from Aegilops tauschii and Thinopyrum ponticum, species that had been fighting rust for millennia before humans domesticated anything. The genes existed. The problem was getting them into a plant that also yielded well, milled properly, and matured in one hundred and twenty days.

In December the crossing began. Each morning Lena and the technicians carried their stools and paper bags and tweezers to the crossing blocks and sat down among the wheat. The procedure was the same as it had been in 1945 when Norman Borlaug had first sat in these fields: select a head that has not yet opened its florets, clip the awns, open each floret with tweezers, remove the three anthers before they shed pollen, bag the emasculated head, wait two days, then apply pollen from the selected male parent. Repeat six thousand times.

Rosario was the fastest. She could emasculate a head in forty seconds — open, clip, strip, bag — her fingers moving with a precision that Lena had never matched. Rosario had been doing this for twenty-seven years. She did not read journal articles or attend conferences. She knew wheat the way a surgeon knows a particular joint: by touch, by repetition, by the accumulated intelligence of her hands.

They worked in rows, moving down the block on their low stools, the sun on their necks, the paper bags rustling in the wind. Each bag was labeled with the cross number, the date, the initials of the technician. LxR-26-1201-RA. LxR-26-1202-RA. The numbers climbed.

In the afternoons Lena pollinated the crosses from two days prior. She removed the paper bags, checked that no foreign pollen had contaminated the emasculated heads, and brushed the selected male pollen across the exposed stigmas with a camel-hair brush. The pollen was a fine yellow dust. When the transfer was complete she rebagged the head and wrote the date and the male parent number on the bag. She did this three hundred times a day.

Some days the wind was bad. A north wind off the desert that rattled the bags and dried the stigmas before she could pollinate. On those days she threw away one cross in five and started again. Some days the heat killed the pollen within minutes of collection and she had to harvest fresh anthers between every third or fourth pollination. Some days everything worked and she moved down the rows in a steady rhythm, bag after bag, and the numbers on her clipboard climbed toward the day’s target.

Tomas asked her once why they did not automate the crossing. There were robotic systems now, he said. Machines that could strip anthers and apply pollen faster than any human hand.

She told him the machines existed but they cost two million dollars and they could not distinguish between a floret that was ready and one that had already shed its pollen. Besides, she said. The machines break. Tweezers dont break.

Rosario laughed at this without looking up from her stool.

The reports from South Asia worsened through December and January. Seven had crossed into Iran. Wheat prices on the Chicago Board of Trade broke records three times in two weeks. The Indian government imposed export restrictions. Pakistan followed. Egypt, which imported more wheat than any country on earth, began rationing.

Lena watched the satellite overlays each evening. The red crept westward. The models predicted Seven would reach Turkey by spring. If it established in the Black Sea wheat belt — Ukraine, southern Russia — the global supply chain would not bend. It would break.

She did not share the models with the technicians. They knew. The news was on their phones, in the rising price of flour at the market in Obregon, in the longer lines at the government tortilla distribution points.

Lena called her mother in Guadalajara. Her mother asked when she was coming home. Lena said she did not know. Her mother said the price of tortillas had doubled since October. Lena said she knew. Her mother said, Is what you’re doing going to fix it. Lena said, I dont know. Maybe. Eventually. Her mother was quiet for a moment and then said, Well. Keep going then.

In February she began the disease screening. The spreader rows had been inoculated with a cocktail of rust races — not Seven itself, which was too dangerous to import, but its closest available analogs. She walked the nursery with her hand lens three times a day, looking for the first pustules.

They appeared on a Tuesday. Small raised blisters on the stem of a susceptible check plot near the south border, each one splitting open to reveal a mass of bright orange spores. The color was startling against the green. She touched one with her finger and it left an orange smear on her skin that would not wash off for days.

Under the lens the pustule was a landscape. The epidermis of the wheat stem had ruptured and the fungal tissue beneath was producing urediniospores by the million — each one a single cell, football-shaped, bright orange, light enough to ride the wind for hundreds of kilometers. A single pustule could produce a hundred thousand spores in a day. A single infected field could fill the air with an orange haze visible from the road.

This was what Puccinia graminis had been doing for sixty million years, long before wheat existed, long before grasses had evolved into the form that humans would eventually domesticate and depend upon for thirty percent of their calories. The fungus was not new. Its strategy was not new. Only its latest iteration — this particular arrangement of virulence genes that happened to defeat this particular arrangement of resistance genes — was new. And even that novelty was, from the pathogen’s perspective, routine. It had broken resistance genes before. It would break them again. It would keep breaking them for as long as there were hosts to infect and spores to carry, which was to say for as long as there was wind and rain and sunlight.

The wheat stood in its rows and grew and filled its grain and waited.

By late February the susceptible plots were orange. The resistant parents — her candidates — stood green among the wreckage. She scored each one on a zero-to-nine scale. Infection type, severity, response. The numbers went into her laptop. The laptop connected to a database in Texcoco where other breeders at other stations were entering their own numbers, their own scores, their own small victories and comprehensive defeats.

Forty-seven of her two hundred parent lines showed resistance to the analog cocktail. She did not know if any of them would resist Seven itself. The analog was close but close meant almost nothing in genetics where a single nucleotide change could turn immunity into susceptibility.

She made her selections. She planned the second round of crosses.

She packed the harvested seed — each cross in its own envelope, each envelope in a box, each box labeled and inventoried — and prepared to ship them to Toluca for the summer cycle.

The first cycle was over. She had completed 6,214 crosses. The rust had crossed into southeastern Turkey.

Second Cycle: May

Toluca sat at 2,640 meters above sea level in a valley west of Mexico City. The air was thin and cool and the afternoon rains came like clockwork. The fields were green in a way the Obregon fields never were — dark green, almost blue, the wheat dense and tall in the highland soil.

Lena planted the F1 generation from her best crosses and began again. New parents. New combinations. The segregating populations from the first cycle growing alongside the new plantings, each one a question she had asked the genome six months ago and was now waiting to hear answered.

The work was the same. Stools, tweezers, bags. Rosario had come up from Obregon. Pilar and Ana had not — budget cuts. Asante had lost a third of his operational funding when the EU redirected agricultural research money to emergency food aid. The world was running out of bread so the money that might produce more bread was spent buying bread that was running out.

Tomas had come. He worked quietly and quickly. He had a good eye for disease symptoms — better than Lena’s, if she was honest. He could spot a nascent pustule on a flag leaf at three meters while she was still fumbling with her lens. She had begun teaching him the crossing technique. His hands were large for the work but he compensated with care. He would be a good breeder if there was still a program to breed for.

The reports continued. Seven in Turkey, established and spreading. The Black Sea wheat belt under threat. Russia announced a complete export ban. Futures hit four hundred dollars a metric ton, then five hundred. In Cairo there were riots. In Lahore the army distributed flour from military stockpiles.

She closed the laptop. She went to the field.

The Toluca rainy season was different from anything in Obregon. By two oclock most afternoons the clouds built over the western mountains and by three they broke and the rain came down in sheets that turned the paths between plots into brown rivers. They would retreat to the concrete block that served as the Toluca laboratory — a single room with a steel bench and a dissecting microscope and a window that faced the field — and wait for the rain to stop.

During the rains she taught Tomas to read the disease scores. She showed him how to distinguish an infection type 2 from an infection type 3 — the size of the pustule, the presence or absence of chlorosis, the necrotic flecking that indicated a hypersensitive response. She gave him the modified Cobb scale and had him practice on dried specimens she kept in a box: pressed stems mounted on cardboard, the pustules preserved under clear tape, each one labeled with its score.

He learned fast. He had the patience for it, which was the quality that mattered most. Breeding was not about intelligence. Intelligence was common. Breeding was about the willingness to look at the same thing ten thousand times and notice the one time it was different.

She thought sometimes about what would happen to Tomas if the program lost its funding. He would go back to Hermosillo. He would teach, perhaps. Or work for a seed company, crossing maize hybrids for yield and herbicide tolerance. Nobody got rich breeding wheat. You bred wheat because three billion people ate it.

In July the Toluca screening plots showed their results. She had inoculated with fresh isolates — closer to Seven now, collected from Turkish fields by a colleague who had carried the samples out in a diplomatic pouch because normal channels had broken down.

The results were worse than Obregon. The Turkish isolates were more aggressive. Lena watched her resistant parents fall one by one over a week. Plots clean on Monday showed heavy infection by Friday. The orange appeared on stems, on leaves, on the sheaths. It spread down the rows like a rumor.

Of forty-seven lines that had resisted the Obregon analogs, eleven resisted the Turkish isolates. Of those eleven, she trusted four. The others had shown partial resistance — reduced severity, smaller pustules, slower spread — which meant the genes were buying time, not solving the problem. Partial resistance where every farmer planted the same variety and the wind carried spores across borders was delay, not defense.

She sat on her stool in the Toluca field with her clipboard on her knees and looked at the numbers. Four lines. She would not know if any of them truly resisted Seven until she could test against Seven itself, which she could not do because the containment protocols forbade it. She was breeding against an enemy she had never met, using proxies and accumulated intuition.

The genome browser showed her candidate genes. Sr62 from Aegilops sharonensis. A novel QTL on chromosome 6A from a wild emmer accession. Two uncharacterized loci that had shown up in association mapping but had never been validated in the field. She was building a house in the dark.

She made her crosses. Fewer this time. Targeted. The four best resistant parents crossed to the best-adapted local varieties. She worked faster than in Obregon. Time was narrowing. She could feel it in the reports, in the price of a kilo of flour in the Toluca market, in the way Asante’s emails had gotten shorter, the formalities stripped away.

The harvested seed went into boxes. The boxes went into a truck. The truck drove south to Obregon for the winter cycle.

Before they left Toluca, Lena walked the field one last time. The afternoon rain had passed and the air was clean and cold and the mountains to the west were sharp against a sky the color of wet slate. She bent down and pulled an infected stem from a susceptible check plot. The pustules were mature — dark orange, almost brown, the spore masses dry and powdery. She rolled the stem between her fingers and the spores came off on her skin. Under a microscope each grain of that powder was a single cell containing the genetic information necessary to locate a wheat stomata, germinate, penetrate the host tissue, establish a feeding structure, redirect the plant’s nutrients to its own reproduction, and produce ten thousand more spores identical to itself. All of this encoded in a genome smaller than the wheat genome by a factor of forty. The rust did more with less. It had always done more with less.

She wiped her fingers on her jeans and walked back to the van.

Third Cycle: November Again

Back in the desert. The same compound, the same fence, the same dog beneath the same truck or its successor. Seven had reached the Mediterranean basin. Southern Italy, Greece, the Levant. The FAO declared a Level 3 emergency. The word famine appeared in official documents for the first time in decades.

Lena planted her F2 populations. Segregating, variable, each plant a different combination of its parents’ genes. Some would carry the resistance. Most would not. She was buying six thousand tickets and the odds were what the odds always were.

Rosario was there. Tomas was there. The crew was smaller. Pilar had taken a job with a private seed company. Ana had gone home to Oaxaca — her father was sick and the station could not match the salary the seed company had offered her. Lena did not blame them.

She and Rosario and Tomas did the work of five. They started earlier. They stayed later. Rosario’s hands were beginning to show the damage — the knuckles swollen, the joints stiff in the morning cold, a tremor in the right hand that she compensated for by bracing her wrist against her knee. She said nothing about it. Lena said nothing about it.

She walked the rows at dawn. The plants emerged. They tillered. They grew. She watched them the way you watch someone you love in a hospital — checking for signs, interpreting every change, knowing that your watching changed nothing.

In January she inoculated. Not with analogs this time. The Turkish colleague had come himself, carrying sealed vials of Seven in a nitrogen dewar, and together they set up a containment protocol in the far corner of the nursery, triple-bordered with susceptible spreader rows, a hundred meters from the nearest experimental plot.

She released the spores on a Tuesday morning. They wore masks and gloves though the masks were for protocol — the spores could not infect humans. They dusted the spreader rows and walked away.

The wind did the rest.

On the third day she saw the first pustule on a lower leaf of a susceptible check. By the fifth day there were hundreds. By the seventh day the spreader rows were orange, the stems furred with spore masses that caught the morning light and turned it the color of iron oxide. The orange was beautiful if you did not know what it meant.

Within two weeks the color had migrated to the experimental plots. She scored her populations each morning, moving down the rows, her clipboard filling with numbers that were mostly bad.

Of six thousand F2 plants, she identified one hundred and twelve that showed resistance. She scored them again. Eighty-nine held. She scored them a third time. Seventy-one.

Seventy-one plants in a world that grew seven hundred and eighty million metric tons of wheat a year and was about to grow considerably less.

She selected the best forty. She bagged their heads individually. She would harvest the seed from each one, grow them out, test the progeny, select again. The process would take two more cycles — another year at minimum. She did not have another year. But the work was measured in generations, not in years, and a generation was one planting and one harvest and one selection and one planting again.

In March the Obregon wheat was mature. She walked the plots for the last time. The resistant plants stood green among the dead and dying. The infected plots were broken stems and shriveled grain, the orange pustules cracked and dark, the spores spent. It looked like a war had passed through. It looked like the aftermath of something that did not care what it destroyed because destruction was not the point. Reproduction was the point.

She harvested the forty selections. Each head clipped with scissors, placed in a paper bag, labeled. Rosario helped. They did not speak.

Tomas loaded the boxes into the truck.

Where are these going, he said.

Lena said, Toluca. And some to El Batan. And the best forty go to Njoro.

Njoro was in Kenya. The original screening site where Seven had first been identified. Sending resistant material back to the place where the pathogen originated was either strategic or ironic and she had stopped trying to decide which.

She also packed a separate box. Twenty lines — her top selections from all three cycles, each packet containing enough seed for a small plot trial. She labeled this box by hand:

Dr. Sofia Reyes CIMMYT Highland Station Km 42, Carretera Toluca-Atlacomulco CP 50740, Toluca, Estado de Mexico

Sofia had been her student eight years ago. She ran the highland station now, where the thin air and cool temperatures slowed the rust’s reproductive cycle. Not stopped it. Slowed it. A field at altitude bought you two weeks before the pustules appeared. In breeding, two weeks was a generation. A generation was a chance.

She wrote a note on a piece of station letterhead and folded it into the box:

20 lines. Best F2:3 from TTKTT-7 screen, Obregon winter cycle. Sr62 + 6A QTL + 2 unknown. Heterogeneous — select hard. Crossing data attached.

She looked at the note. She considered adding something personal. She did not add anything. She folded the data sheets — crossing numbers, dates, inoculation records, severity scores, fifty pages of numbers that represented eighteen months and eighteen thousand crosses performed by hand — and placed them in the box on top of the seed packets.

Tomas sealed the box with packing tape.

That evening she sat in her room and opened the genome browser one last time. She looked at the chromosome 6A region where her best QTL mapped. The gene — if it was a gene, if it was one gene and not several, if the association was real and not an artifact of population structure — sat in a stretch of DNA that had been conserved across wheat species for approximately four million years. Long before humans. Long before agriculture. The wild grasses that grew on the hillsides of the Fertile Crescent had carried this sequence, or something like it, through drought and fire and ice and the slow creep of continents. She was not the first to need it. She was only the latest.

The rust had its own history, equally old. Puccinia graminis had diversified alongside the grasses, tracking its host through every evolutionary turn, every defense. When the grass evolved a resistance gene, the rust evolved a virulence gene to match. When the grass stacked resistances, the rust stacked virulences. The arms race had no beginning that science could identify and no end that anyone could predict. It was older than the Himalayas. It would continue long after the last wheat breeder had put down her tweezers and gone home.

She closed the laptop. The screen went dark and the room was dark and outside the Sonoran desert was dark under a sky of hard white stars.

The next morning she watched the truck leave with the boxes. Rosario stood beside her. They watched until the truck was gone.

You going home, Rosario said.

No.

Rosario looked at her. What then.

Lena did not answer right away. The sun was up and the desert was pale and flat to the horizon. She thought about the forty selections in their paper bags on the truck. Seventy-one plants out of six thousand that had resisted Seven, narrowed to forty, narrowed further by the time Sofia grew them out and scored them and selected again. The funnel of breeding: wide at the top, narrow at the bottom, most of what you started with lost along the way.

She thought about what Asante had said in the first briefing. We have done this before. We will do it again. And Borlaug before him, sitting on a stool in this same field in 1954, watching Race 15B destroy his resistant lines, starting over. And before Borlaug, the nameless farmers of the Fertile Crescent who had noticed that some plants survived the red dust and had saved their seeds. Ten thousand years of saving seeds.

I’m going up, she said.

To Toluca.

To the highland station. I want to be there when Sofia plants the trial.

Rosario nodded. She did not ask why. She had been doing this work for twenty-seven years.

She took the bus from Obregon to Mexico City and then a van to Toluca and then another van up the winding road to the highland station. The desert fell away and the hills were green with pine and oak and the air was thin and cold and smelled of woodsmoke. Clouds sat in the valleys like something poured.

Sofia met her at the gate. She wore a fleece jacket and rubber boots and her hair was tied back with a piece of twine. She held the box.

Twenty lines, Sofia said.

Twenty.

The data sheets are thorough.

Lena nodded. They stood at the gate for a moment. Eight years since Sofia had finished her doctorate, since she had sat on a stool in the Obregon field and learned to strip anthers under Lena’s supervision, since she had gone north to a postdoc and then come back to run this station at twenty-six hundred meters. The last email between them had been about seed shipments. The one before that about a conference neither attended. The relationship had narrowed to its professional minimum, which was maybe the only width that could survive the pressure of the work.

You look tired, Sofia said.

I am tired.

The twenty lines. Are they good.

They’re what I have, she said.

Sofia nodded. In breeding, what you had was always what you started with.

They walked to the field. The highland plots were freshly tilled, the dark soil turned in long rows, the irrigation channels cut and waiting. The mountains rose on three sides. Sofia had already prepared the planting plan. Twenty lines in replicated strips, with susceptible checks and border rows. Southern exposure, good drainage, close to the weather station. At this altitude the wheat would grow slowly — a hundred and forty days to maturity instead of a hundred and twenty. But the rust would grow slowly too.

They planted that afternoon. Lena and Sofia and two field workers whose names Lena learned and forgot and then learned again. A hand planter, the cone pushed into the soil, the seed dropped, the soil pressed closed with the heel of a boot. Row by row. The seed was small and pale and hard. Each one contained an embryo, a starch reserve, and whatever genes Lena had managed to assemble across three cycles and eighteen months and eighteen thousand crosses. The field would sort them. The rust would sort them. Her job was over. She had put the question into the ground. The sun moved across the sky. The shadows of the mountains crept across the valley.

They finished at dusk. The planted rows were invisible — dark soil in a dark field under a darkening sky. You could not see the seeds. You could not see anything that had been done.

Sofia asked if she wanted dinner. Lena said yes. They ate rice and beans and tortillas and talked about the planting plan and the scoring schedule and the inoculation protocol and did not talk about the bread lines in Cairo or the army distributions in Lahore or the seven hundred and eighty million metric tons that were now projected to be five hundred and sixty million.

After dinner Lena went outside. The highland air was cold and the stars were hard and white overhead. Somewhere below her, at sea level, in the Sonoran desert, in the Punjab, in the Nile Delta, the rust was doing what it had always done — producing spores, riding wind, finding hosts, feeding, reproducing. It did not rest. It did not know what wheat was or what humans were or what bread was. It found cells to infect and it infected them.

She looked at her hands. The nails were cracked and dark with soil. Between her fingers, in the lines of her knuckles, a faint orange stain that might have been from the pustules she had scored in the Obregon nursery weeks ago, or from the red highland earth she had pressed seeds into that afternoon. She could not tell which.