📚 4.3 Classification and Evolution: Experimental Design | 4.3 分类与进化：实验设计
Classification and evolution are cornerstones of modern biology, enabling us to organise the immense diversity of life and to understand how species have changed over time. Designing experiments to investigate classification and evolution involves a range of approaches — from traditional morphological observations to cutting-edge molecular techniques. In the A-Level Biology curriculum, you are expected to know how to design, conduct and evaluate investigations related to taxonomy, phylogenetics, natural selection and biodiversity. This article provides a comprehensive guide to experimental design in this topic, covering key methods, controls, variables and data analysis.
分类与进化是现代生物学的基石，使我们能够整理生命的巨大多样性，并理解物种如何随时间变化。设计实验来研究分类与进化涉及一系列方法——从传统的形态观察到前沿的分子技术。在 A-Level 生物学课程中，你需要了解如何设计、实施和评价与分类学、系统发育、自然选择和生物多样性相关的调查。本文提供了该主题实验设计的全面指南，涵盖关键方法、对照、变量和数据分析。
1. Principles of Experimental Design in Classification and Evolution | 分类与进化中的实验设计原则
Any investigation in biology must start with a clear hypothesis and well-defined variables. In classification experiments, the independent variable might be the choice of characteristic (morphological, molecular) used for grouping, while the dependent variable is the resulting taxonomic grouping or similarity score. Controlled variables include the measurement scale, the age of specimens and the environmental conditions of observation.
任何生物学调查都必须从明确的假设和清晰界定的变量开始。在分类实验中，自变量可能是用于分组的特征选择（形态、分子），而因变量是产生的分类分组或相似性得分。控制变量包括测量尺度、标本年龄和观察的环境条件。
Reliability is improved by using a sufficiently large sample size and by taking repeated measurements. For example, if you are measuring leaf length to classify plants, you should measure at least 30 leaves per species from multiple individuals. This reduces the impact of random variation and allows calculation of a mean and standard deviation.
通过使用足够大的样本量和重复测量可以提高可靠性。例如，如果你通过测量叶长来分类植物，你应当测量每个物种至少30片叶，源自不同个体。这可以减少随机变异的影响，并允许计算平均值和标准差。
Ethical considerations are paramount when working with living organisms. In evolution experiments, such as those observing antibiotic resistance in bacteria, strict aseptic technique and safe disposal are required. For field studies on biodiversity, sampling should minimise habitat disturbance and organisms should be returned to their habitat promptly.
在处理活体生物时，伦理方面的考虑至关重要。在进化实验中，例如观察细菌抗药性的实验，需要严格的无菌操作和安全处理。对于野外生物多样性研究，采样应尽量减少对栖息地的干扰，并将生物尽快送回栖息地。
2. Morphological Classification and Dichotomous Keys | 形态分类与二分法检索表
One of the classic experiments is to construct a dichotomous key using a set of supplied specimens. You are given leaves, shells or insect images and you must design a key that allows another person to identify each specimen correctly. The success of your key is tested by having a peer use it blindly.
经典实验之一是使用一组提供的标本构建二分法检索表。你会拿到树叶、贝壳或昆虫图片，必须设计一个检索表，让另一个人能正确识别每个标本。你的检索表是否成功可以通过让同伴盲试来测试。
In your plan, clearly state that a fresh specimen should be observed under a hand lens, and all measurements should be taken with a ruler calibrated in millimetres. Record features such as leaf margin type, venation pattern, presence of hairs and length-to-width ratio. The independent variable is the specimen identity, and the dependent variable is the route taken through the key. A control could involve comparing identifications made using the key against a published taxonomic key.
在你的计划中，要明确说明新鲜标本应使用手持放大镜观察，所有测量都应用毫米刻度的直尺进行。记录叶缘类型、叶脉样式、有无毛被以及长宽比等特征。自变量是标本的身份，因变量是检索表给出的分类路径。对照可以是将使用该检索表的鉴定结果与已发表的分类检索表进行比较。
A common method is to construct a table of character states:
常用方法是构建特征状态表：
| Specimen / 标本 | Leaf margin / 叶缘 | Venation / 叶脉 | Hairs on lower surface / 下表面毛被 |
|---|---|---|---|
| A | Entire / 全缘 | Parallel / 平行 | Absent / 无 |
| B | Serrate / 锯齿 | Reticulate / 网状 | Present / 有 |
From this table, couplets can be written – for example, ‘1a leaf margin entire …….. go to 2’ and ‘1b leaf margin serrate ……. go to 3’. The key is then trialled and refined. The experiment demonstrates that classification based solely on morphology can be subjective, but standardised measurements improve objectivity.
通过该表可以编写对句——例如，“1a 叶缘全缘…….. 转到2”和“1b 叶缘锯齿…….. 转到3”。然后测试并优化检索表。该实验表明，仅基于形态的分类可能带有主观性，但标准化测量可以提高客观性。
3. Using DNA Barcoding to Identify Species | 使用DNA条形码鉴定物种
DNA barcoding uses a short, standardised region of the genome to identify species. In animals, the mitochondrial cytochrome c oxidase subunit I (COI) gene is most commonly used. For plants, the chloroplast genes rbcL and matK are preferred. An A-Level experiment might involve extracting DNA from an unknown tissue sample, amplifying the barcode region by PCR, and then sending the product for sequencing.
DNA条形码利用基因组中一段短的、标准化的区域来鉴定物种。在动物中，最常用的是线粒体细胞色素c氧化酶亚基I (COI)基因。对于植物，则更常用叶绿体基因rbcL和matK。一项A-Level实验可能涉及从未知组织样本中提取DNA，通过PCR扩增条形码区域，然后将产物送测序。
To design this investigation, the independent variable is the source of the DNA sample (different species or unknown vs. known). The dependent variable is the sequence obtained and its percentage identity match in a database like BOLD or GenBank. Controlled variables include the PCR cycling parameters, primer sequences and DNA polymerase concentration. The experiment must include positive and negative controls: a known DNA template and a no-template control, respectively.
为设计该调查，自变量是DNA样本的来源（不同物种或未知与已知）。因变量是获得的序列及其在BOLD或GenBank等数据库中的相似度百分比匹配。控制变量包括PCR循环参数、引物序列和DNA聚合酶浓度。实验必须包含阳性和阴性对照：分别是已知DNA模板和无模板对照。
After obtaining the trace files, a bioinformatics practical can be done: the sequence is edited and aligned, then a BLAST search reveals the closest matches. A table of results might show percentage identity and E-value, where an E-value near 0 indicates a highly significant match. Ethical approval is needed if the DNA comes from a protected species.
获得测序峰图文件后，可以进行生物信息学实操：编辑并比对序列，然后通过BLAST搜索显示最接近的匹配。结果表可能显示相似度百分比和E值，E值越接近0表示匹配越显著。如果DNA来自受保护物种，则需要伦理审批。
4. Protein Gel Electrophoresis for Evolutionary Relationships | 蛋白质凝胶电泳探究进化关系
Before the widespread use of DNA sequencing, protein electrophoresis was a key tool for studying evolutionary relationships. Proteins extracted from different species are run on a polyacrylamide gel, and the banding patterns are compared. Differences arise from amino acid substitutions that alter the protein’s net charge or size. A classic experiment uses the enzyme lactate dehydrogenase (LDH) from different vertebrates.
在DNA测序广泛应用之前，蛋白质电泳是研究进化关系的重要工具。从不同物种中提取的蛋白质在聚丙烯酰胺凝胶上分离，比较条带模式。条带的差异源于氨基酸替换，改变了蛋白质的净电荷或大小。一个经典实验使用不同脊椎动物的乳酸脱氢酶 (LDH)。
To plan this, you would prepare protein extracts under identical conditions (temperature, buffer, protease inhibitors). The independent variable is the species origin of the protein; the dependent variable is the relative mobility (Rf value) of each band. A standard protein ladder is included as a control. The number of shared bands can be used to construct a simple similarity matrix, and from that, a dendrogram can be drawn suggesting evolutionary distances.
为计划该实验，你将在相同条件（温度、缓冲液、蛋白酶抑制剂）下制备蛋白质提取物。自变量是蛋白质的物种来源；因变量是每条带的相对迁移率（Rf值）。实验中包含标准蛋白质分子量梯作为对照。共享条带的数目可用于构建简单的相似性矩阵，并据此绘制树状图来推测进化距离。
Limitations of this technique include the fact that many mutations are silent at the protein level and that comigrating bands may not represent identical proteins. Nevertheless, it provides a valuable illustration of the principle that closely related species have more similar protein profiles.
该技术的局限性包括许多突变在蛋白质水平上不表现出来，以及共迁移的条带可能不代表同一蛋白质。尽管如此，它仍有力地说明了亲缘关系较近的物种具有更相似的蛋白质谱这一原则。
5. Simulating Natural Selection: The Peppered Moth Investigation | 模拟自然选择：椒花蛾调查
The peppered moth, Biston betularia, is a classic example of natural selection. An excellent classroom simulation can be set up using a piece of dark or light ‘tree bark’ fabric, and dark and light paper ‘moths’ scattered on it. Students acting as predators have 10 seconds to pick up as many moths as possible, and the survival rates are recorded.
椒花蛾 (Biston betularia) 是自然选择的经典范例。可以用一块深色或浅色的“树皮”布料，并在上面撒放深色和浅色纸“蛾”，搭建极好的课堂模拟实验。扮演捕食者的学生用10秒钟尽可能多地捡起飞蛾，记录存活率。
In your experimental design, the independent variable is the background colour (light vs. dark bark) or the initial proportion of moth colour forms. The dependent variable is the proportion of each form ‘surviving’ after a predation bout. Control variables include the time allowed, the number of released moths, the distance of the predator from the bark and the lighting intensity. You should repeat the trial several times with the same conditions and also rotate the predator role to minimise learning bias.
在你的实验设计中，自变量是背景颜色（浅色与深色树皮）或蛾体色型的初始比例。因变量是经历一次捕食回合后每种体色的“存活”比例。控制变量包括允许的时间、投放的蛾数、捕食者与树皮的距离以及光照强度。你应当在相同条件下重复多次试验，并轮换捕食者角色以最小化学习偏差。
Data can be recorded in a table and then analysed using a chi-squared test to see whether the observed survival frequencies differ significantly from those expected under no selection. A suitable null hypothesis states that survival is independent of moth colour. If the calculated chi-squared value exceeds the critical value at p=0.05, you reject the null hypothesis and infer directional selection.
数据可记录在表格中，然后用卡方检验分析，看观察到的存活频率是否与零选择下的预期频率显著不同。合适的零假设为：存活率与蛾体色无关。如果卡方计算值超过p=0.05的临界值，则拒绝零假设，推断存在定向选择。
χ² = Σ (O – E)² / E
This investigation vividly demonstrates how environmental change can shift allele frequencies over generations.
该调查生动地展示了环境变化如何在世代间改变等位基因频率。
6. Investigating Antibiotic Resistance Evolution in Bacteria | 研究细菌抗药性的进化
The evolution of antibiotic resistance is a serious global health threat and can be modelled in the school laboratory using a non-pathogenic strain of Bacillus subtilis or E. coli K12. One common experiment involves exposing bacteria to gradually increasing concentrations of an antibiotic in a gradient plate, observing where colonies can grow.
抗生素耐药性的进化是一个严重的全球健康威胁，可在学校实验室使用非致病菌株，如枯草芽孢杆菌或大肠杆菌K12来模拟。一个常见的实验是让细菌暴露在梯度平板中逐步升高的抗生素浓度中，观察菌落能在哪里生长。
For a more evolution-focused investigation, you can carry out serial passaging: a liquid culture of bacteria is grown overnight in a sub-inhibitory concentration of ampicillin, then a sample is transferred to fresh medium with a slightly higher concentration. By repeating this for many generations, you can measure the minimum inhibitory concentration (MIC) over time. The independent variable is the generation number or antibiotic concentration; the dependent variable is the MIC or growth rate. Control flasks without antibiotic maintain the viability of the original sensitive strain for comparison.
要为更侧重进化的研究，可进行连续传代：让细菌液体培养物在亚抑制浓度的氨苄青霉素中过夜生长，然后将样品转移到含有略高浓度的新鲜培养基中。通过多次重复传代，可以测量最低抑菌浓度 (MIC) 随时间的变化。自变量是传代次数或抗生素浓度；因变量是MIC或生长速率。不含抗生素的对照瓶用于维持原始敏感菌株的活力以供比较。
Safety is paramount: all bacterial work must use aseptic technique in a biosafety cabinet or near a Bunsen burner flame. All cultures and contaminated materials must be autoclaved before disposal. The results typically show a stepwise increase in MIC, demonstrating evolution by mutation and natural selection. You can also plate samples to check for the presence of resistant colonies.
安全至关重要：所有细菌操作都必须使用无菌技术，在生物安全柜或本生灯火焰附近进行。所有培养物和受污染的材料必须在丢弃前高压灭菌。结果通常显示MIC逐步上升，展示了通过突变和自然选择发生的进化。你也可以平板划线样品以检查是否有耐药菌落出现。
7. Estimating Biodiversity: Sampling Techniques | 估计生物多样性：采样技术
Fieldwork is essential for collecting data on species distribution and abundance. Two key techniques are quadrat sampling for sessile organisms (plants, lichens) and mark-release-recapture for motile organisms (snails, woodlice). A well-designed investigation compares biodiversity between two habitats, such as a mown playing field and an unmanaged meadow.
野外工作是收集物种分布和丰度数据的基础。两项关键技术是用于固着生物（植物、地衣）的样方采样和用于活动生物（蜗牛、潮虫）的标记重捕法。一个精心设计的调查可比较两个栖息地之间的生物多样性，例如修剪过的运动场和未管理的草甸。
In a quadrat study, the independent variable is the habitat type; the dependent variables are species richness (number of species) and species evenness. Random placement of quadrats using random number coordinates ensures unbiased sampling. A minimum of 10 quadrats per site is recommended. For each quadrat, list all species present and their percentage cover. Controlled variables include quadrat size (e.g. 0.25 m²) and time of day. Simpson’s Diversity Index (D) can be calculated:
在样方研究中，自变量是栖息地类型；因变量是物种丰富度（物种数）和物种均匀度。使用随机数坐标随机放置样方可以确保无偏采样。每个地点建议至少放置10个样方。在每个样方中，列出所有存在的物种及其覆盖率百分比。控制变量包括样方大小（如0.25 m²）和一天中的时间。可计算辛普森多样性指数 (D)：
D = 1 – ∑ (n/N)²
where n is the number of individuals of a particular species, and N is the total number of individuals. A higher D value indicates greater diversity.
其中 n 是某个特定物种的个体数，N 是个体总数。D值越高表示多样性越高。
For mark-release-recapture, the Lincoln Index estimates population size: N = (M × C) / R, where M is the number initially marked, C is the total number caught in the second sample, and R is the number of marked individuals recaptured. Key assumptions – such as no immigration, no emigration, and that marking does not affect survival – must be discussed as limitations.
对于标记重捕法，林肯指数估算种群大小：N = (M × C) / R，其中 M 是最初标记的数量，C 是第二次捕获的总数，R 是重捕到的已标记个体数。关键的假设——如没有迁入、没有迁出、以及标记不影响存活——必须作为局限性加以讨论。
8. Constructing Phylogenetic Trees from Molecular Data | 从分子数据构建系统发育树
Modern phylogenetics relies heavily on DNA sequence alignment. You can give students pre-prepared sequences of the same gene from several primate species, along with an outgroup (e.g. mouse). By counting the number of nucleotide differences between each pair of sequences, a distance matrix is constructed. This is then used to build a simple tree using the neighbor-joining method.
现代系统发育学高度依赖DNA序列比对。你可以给学生提供预先准备好的同一基因的序列，来自若干灵长类物种，以及一个外群（如小鼠）。通过计算每对序列之间的核苷酸差异数，构建距离矩阵。然后利用该矩阵通过邻接法构建一棵简单的系统树。
In an exam-style experimental design question, you might be asked to outline how to obtain such data. Your plan would mention extracting DNA, PCR amplifying the chosen gene region, Sanger sequencing, and then editing chromatograms to obtain consensus sequences. The independent variable is the species; the dependent variable is the genetic distance or tree topology. Alignment software such as Clustal Omega is used under controlled parameters (gap opening penalty). Bootstrapping values (e.g. 1000 replicates) are calculated to assess the reliability of branches.
在考试风格的实验设计题中，你可能会被要求概述如何获得这类数据。你的计划将包括提取DNA、PCR扩增所选基因区域、桑格测序，然后编辑测序峰图以获得一致序列。自变量是物种；因变量是遗传距离或树拓扑结构。使用如Clustal Omega等比对软件，在受控参数（空位开启罚分）下进行分析。计算自举值（如1000次重复）以评估分支的可靠性。
Such an investigation demonstrates that gorillas and chimpanzees share a more recent common ancestor with humans than with orangutans, and that molecular data can resolve relationships that are ambiguous from morphology alone.
这样的调查表明，大猩猩和黑猩猩与人类共享一个比红毛猩猩更近的共同祖先，并且分子数据可以解决单凭形态无法确定的亲缘关系。
9. Bioinformatics Tools: BLAST and Sequence Alignment | 生物信息学工具：BLAST与序列比对
An accessible lab experiment uses the online BLAST (Basic Local Alignment Search Tool) to identify an unknown DNA sequence. Students copy a provided nucleotide sequence, paste it into the NCBI BLASTn interface, and analyse the output. The percentage identity, query cover and E-value help determine the most likely species. The independent variable could be different query sequences; the dependent variable is the top hit species name and its statistical support.
一个容易开展的实验室实验使用在线BLAST（基本局部比对搜索工具）来鉴定未知DNA序列。学生复制一条提供的核苷酸序列，粘贴到NCBI BLASTn界面中，然后分析输出结果。相似度百分比、查询覆盖度和E值有助于确定最可能的物种。自变量可以是不同的查询序列；因变量是最佳匹配的物种名称及其统计支持度。
A proper experimental design must include controls: searching with a sequence of known identity verifies that the BLAST algorithm is working, while a nonsense sequence should yield no significant matches. You can extend the investigation by downloading several matching sequences, aligning them, and using MEGA software to draw a phylogenetic tree that visualises evolutionary relationships.
恰当的实验设计必须包括对照：使用已知身份的序列进行搜索，验证BLAST算法工作正常，而一条无意义的序列不应产生显著匹配。你可以扩展调查，下载几条匹配的序列，进行比对，并使用MEGA软件绘制系统发育树，直观展示进化关系。
Limitations include database completeness and the fact that horizontal gene transfer in prokaryotes can obscure species boundaries. However, this experiment reinforces that classification is increasingly based on genetic similarity, not just observable traits.
局限性包括数据库的完整性，以及原核生物中的水平基因转移可能模糊物种界限。但该实验强化了分类越来越基于遗传相似性，而不仅仅是可观察的特征。
10. Statistical Analysis and Interpretation | 统计分析与解释
Good experimental design goes hand in hand with appropriate statistical analysis. In a natural selection simulation, the chi-squared goodness-of-fit test determines whether the observed frequencies deviate from a null model. For comparing the mean leaf lengths of two groups, a Student’s t-test is used if the data are normally distributed. When comparing Simpson’s Diversity Index values between two habitats, a Mann-Whitney U test is more suitable if data are not normally distributed.
好的实验设计与适当的统计分析密不可分。在自然选择模拟中，卡方拟合优度检验可判断观察频率是否偏离零模型。为了比较两组的平均叶长，如果数据呈正态分布，则使用学生t检验。比较两个栖息地之间的辛普森多样性指数值时，若数据非正态，则曼-惠特尼U检验更为合适。
When presenting data, always include error bars (e.g. representing ±1 standard deviation) on bar charts. The number of replicates (n) must be stated. In the caption, explain what the error bar indicates. You should also discuss whether the results are statistically significant at the p<0.05 level and what further experiments could refine the conclusions.
展示数据时，始终在条形图中加入误差条（例如表示±1个标准差）。必须注明重复数 (n)。在图注中说明误差条的含义。你还应讨论结果在p<0.05水平下是否具有统计显著性，以及哪些进一步实验可以完善结论。
Recognising sources of error is crucial. In a morphology-based classification, observer bias can be reduced by using a blind protocol. In a DNA barcoding experiment, cross-contamination can produce false positives. Always suggest improvements: e.g. increase sample size, use a more precise balance, or calibrate the PCR machine regularly.
识别误差来源至关重要。在基于形态的分类中，可使用盲法协议减少观察者偏差。在DNA条形码实验中，交叉污染可能产生假阳性。始终提出改进建议：例如增加样本量、使用更精密的天平，或定期校准PCR仪。
Published by TutorHao | Biology Revision Series | aleveler.com
更多咨询请联系16621398022（同微信）
屏轩国际教育cambridge primary/secondary checkpoint, cat4, ukiset,ukcat,igcse,alevel,PAT,STEP,MAT, ibdp,ap,ssat,sat,sat2课程辅导，国外大学本科硕士研究生博士课程论文辅导