树皮变黑之后，一只浅色桦尺蛾会怎样？The bark turns black. What happens to a light peppered moth?
19 世纪的英国工业城市里，煤烟把长着地衣的浅色树干熏成了黑色。桦尺蛾白天停在树干上，靠和树皮颜色相近躲过鸟。几十年后，曼彻斯特附近的桦尺蛾几乎全变成了黑色。“变成”这个词，到底是谁变了？
In 19th-century industrial Britain, coal smoke blackened tree trunks that had been pale with lichen. Peppered moths rest on trunks by day and escape birds by matching the bark. Within decades, almost every peppered moth near Manchester was black. But what exactly “became” black?
01预测Predict
02繁殖几代Breed the population
左边是一片树皮上的桦尺蛾，右边是每一代的记录。黑色由显性等位基因 C 决定（CC 和 Cc 都是黑色），浅色是 cc。每一代，颜色和树皮差得越多的蛾越容易被鸟吃掉，活下来的蛾产下下一代。
On the left, moths on a patch of bark; on the right, the record of every generation. Black is set by a dominant allele C (both CC and Cc are black); light moths are cc. Each generation, birds eat more of the moths that stand out from the bark, and the survivors breed the next generation.
你来当鸟Be the bird
8 秒内点击你能找到的蛾，被点中的就被吃掉了。下一代由你没找到的蛾繁殖。You have 8 seconds to click every moth you can spot; each click eats one. The moths you miss breed the next generation.
换一个种群Change the population
改变下面三项会从第 0 代重新开始。Changing these three starts again from generation 0.
03解释Explain
没有一只蛾改变过颜色。每只蛾一生都保持出生时的颜色。变的是整群蛾里黑色和浅色的比例，而且只在代与代之间变。这需要三个条件同时成立：
- 变异：群体里本来就有黑色和浅色两种。
- 遗传：颜色由基因决定，会传给后代。
- 繁殖差异：在黑树皮上，黑色蛾活得更久、留下更多后代。
拿掉任何一个，选择就停了：把初始频率调到 0，不允许突变，树皮再黑也不会出现一只黑蛾；把捕食强度调到 0，比例只会随机晃动。所以说演化的是种群，不是个体。
为什么回来得那么慢
浅色是隐性的。浅色等位基因 c 变少以后，大多数藏在看起来是黑色的 Cc 杂合子里，鸟看不见它，选择也就碰不到它。
No moth ever changed color. Each moth keeps the color it was born with. What changed was the mix of black and light moths in the population, and only from one generation to the next. That takes three things at once:
- Variation: the population already contains both black and light moths.
- Heredity: color is set by genes and passed to offspring.
- Differential reproduction: on dark bark, black moths live longer and leave more offspring.
Remove any one and selection stops. Set the starting frequency to 0 with mutation off, and no black moth ever appears however dark the bark gets. Set predation to 0, and the mix only wanders at random. Populations evolve; individuals don’t.
Why the way back is so slow
Light is recessive. Once the light allele c is rare, most copies sit in Cc heterozygotes that look black. Birds can’t see it, so selection can’t reach it.
- 先看树皮上的蛾：每一只都保持出生时的颜色，没有一只变过。Start with the moths on the bark: each keeps the colour it was born with; none ever changes.
- 再看存活率：在黑树皮上，黑色蛾活下来的比例更高。Then survival: on dark bark, a larger share of black moths lives to breed.
- 看记录里的柱子：黑色蛾的比例一代一代往上走，变化只发生在代与代之间。Now the bars in the record: the black share climbs generation by generation, and only between generations.
- 最后看等位基因频率：它比黑色蛾比例低，因为黑色蛾里有一部分是带着浅色基因的 Cc。Finally the allele frequency: it sits below the black share, because some black moths are Cc and carry the light allele.
试一下：先让黑色蛾占到九成以上，再把树皮调回浅色（1956 年英国《清洁空气法》之后的情形），继续繁殖。浅色蛾一开始几乎不增加，因为它们大多还藏在杂合子里。
这个故事的证据
1848 年曼彻斯特首次记录到黑色型，到 1895 年当地约 98% 是黑色。霍尔丹（1924）算出，要这么快扩散，黑色型每代需要约 30% 的生存优势。凯特尔韦尔在 1950 年代做了标记重捕实验，后来被批评释放密度过高、把蛾放在不自然的位置，甚至有书指控造假，但这一指控没有得到证实。马杰勒斯在 2001–2007 年用近 5000 只蛾重做了实验，结果在他去世后发表（Cook 等，2012）：鸟的选择性捕食足以解释黑色型的兴衰。2016 年，研究者找到了黑色型的突变：cortex 基因里约在 1819 年插入的一段转座子（van’t Hof 等，2016）。
Try it: let black moths pass 90%, then turn the bark pale again (the situation after Britain’s 1956 Clean Air Act) and keep breeding. Light moths barely increase at first, because they still hide in heterozygotes.
The evidence behind the story
The black form was first recorded in Manchester in 1848; by 1895 about 98% of the local moths were black. Haldane (1924) calculated that so fast a spread needed roughly a 30% survival advantage per generation. Kettlewell’s mark–release–recapture experiments in the 1950s were later criticized for high release densities and unnatural resting sites; one book even alleged fraud, a charge that was never substantiated. Michael Majerus repeated the work with nearly 5,000 moths in 2001–2007; published after his death (Cook et al., 2012), it found that selective bird predation is enough to explain the rise and fall of the black form. In 2016 the mutation itself was found: a transposable element inserted into the cortex gene around 1819 (van’t Hof et al., 2016).
“黑色蛾变多，不是因为 ___，而是因为 ___。如果一开始 ___，就不会发生。”“Black moths became common not because ___, but because ___. It would not have happened if ___.”
04独立练习Independent practice
这一步拿掉辅助：左边的记录图、读数和图例都会隐藏，控件锁定。每道题先自己回答，再看模型给出的结果。
This step removes the supports: the record plot, readouts and legend are hidden and the controls are locked. Answer each case yourself, then see what the model does.
05迁移：抗生素耐药Transfer: antibiotic resistance
这个模型的适用范围Where this model applies
只有一个基因座、两个等位基因，世代不重叠，随机交配（每代开始时符合哈迪–温伯格比例）。存活只取决于颜色和树皮的差别，用一个“捕食强度”参数表示；真实的存活还受迁入迁出、蛾的停歇位置和其他天敌影响，所以真实的下降比这里快。
突变率 0.1% 比真实值大得多，只为在几十代内看到效果。随机漂变用固定随机种子抽样，同样的设置每次得到同样的结果。“你来当鸟”是一个游戏，屏幕上的 40 只蛾代表整个种群的比例，不是数据。
One locus with two alleles, non-overlapping generations, random mating (Hardy–Weinberg proportions at the start of each generation). Survival depends only on the mismatch between moth and bark, set by one “predation strength” parameter. Real survival also depends on migration, resting sites and other predators, which is why the real decline was faster than here.
The 0.1% mutation rate is far higher than real rates, so its effect shows within decades. Drift is sampled from a fixed random seed, so the same settings always give the same run. “Be the bird” is a game: the 40 moths on screen stand for the population’s proportions, not data.
离线可用。用 Tab 切换控件，方向键调节滑块；捕食时蛾也可以用 Tab 选中、回车吃掉。没有快捷键会干扰你的写作。Works offline. Use Tab to reach controls and arrow keys on sliders; during a hunt, moths can be reached with Tab and eaten with Enter. No keyboard shortcuts override your writing.
- 浅色蛾 ccLight moth (cc)
- 黑色蛾 CC / CcBlack moth (CC / Cc)
- 每代黑色蛾比例Black share per generation
- 黑色等位基因频率 pDark-allele frequency p
- 没有漂变时的预期Expected without drift