Capstone: plants, crops and climate

This final lesson brings the course together around a practical question: how will the plants people depend on cope with a changing climate, and what tools do plant biologists and breeders have to help them? Answering it draws on every earlier lesson. Photosynthesis and its response to CO₂ and temperature (lessons 3 and 4), water relations and stomata (lesson 5), nutrients and symbioses (lesson 6), hormones (lessons 7 and 8), the timing of flowering (lesson 9), reproduction and seed set (lesson 10) and stress and disease (lesson 11) all determine how a crop or a forest responds to warming, drought and rising CO₂. The lesson first introduces Arabidopsis as the model that made modern plant biology possible, then surveys plant biotechnology and how Canada and the United States regulate it, then examines climate impacts on crops and forests, and ends with a worked capstone analysis of a crop season.

Arabidopsis: a weed that changed biology

Many of the discoveries in this course, from the ABC model to the auxin receptor and the ABA signaling pathway, were made in Arabidopsis thaliana, thale cress, a small weed of the mustard family with no economic value. It became the dominant model for plant biology in the 1980s because it has a short life cycle of about six to eight weeks, is small enough to grow thousands of plants in a room, self-pollinates and produces thousands of seeds per plant, and has one of the smallest genomes of any flowering plant, about 135 million base pairs with roughly 27,000 protein-coding genes. Its genome, published in 2000, was the first plant genome sequenced. Arabidopsis can be transformed simply by dipping its flowers into a suspension of Agrobacterium, which led to collections of insertion mutants covering almost every gene. The 1001 Genomes project has sequenced natural accessions from across its range, from North Africa to Scandinavia, revealing the genetic basis of local adaptation in flowering time, cold tolerance and other traits. Discoveries made in Arabidopsis are routinely transferred to crops such as canola, a close relative, and to more distant species.

Tools of plant biotechnology

Plant genetic engineering began in the early 1980s with the bacterium Agrobacterium tumefaciens, which, as lesson 7 described, naturally transfers a segment of DNA, the T-DNA, from its tumor-inducing plasmid into plant chromosomes. Scientists replaced the tumor-causing genes in the T-DNA with genes of interest and a selectable marker, turning the bacterium into a vehicle for introducing new genes. Transformed cells are selected in tissue culture and regenerated into whole plants using the auxin–cytokinin principles of lesson 7. For species that Agrobacterium transforms poorly, notably many cereals in the early years, the biolistic or gene-gun method shoots DNA-coated gold or tungsten particles into cells. A plant carrying a gene introduced from another organism is called transgenic.

Genome editing, especially with CRISPR–Cas systems since about 2013, has changed the field again. An editing reagent, a Cas nuclease guided by a short RNA, is delivered into plant cells, usually as a transgene, and cuts the genome at a chosen sequence. The cell’s own repair machinery then rejoins the cut, often with small insertions or deletions that knock out the gene; with more elaborate designs, base editors and prime editors make precise single-base changes without a double-strand break. Once the desired edit has been made, the transgene encoding the editing machinery can be removed by ordinary crossing and segregation, so the final plant may carry only a small change indistinguishable from a natural mutation. This distinction between transgenic plants and edited plants without foreign DNA has become central to how countries regulate biotechnology.

import math

# A T0 rice plant carries one copy of a Cas9 transgene (hemizygous, unlinked to the target)
# and is heterozygous for a knockout edit at the target gene. It is self-pollinated.
# Wanted in T1: homozygous for the edit AND free of the Cas9 transgene.
p_cas9_free = 1 / 4          # hemizygous transgene: 1/4 of selfed progeny lack it
p_edit_homo = 1 / 4          # heterozygous edit: 1/4 homozygous edited
p_target = p_cas9_free * p_edit_homo     # independent if unlinked
print(f"probability a T1 plant is edit-homozygous and transgene-free: {p_target:.4f} (1 in {1/p_target:.0f})")

# How many T1 seedlings must be screened to find at least one such plant?
for conf in (0.90, 0.95, 0.99):
    n = math.ceil(math.log(1 - conf) / math.log(1 - p_target))
    print(f"to be {int(conf*100)}% sure of at least one: screen {n} plants")

# If the plant is instead biallelic (two different edits) the edit is fixed in all progeny
print(f"if T0 is biallelic: transgene-free and edited = {1/4:.2f} of T1 (screen far fewer)")

probability a T1 plant is edit-homozygous and transgene-free: 0.0625 (1 in 16)
to be 90% sure of at least one: screen 36 plants
to be 95% sure of at least one: screen 47 plants
to be 99% sure of at least one: screen 72 plants
if T0 is biallelic: transgene-free and edited = 0.25 of T1 (screen far fewer)

The calculation shows why editing projects plan their screens with Mendelian arithmetic. If the original plant carries one transgene copy and one edited allele, only one progeny plant in 16 has both the edit fixed and the transgene gone, so screening about 50 seedlings gives 95 percent confidence of finding at least one. If editing was efficient enough to modify both alleles in the first plant, every progeny inherits an edit and a quarter of them are transgene-free, so far fewer plants are needed. PCR screens for the transgene and sequencing of the target site identify the plants to keep.

How Canada and the United States regulate biotech crops

The United States regulates biotechnology under the Coordinated Framework for the Regulation of Biotechnology of 1986, which divides responsibilities among three agencies using existing laws. The US Department of Agriculture’s Animal and Plant Health Inspection Service (APHIS) regulates genetically engineered organisms that could be plant pests, under 7 CFR part 340; the Environmental Protection Agency regulates pesticidal substances produced in plants, such as Bt toxins, as plant-incorporated protectants; and the Food and Drug Administration consults with developers on food safety. In 2020 APHIS replaced its older part 340 rules with the SECURE rule, which exempted many gene-edited plants and streamlined review. On 2 December 2024 a federal district court in California vacated the SECURE rule, and APHIS announced that it was restoring its pre-2020 framework, including permits, petitions for non-regulated status and its ‘Am I Regulated’ letters, through which developers can ask whether a particular edited plant falls under the rules. In May 2026 USDA published a request for information seeking comments, due by mid-June, on a future regulatory framework for modified organisms under the Plant Protection Act. Foods containing detectable modified genetic material must be labeled under the National Bioengineered Food Disclosure Standard, mandatory since 2022. Students should check the APHIS, EPA and FDA websites for current rules, because this area changes quickly.

Whatever the regulatory system, scientific assessments of transgenic crops have consistently concluded that the method of breeding does not in itself make a crop more or less safe, and that each product should be judged by its traits. Major reviews, including a 2016 report of the US National Academies of Sciences, Engineering, and Medicine, found no substantiated evidence that approved transgenic crops posed greater risks to human health than conventional crops, while documenting real issues of weed and insect resistance that arise from overreliance on single traits. The weed resistance problem is a direct application of lesson 11’s arms-race logic: repeated use of one herbicide, such as glyphosate on glyphosate-tolerant crops, selected for resistant weeds such as Palmer amaranth in the United States and kochia on the Prairies.

Climate change and crops

The climate of the Canadian Prairies and the US Great Plains has warmed by more than the global average since the early twentieth century, with the greatest warming in winter and at night. Atmospheric CO₂ has risen from about 280 ppm before industrialization to more than 420 ppm. The effects on crops combine several processes covered in this course, some favorable and some harmful. Higher CO₂ increases photosynthesis and water use efficiency in C3 crops such as wheat, canola, barley and soybean (lesson 4), though by less in the field than in early chamber experiments, and it lowers grain protein and mineral content. Warmer temperatures lengthen the frost-free season and let crops and varieties move north, but they also speed development, shortening the grain-filling period, and increase the frequency of damaging heat during flowering (lesson 11). Higher temperatures raise evaporative demand, so droughts become more intense even where rainfall does not change (lesson 5). And pests and diseases expand their ranges.

Plant development is driven more by accumulated warmth than by calendar time. Agronomists track it with growing degree days (GDD), the sum over days of the amount by which the mean daily temperature exceeds a base temperature below which the crop does not develop. Base temperatures differ among crops: about 5 °C for wheat and canola, and 10 °C (50 °F) for maize and soybean, with maize calculations also capping the daily maximum at 30 °C (86 °F). Each growth stage requires a roughly constant number of degree days, so degree-day sums predict when a crop will flower or mature. In Ontario and Quebec, a related index called corn heat units is used to match maize hybrids to regions.

import numpy as np

rng = np.random.default_rng(340)
# Invented daily temperatures (F) for a corn field near Ames, Iowa, planted 1 May (day 0)
days = np.arange(0, 140)
noise_max, noise_min = rng.normal(0, 5, days.size), rng.normal(0, 4, days.size)    # same weather noise in both runs
def season(warming_F):
    mean = 60 + 18 * np.sin(np.pi * (days + 10) / 175) + warming_F
    return mean + 11 + noise_max, mean - 11 + noise_min

flower_gdd = 1350      # approximate GDD (base 50 F, cap 86 F) from planting to silking for a mid-season hybrid
for name, warm in (("baseline season", 0.0), ("season 3.6 F warmer", 3.6)):
    tmax, tmin = season(warm)
    hi = np.clip(tmax, 50, 86); lo = np.clip(tmin, 50, 86)  # standard modified GDD method
    gdd = (hi + lo) / 2 - 50
    cum = np.cumsum(gdd)
    start = int(np.argmax(cum >= flower_gdd))
    window = slice(start, start + 14)                       # about 2 weeks around silking
    hot = int((tmax[window] >= 95).sum())
    print(f"{name:20s} silking day {start:3d} after planting; "
          f"days >= 95 F around silking: {hot}; season GDD: {cum[-1]:.0f}")

baseline season      silking day  63 after planting; days >= 95 F around silking: 1; season GDD: 3176
season 3.6 F warmer  silking day  57 after planting; days >= 95 F around silking: 5; season GDD: 3549

The capstone model shows two effects of warming at once. In the warmer season, the corn crop reaches silking several days sooner after planting, because degree days accumulate faster, and silking falls in a part of the summer when very hot days are more common, so more days around silking exceed 95 °F, when pollen viability and kernel set decline and drought stress intensifies. The data are invented, but both effects have been documented across the Corn Belt, and they help explain the yield losses in hot, dry summers such as 2012, when the average US corn yield fell by about a quarter from trend. Growers respond with earlier planting, hybrids of different maturity and drought tolerance, and practices that conserve soil water.

Forests and natural vegetation

Wild plants respond too. Spring events such as leaf-out and flowering have advanced by days per decade in many temperate regions, tracking warmer springs, a shift documented in records stretching back more than a century, including Henry David Thoreau’s notes from Concord, Massachusetts, and long-running phenology networks in Canada and the United States. Mismatches can result when plants and their pollinators or herbivores respond to different cues; a plant timed by day length and an insect timed by temperature may drift apart. In the boreal forest, warming and drought stress, increased fire and insect outbreaks such as the mountain pine beetle, which killed pine across roughly 18 million hectares of British Columbia and spread east of the Rocky Mountains into Alberta in the 2000s, are reshaping a biome that stores a large fraction of the world’s terrestrial carbon. Tree species are expected to shift northward and upslope over generations, but trees migrate slowly, and assisted migration, planting seed sources from warmer regions, is being trialled by forest agencies on both sides of the border. Drought-induced hydraulic failure (lesson 5) has become a major cause of tree mortality in western North America.

Breeding for a changing climate

Breeders have several levers drawn directly from plant biology. Flowering time and maturity genes (lessons 9 and 10) can be adjusted so that crops avoid heat and drought at sensitive stages. Root architecture genes such as those controlling root angle (lesson 9) can push roots deeper. Stomatal density and ABA sensitivity (lessons 5 and 8) can be tuned for water use efficiency. Photoprotection and photorespiration (lessons 3 and 4) are targets for raising photosynthetic efficiency, with field trials showing gains in tobacco and soybean. Disease resistance can be strengthened by stacking resistance genes and using durable partial resistance (lesson 11). Wild relatives of crops, conserved in gene banks, supply alleles for heat, drought and disease tolerance that were lost during domestication. Genomic selection, which predicts breeding value from DNA markers across the whole genome, speeds selection, and genome editing allows targeted changes to known genes. No single trait will make crops climate-proof; resilience comes from combining physiology, genetics and management.