Dr. H. Garrison Wilkes

H. Garrison Wilkes graduated cum laude from Pomona College in 1959 and earned his Ph.D. at Harvard University in 1966. He began his academic career as an assistant professor at Tulane University, then moved to the University of Massachusetts Boston in 1970, where he rose to associate professor in 1974 and professor in 1984 — teaching there for the rest of his career.

Alongside his teaching, he served on the Rockefeller Foundation's maize germplasm committee for two decades (1970–1990), on the National Research Council/National Academy of Sciences life sciences assembly (1973–1977), and on a national committee on agricultural genetic resources policy in Washington (1987–1991) — the institutional side of an argument he made everywhere he went: that the wild relatives of our crops are irreplaceable, and that losing them is a kind of extinction we can still choose to prevent.

His 1967 doctoral thesis settled a long-standing taxonomic question: through systematic collection across Mexico and Guatemala and study of morphology and chromosome knobs, he confirmed the division of genus Zea into annual and perennial teosinte and identified the distinct races of annual teosinte still used today.

In 2011, the ASA/CSSA/SSSA convention in San Antonio held a special session, "The Mysteries of Maize," honoring the researchers who shaped the field — among them Wilkes, alongside Major Goodman, Hugh Iltis, Ángel Kato Yamakake, Wilfredo Salhuana, Jesús Sánchez González, and CIMMYT's own Suketoshi Taba and José Crossa. He opened the session with a talk on the in-situ conservation of teosinte and its co-evolution with maize — fieldwork he had, by then, been doing for decades.

In 2022, Wilkes began donating his original 1960s–70s teosinte seed collections — over a dozen boxes, packed and shipped from his home workshop in Massachusetts — to the USDA-ARS North Central Region Plant Introduction Station in Ames, Iowa. Over his career he contributed 49 accessions to that collection, cataloged in 2023 as "The Wilkes Legacy Collection" by researchers Vivian Bernau, Denise Costich, and Wilkes himself.

Delivered on receiving the Meyer Medal in 2020, illustrated with slides from six decades of fieldwork in Mexico — teosinte and its wild relative Tripsacum, maize–teosinte hybrids, archaeological corn from Tehuacán and Mitla, and the CIMMYT gene bank where much of that seed now lives. Transcribed and lightly corrected from the original recording.

I'm Garrison Wilkes, and I am the Meyer Medal lecturer for this year, 2020. I'm old enough to actually have experienced the source of "plant genetic resources." This phrase was coined in Rome at FAO in a night meeting with Sir Otto Frankel, Jack Hawkes, Jack Harlan, Erna Bennett, Bix Sikes, myself, and several others, because Sir Otto believed that plant conservation needed a tag. And the tag that people settled on that night was "plant genetic resources." Plant genetic resources are the genetics selected by farmers from centuries past to the present. They experience three dynamics. One is genetic erosion in farmers' fields. The second is genetic vulnerability, which we as humans are experiencing right now. And the third is extinction. Now, I'm going to use maize as the example for my talk, because of personal experience. When I returned to Cambridge after a year and a half in Mexico, I wrote the National Seed Storage Laboratory and the director, Sam Bass, that I had kilo amounts of teosinte seed, which I would be very willing to send to Fort Collins. And Sam Bass replied to me: "This gene bank does not store wild plants. Thank you very much for your inquiry." And I've spent the next 20 years trying to convince people that the wild relatives of corn — plants like teosinte that are fully fertile with corn — are part of our plant genetic resources. Then I had the experience in the '80s of assisting CIMMYT with rebooting the gene bank. And third, I have a role in the Tripsacum Garden at Tlaltizapán, one of the experimental stations of CIMMYT.

This obviously is the maize plant, with the polystichous central spike. And the central spike of corn distinguishes corn from teosinte, because the central spike is polystichous. The lateral branches of that tassel are distichous, and teosinte is a two-ranked structure, which is distichous. Also, teosinte has a massive lower glume, which encloses a single seed, whereas corn has naked, paired spikelets in the cupule. And lastly, corn has a papery lower glume, and the teosinte lower glume is very massive.

These are two teosinte tassels, and you can see that there's no polystichous central spike.

These are tassels — on the right are teosinte tassels, and on the left is a corn tassel, and you can see the obvious central spike. The six tassels in between the two are maize–teosinte hybrids and backcrosses, so going through the field it's very easy to spot the hybrids because of their minute, central-spike tassel.

Another attribute of teosinte is the branching, and you can see how this is a sort of candelabra, with multiple tassels at the end of lateral branches.

This is the rachis tissue of individual spikes at a node for teosinte. It takes a couple of hours to remove all the husks that surround each spike, so this is a couple of hours' labor to expose all the teosinte spikes. You have to be very careful, because the rachis tissue disarticulates, and then you have broken spikes. A perfect example like this is rare indeed, and there aren't many pictures of it in the literature.

Teosinte produces, on a very luxuriant plant, up to 2,000 individual seeds. Three of these seeds are from a Flannery dig in Mitla, Oaxaca, and they're 3,000 years old. The only one that is recent teosinte is in the lower right-hand cluster, and it's there for scale and morphology. In the upper left you'll see rachis segments of teosinte with the lower glume fully expanded, being pushed out — typical of the maize-luxuriance effect on the endosperm in a maize-contaminated teosinte. Below is maize, and you'll notice how large the embryo is compared to the endosperm, and the upper right is rachis segments identified — as clearly as one can say "pure" — as teosinte. It's always a little maize-contaminated, but pure teosinte, and this site is very important because it had teosinte, evidence of hybridization with maize, and maize. This is in Mitla, Oaxaca.

This is Tripsacum, and there are about 30 species of Tripsacum in the Americas. The country with the largest diversity of Tripsacum is Mexico, but it extends down into South America. You'll see here at the base of the inflorescence the rachis segments, very much like teosinte, but above the flowers are the male pollen — so the spike has both rachis segments and pollen-producing florets.

This is Tripsacum maizar, and it looks like it has a tassel, but these are florets — spikes with seed-bearing rachis segments below and pollen-producing florets. At the top — this plant is about five meters high. It's a very robust Tripsacum, and it's also a 36-chromosome Tripsacum.

This is Tripsacum dactyloides, which is found in the eastern U.S.

Tripsacum has chromosome numbers of 36, 54, and 72. The Tripsacum Garden at Tlaltizapán, Morelos, Mexico, is one of the stations of CIMMYT.

This is a maize field at harvest time, and the corn is put in stacks to dry — you'll notice that teosinte plants aren't cut and put in the stack. But when I work in these fields, I'll often find clusters of F1 hybrids and backcrosses, located in a small area — then there won't be another for 20 meters or so, and that's because the hybrids dried in the stack. They were cut, recognized as corn, and put into the stack, and their seed dribbled out onto the ground for the following year. Teosinte has 20 chromosomes and is fully fertile with corn. Tripsacum can be crossed with corn, but it's a labor of love to do it. But with teosinte, no problem.

This is an example from November, seed-harvesting time for teosinte, and you'll see it growing in an acacia tree or shrub, ungrazed.

This slide of teosinte, dry in late October/November, is here because it shows limestone rock. At 1,500 meters in Mexico, when I found limestone outcroppings, I would carefully look for teosinte, because it's often found on limestone soils at that elevation.

This is the Beadle mutation hunt, and we collected well over 50 kilos of teosinte seed with the help of some macheteros.

This is George Beadle in the field. This seed was given PI numbers, and over 50 kilos is deposited in the bank at Ames, Iowa. When we left the field that day, George — who had been a Nobel laureate and president of the University of Chicago — said, "You know, I think this is the thing I'm most proud of having done."

On the right is Conico Norteño, on the central plateau, about 1,850 meters elevation, and on the far left are maize–teosinte hybrids from the same fields. And if I take the maize–teosinte hybrids and backcross them to corn, within two years I get corn back. Making corn out of teosinte F1 hybrids is no problem.

This is from Tehuacán in Puebla, about 200-plus miles from Mitla, the archaeological material I showed you earlier. And you can see all the rachis segments on the left, typical of F1 hybrids of teosinte and maize. And this site has both maize cobs and F1s of teosinte and maize, but no pure teosinte fruit cases, as in the slide you saw earlier from Mitla.

Mitla had these wonderful spikes of maize-contaminated teosinte from 3,000 years ago.

This is the crown jewel of the excavations at Tehuacán by MacNeish. This is found at the lowest level. It's dated to about 7,000 years ago. It has a very small rachis, or cob, in the center, and very papery glumes. This is polystichous. These are paired spikelets. As you see, it's about an inch in length. It's the oldest material we have, and it shows no evidence of lignification. It shows no evidence of teosinte germplasm. In fact, it's almost the opposite of teosinte germplasm. It isn't until about 5,000 years ago that we find evidence of teosinte germplasm in Tehuacán. So there are two timeframes for hybridization of maize and teosinte: one about 5,000 years ago, and the other about 3,000 years ago.

In the foreground is a less-than-robust teosinte plant, and this is clearly a maize field — you'll see almost one central spike for this corn, which is Chalqueño, but in the foreground is a teosinte plant.

The plant in the center is obviously teosinte, and it has a multi-branched tassel. This is Chalqueño, and in the furrow there aren't many teosinte plants, but in the row with corn, the farmer doesn't cut out the teosinte plants because he can't tell whether it's teosinte or corn. Teosinte here has a very broad leaf, and a sun-red color. It mimics corn in this region, but at about this point the farmer will cut it out, take it home, and feed it to the livestock — the cow — and cows being ruminants, the seeds don't germinate very well, but the rachis segments will pass right through the donkey, mule, or horse, and he'll take that manure and put it back on his cornfield the next year, so there's a cycle of replanting teosinte even when they've rogued it out of the field. This is a really clear example of a backcross to teosinte in a maize field.

This is a maize cob with papery glumes, but a larger cob, about 3,000 years ago, and in the next slide, which is indurated, we'll clearly see teosinte hybridization.

The cupules here are deep. The induration of the rachis is pronounced, and this is a very lignified cob.

In the 1960s, this was the largest population of teosinte, covering hundreds of square miles in the Balsas River basin. This population is now very much in danger, because there's no money in growing corn on these steep slopes, on account of the lower price of U.S. corn, and all these farmers have migrated to California to harvest the crops.

This is in the Balsas, and you may know this teosinte as Parviglumis, but I feel it's much more accurate to consider it a race of Mexican teosinte.

These are ranchos, or ranch houses, and when you go into these mountains, you won't find any males between the ages of 15 and 70 in the village, because they've all migrated to the [United] States, because there's no money in harvesting or selling corn on these mountain slopes.

These three people have contributed most of the teosinte seed that is in the U.S. gene bank. On the left is Jesús Sánchez González, a Mexican who has worked extensively in the years after I did the thesis work. In the center is Ángel Kato, and Kato is at the Escuela Nacional de Graduados in Montecillos, in the Valley of Mexico, and is a cytologist trained by Barbara McClintock. And then, yours truly, in the blue shirt — we're harvesting a very unique corn that grows in the dry sector of the Valley of Mexico: a three-month teosinte. This is highly unusual, because most teosinte plants have a five- to eight-month maturity, but this — in the drought — this teosinte practices drought avoidance by being very quick in its maturation.

This is Efraím Hernández Xolocotzi, and Xolocotzi is an ethnobotanist at the National School of Agriculture, and this picture was probably taken in the 1970s, when he was about 50 years old. He is the plant collector who collected for the Rockefeller Foundation in the 1940s. His mother migrated from Tlaxcala to New York City, and he was raised in New York, went to Cornell as an undergraduate, and then, on graduation, returned to Mexico, and was later a student at Harvard University. He was a mentor of mine, and probably the most knowledgeable person on Mexican ethnobotany.

So far we've been looking at archaeological material and wild relatives of corn in the field, but this is preparing seed for the Arctic Circle gene bank.

I'm going to take you to places that most people don't have the privilege to see. This is the head of the maize gene bank for 30 years at CIMMYT, Suketoshi Taba. He and I were colleagues, and this is the door to the elevator that will take passengers 25 feet down below ground level to the active gene bank vault, and then, once a year, they go into the vault below that, below the water table, which is the long-term storage gene bank.

The gene bank vault — the gene bank — is fitted with compactors and is about the size of, a little bigger than, two professional tennis courts.

Here I am in the vault with the director of the gene bank for the last eight years, Denise Costich. She reached, last week, the mandatory retirement age, which is 65 at CIMMYT, and so there's a new head of the maize gene bank.

Seed is stored in these plastic containers, where you can see the level of seed left. This is in the active collection. Each one of these canisters represents one accession.

These are the canisters at the active gene bank. The temperature is just above the freezing point, and you can see the barcode. The barcode gives the position in the gene bank — which shelf, which row — the racial composition of the seed, the maize-race composition, and then, below that, the data on where the seed was regenerated.

Gene banks are a solution — maybe not the best solution, but a solution — for genetic erosion and extinction. But there's one other thing I want to emphasize, and that is knowledge. This is from a Zapotec village in Oaxaca, and this grouping of ears of corn represents seven races. These are landraces that these cultivators are able to keep from out-crossing with other fields — meaning that where they grow Palomero Toluqueño, or Bofo, depends on what their neighbors are growing, and where the wind currents run. This is a lot of knowledge that we don't have in a gene bank.

I like this slide — it's from India, but it's about talking between generations. This is something gene banks need to pay a little more attention to in their holdings: what selection pressures shaped certain populations, and so on. An example that comes to mind is Bix Sikes, in the Norwegian gene bank, saying that we have seed of five-month barley in the gene bank, and we have seed of three-month barley in the gene bank — but when farmers in Sweden grew the barley, they used to mix the seed of both three-month and five-month barley and plant it together in the field, because if it was an early fall, they had three-month barley to harvest; if it was a year with a late fall, they had both five-month and three-month barley to harvest. That is, they ran the field as an insurance policy by having two genotypes. That's what I meant by "smarts." Unfortunately, we haven't collected some of that "smarts" material for the gene bank.

This is the 1960s, and teosinte was abundant in Mexico on dirt roads. It's very hard sometimes to find it on dirt roads now, because most of this land has been converted to raising livestock, since there's no market for corn.

In the last slide I said there's no market for corn there. Obviously there is a market for corn, but the cost for a campesino to grow that corn is much higher than the cost of the U.S. dumping corn in Mexico, so that Mexican-grown corn is, for many areas of Mexico, not cost-effective.

This is that three-month teosinte you saw earlier, with the three macheteros — Jesús, Kato, and myself. This picture was taken in the 1960s. It's exactly the same site, and I know it's the same site because the cooperative dairy had a retail store across the way, out in the middle of the corn fields.

This is exactly the location of the previous slide. This is what we would call extinction. This was a unique three-month teosinte. No other population of teosinte in the Americas is three-month, and if it wasn't for the gene bank, we wouldn't have it anymore.

My hope for the future is that there will still be room for teosinte to grow in the maize fields of Mexico. Here you see the maize in the stacks, and the teosinte plants left standing in the field. This is Amecameca, 2,250 meters, with the sleeping lady, Iztaccíhuatl, and the volcano Popocatépetl on the right. My hope is that there will be places in Mexico in the future — 50 years from now — where teosinte will still have a niche, and will not have gone extinct.

1967 — Teosinte: The Closest Relative of Maize

Doctoral thesis, Bussey Institute, Harvard University. 159 pp.

Dr. Wilkes' foundational work. It established the taxonomic division of genus Zea into annual and perennial teosinte and identified the distinct annual teosinte races: the classification framework later maize-origin research was built on.

1970 — Teosinte Introgression in the Maize of the Nobogame Valley

Botanical Museum Leaflets, Harvard University 22(9): 297–311.

A detailed field case study documenting natural gene flow between teosinte and maize in a specific Mexican valley: early evidence, grounded in direct observation, for the introgression patterns Dr. Wilkes argued for more broadly.

1972 — Maize and Its Wild Relatives

Science 177(4054): 1071–1077.

Dr. Wilkes' most-cited paper. Synthesized the relationship between maize, teosinte, and Tripsacum, and made the case for preserving wild relatives as a genetic reservoir for crop improvement: still cited in current maize genomics research.

1977 — Hybridization of Maize and Teosinte, in Mexico and Guatemala, and the Improvement of Maize

Economic Botany 31: 254–293.

Documented maize–teosinte introgression across Mesoamerica at field scale and linked it directly to practical plant-breeding value.

1989 — Maize: Domestication, Racial Evolution and Spread

In Foraging and Farming: The Evolution of Plant Exploitation (D.R. Harris & G.C. Hillman, eds.), pp. 440–455.

A synthesis chapter for a major archaeobotany volume, tying together maize's domestication, racial diversification, and geographic spread: it bridged Dr. Wilkes' earlier field-based papers and Dr. Wilkes' later conservation advocacy.

1995 — Mystery and Missing Links: The Origin of Maize

With M.M. Goodman. CIMMYT, Mexico, D.F.

A joint synthesis with maize geneticist Major Goodman, revisiting the origin-of-maize debate in light of two more decades of evidence. It marks a shift toward collaborative, consensus-building work late in Dr. Wilkes' career.

2006 — Urgent Notice to All Maize Researchers

Disappearance and extinction of the last wild teosinte populations is more than half completed: a modest proposal for teosinte evolution and conservation in situ, the Balsas, Guerrero, Mexico.

A late-career call to action on habitat loss in the Balsas region: the last wild refuge of maize's direct progenitor. It reflects Dr. Wilkes' shift from taxonomy toward genetic-resource conservation policy, and connects directly to CIMMYT naming its maize wild-relatives greenhouse after Dr. Wilkes in 2020.