研究初稿 R1 | 2026年11月9日
Research draft R1 | 9 November 2026
本章研究新叶出现、叶片扩展、绿叶衰老与茎群更新怎样共同决定群体叶面积。 读者需要把“叶多”“叶大”“冠层封行”和“叶面积指数高”拆成不同变量,才能解释生长差异及测量误差。
This chapter examines how leaf appearance, blade expansion, senescence and stalk-population turnover jointly determine canopy leaf area. Leaf number, blade size, canopy closure and leaf area index are different variables; separating them is necessary to explain growth responses and measurement error.
核心判据:先定义面积对象与地面分母,再讨论动态和生产意义。 绿叶面积指数不是一个从任意照片、冠层仪或品种名称直接读出的常数。
Core rule: define the area being counted and its ground-area denominator before interpreting dynamics or production relevance. Green LAI is not a constant recoverable from an arbitrary photograph, instrument reading or cultivar name.
绿叶LAI / Green LAI = 单面绿叶片面积 / 水平地面面积
本章以现存、按绿色比例折算的叶片面积为主要对象。 衰老叶、叶鞘、茎秆和杂草分别登记;方法采用其他口径时保留原名称。 [M01]
The main target is standing one-sided lamina area weighted by its green fraction. Senescent tissue, sheaths, stalks and weeds are recorded separately. Alternative definitions remain explicitly named. [M01]
状态:双语初稿与首轮来源核查;仍有全文、独立重复、图像与原始数据缺口。 不是独立专家审定或全书出版批准。
Status: bilingual draft with initial source review and explicit full-text, replication, media and raw-data gaps. This is not independent expert approval or whole-book publication clearance.
016.01 研究对象与核心问题
Objects and key questions
1.1 单叶面积、单茎面积与群体面积
Single-leaf, stalk and crop area
单叶面积描述一个叶片;单茎面积是指定茎上叶片的合计;群体面积还受到每平方米茎数及主茎、早生蘖和晚生蘖比例影响。 抽样时只选最大的主茎,可能高估群体平均单茎面积,即使单叶测量本身准确。
A blade is the unit for single-leaf area. Stalk area sums blades on a defined stalk. Crop area additionally depends on stalk density and the mixture of main stalks and tiller cohorts. Selecting only the largest main stalks can bias the crop mean even when each individual blade is measured accurately.
1.2 绿叶面积与总叶片面积
Green area versus total lamina area
一片叶可以同时包含绿色、黄化和坏死部分。 本章将绿色比例记为g,取值0—1;叶片面积乘以g得到绿叶面积。 视觉绿色并不保证光合能力相同,因此LAI用于描述面积,不能替代气体交换或叶绿素测量。
A blade may contain green, yellow and necrotic tissue. Its green fraction g ranges from zero to one; multiplying blade area by g gives green area. Visible greenness does not ensure equal photosynthetic capacity. LAI describes area and does not replace gas-exchange or chlorophyll measurements.
表016-01 必须分开的变量
Table 016-01. Quantities that must remain distinct
变量 / Quantity | 分子与分母 / Numerator and denominator | 解释 / Interpretation |
|---|---|---|
单叶面积 / Blade area | 一片叶的单面面积 / One-sided area of one blade | m²或cm²;注明展开状态 / State expansion status |
单茎绿叶面积 / Green area per stalk | 该茎绿色叶片面积合计 / Sum on a stalk | m²/stalk;注明叶位与绿色比例 / Rank and fraction needed |
绿叶LAI / Green LAI | 绿叶面积 / 水平地面面积 | m²/m²;可大于1 / May exceed one |
有效光学指数 / Effective optical index | 由光隙反演 / Inferred from gaps | 依赖结构与算法 / Structure- and algorithm-dependent |
覆盖率 / Canopy cover | 投影覆盖面积 / 地面面积 | 0—1;不是LAI / Not LAI |
1.3 本章要回答的动态问题
The dynamic questions
同样的LAI上升,可能来自叶片变大、出叶增加或茎数增加;同样的下降,可能来自衰老、折损或死蘖。 研究应同时记录这些分量,而不是只拟合一条总量曲线。
An LAI increase may arise from larger blades, additional leaves or additional stalks; a decrease may arise from senescence, damage or tiller death. Record the components rather than fitting only an aggregate curve.
016.02 结构基础与过程机制
Structural basis and process mechanisms
2.1 出叶与面积扩展不是同一过程
Leaf appearance and expansion are different processes
可见叶尖、叶领出现和叶片完全展开是不同事件。 叶片数量增加时,新增叶可能尚小;叶片数量不变时,已经出现的叶仍可扩展。 因此需要分别建立“出现时间”和“面积轨迹”,并明确是否纳入心叶中已经露出的部分。 [R01、R03]
Visible tip, visible ligule and full expansion are distinct events. Newly counted leaves may still be small, while previously emerged leaves may expand without changing leaf number. Track appearance time and area trajectories separately, stating whether exposed portions of emerging blades are included. [R01, R03]
2.2 叶位与最终叶片大小
Leaf rank and final blade size
Q117研究发现叶片大小随叶位增加,并在约第17叶及以后出现最大的叶片;研究还报告平均每茎有3.7片叶处于扩展中。 这里的数字是特定材料和条件下的结果,不能把所有叶片都视为“第+3叶的复制品”。 [R01]
In the Q117 study, blade size increased with rank and the largest blades occurred around rank 17 and above; an average of 3.7 leaves per stalk were still expanding. These are material- and condition-specific results, not grounds for treating every leaf as a copy of the +3 leaf. [R01]
2.3 从长度和宽度到面积
From length and width to area
aᵢ = kᵢ × Lᵢ × Wᵢ;aᵢ,green = aᵢ × gᵢ
L为规定的叶片长度,W为最大叶宽,k为经校准的形状系数。 叶片弯曲、缺损、卷叶和量取位置都需记录。 长度沿叶片量取,不能用倾斜投影长度代替。
L is the prescribed blade length, W its maximum width and k a calibrated shape factor. Record curvature, missing tissue, rolling and measurement positions. Measure along the blade rather than substituting its inclined projection.
佛罗里达研究报告的0.72系数在巴西研究中被作为假设沿用。 两篇论文使用同一个系数不等于该系数已经在两地独立校准。 叶形差异较大时,应使用扫描或面积仪建立本地参照。 [R02、R03]
The Florida coefficient of 0.72 was assumed in the Brazilian study. Reuse of a coefficient is not an independent calibration at a second location. When blade shape differs, establish a local reference with scanning or an area meter. [R02, R03]
2.4 主茎、分蘖与密度的共同贡献
Main stalks, tillers and density
Q117研究中,分蘖对总LAI的贡献在5叶阶段为60%—90%,在15叶阶段为20%—50%,且这些比例由总量与主茎量之差求得。 它说明群体组成会随阶段改变,不意味着其他品种可直接套用这些比例。 [R01]
In Q117, inferred tiller contributions were 60–90% of total LAI at the five-leaf stage and 20–50% at the fifteen-leaf stage. These fractions were obtained by difference from total and main-stalk LAI. They demonstrate changing population composition, not transferable fixed fractions for other cultivars. [R01]
LAIgreen = Σₕ(nₕ × āgreen,h)
h代表主茎或分蘖世代等分层;n为茎/m²,ā为m²/茎。 分层面积与茎密度必须来自同一时间和相同地面范围。
h indexes main stalks or tiller cohorts; n is stalks per m² and ā is m² per stalk. Areas and densities must correspond to the same date and ground support.
2.5 绿叶面积收支
A green-area budget
ΔAgreen = 扩展增加 + 新登记茎面积 − 衰老损失 − 其他损失
这是记账关系,不是拟合参数。 已经在“新登记茎面积”中计入的叶片不能再次计为同期扩展;死蘖的全茎绿叶损失与逐叶衰老损失也不能重复扣除。
This is an accounting identity, not a fitted parameter. Area counted upon recruitment must not also be counted as contemporaneous expansion; whole-stalk losses and individual-leaf senescence must likewise not be deducted twice.
2.6 平台期仍可能高速周转
Rapid turnover can occur during an LAI plateau
若每日新增长的绿叶面积恰好抵消衰老损失,总LAI看似不变。 两个相同LAI的冠层可以具有不同叶龄结构、扩展成本和光合能力。 需要追踪永久叶号及绿色比例,才能判断是保持健康叶片还是不断以新叶替换老叶。
When new green area offsets senescent losses, total LAI appears stable. Two canopies with the same LAI may have different leaf-age distributions, expansion costs and photosynthetic capacities. Permanent leaf identities and green fractions distinguish retained foliage from rapid replacement.
016.03 生育阶段与环境响应
Developmental stages and environmental responses
3.1 建群期:数量、大小和空间空缺
Establishment: number, size and spatial gaps
早期LAI较低时,应同时观察出苗与出蘖、叶片扩展和行间空缺。 冠层俯视图提供覆盖率,不能提供被上层叶遮住的叶面积。 缺株区必须留在地面分母中;把采样框只放在浓密蔗蔸上会改变研究对象。
At low early LAI, assess emergence, tillering, blade expansion and inter-row gaps together. A top-view image measures cover but not hidden leaf layers. Missing-stalk areas remain in the ground denominator; placing quadrats only on dense stools changes the population being studied.
3.2 旺长期与LAI峰值
Rapid growth and the LAI peak
群体LAI达到峰值意味着面积增加与损失的平衡发生变化,不代表产量已经达到最大。 可用峰值、峰值日期、达到某一预先声明阈值的时间和绿叶面积持续期描述曲线;阈值仅是研究比较口径,不应宣称为全球生产标准。
Peak LAI marks a change in the balance of area gains and losses, not necessarily maximum yield. Describe the peak, its date, time to a predeclared threshold and green leaf area duration. A comparison threshold is an analytical convention, not a universal production standard.
表016-02 分阶段观测重点
Table 016-02. Stage-specific observations
阶段 / Stage | 同时记录 / Record together | 避免误判 / Avoid |
|---|---|---|
建立 / Establishment | 出苗、茎密度、叶面积、缺株 / Emergence, density, area, gaps | 覆盖率替代LAI / Cover as LAI |
扩展 / Expansion | 叶位、扩展速度、分蘖世代 / Rank, expansion, tiller cohort | 只测大主茎 / Largest stalks only |
平台 / Plateau | 新叶增加、老叶损失 / Area gain and loss | 以为停止出叶 / Assuming no leaf production |
衰退 / Decline | 绿色比例、落叶、死蘖、倒伏 / Greenness, loss, death, lodging | 把所有下降归为成熟 / All decline called ripening |
3.3 衰退期:分清衰老、倒伏与管理移除
Decline: separate senescence, lodging and removal
叶片黄化降低绿叶面积;折叶、剥叶和死蘖改变保留面积;倒伏可能只先改变叶片位置及光隙。 在解释曲线前,应登记事件而不是仅标记“后期”。
Yellowing reduces green area; breakage, leaf removal and stalk death alter retained area. Lodging may initially alter orientation and gap geometry without removing blades. Record events before interpreting a curve as merely late-season decline.
3.4 缺水对增加项与损失项的双重影响
Water deficit affects both gains and losses
缺水试验显示叶片扩展和出叶等过程会改变,随后绿叶数量也可能下降。 本章只采用原研究可核查的定性结果,不把某一土壤水分阈值作为所有蔗区的统一灌溉触发点。 需同步测量土壤水分、叶片状态和扩展速度。 [R07]
Water-deficit experiments report changes in leaf extension and appearance, followed by altered green-leaf numbers. This chapter uses the verified qualitative finding, not a universal soil-water trigger for irrigation. Soil water, blade status and expansion rate should be measured together. [R07]
卷叶会改变投影和光学透过率,但展开后的实体叶片面积未必立即减少。 复水后的光隙变化也不能单独证明重新长出了相同面积的新组织。 应分别保存原位姿态、展开面积及绿色比例。
Rolling changes projection and transmittance without necessarily causing an immediate reduction in unfolded lamina area. A post-rewatering change in gaps alone does not demonstrate tissue regrowth. Preserve in-situ posture, unfolded area and green fraction as separate measurements.
3.5 温度:同时影响出现速度与叶片大小
Temperature affects appearance and blade size
温度相关模型可以帮助比较发育进度,但“同积温”不保证同LAI。 叶位、茎数及衰老仍需计入;模型所用基温必须注明。 第006章处理积温定义,本章只把它作为动态曲线的一个时间坐标。 [R01、R03]
Temperature-based models help compare developmental progress, but equal thermal time does not ensure equal LAI. Rank, stalk density and senescence still matter; document the chosen base temperature. Chapter 006 defines thermal time, used here as one possible coordinate for canopy dynamics. [R01, R03]
3.6 养分、病害与复合条件
Nutrients, disease and combined conditions
本章不根据叶面积外观开列肥料剂量。 营养限制、病害、干旱和光竞争可能产生相似表型;应先记录土壤与组织诊断、病斑比例和同期对照。 管理比较还要报告水量与投入条件,防止把供水差异解释成单独肥效。
This chapter does not prescribe fertilizer rates from visual leaf area. Nutrition, disease, drought and light competition can produce similar phenotypes. Record diagnostic measurements, diseased fraction and concurrent controls. Management comparisons must report water and inputs so that water differences are not misattributed to fertilizer alone.
016.04 品种与生态区差异
Genotypic and ecological differences
4.1 “出叶快”不等于“叶面积一定最大”
Fast appearance does not necessarily mean largest leaf area
巴西三克隆试验比较了早期叶片大小与出叶。 单叶面积差异可使出叶顺序与面积排序不同;研究所讨论的群体潜力包含“茎密度相同”的假设。 因此,品种比较必须同时给出面积和群体密度。 [R03]
The Brazilian clone trial separated early blade size from appearance. Blade-size differences can cause leaf-number and leaf-area rankings to differ; the inferred crop potential included an equal-stalk-density assumption. Cultivar comparisons therefore require both leaf area and population density. [R03]
4.2 新植与宿根:共用状态变量,不强制共用曲线
Plant and ratoon crops: common variables, distinct trajectories
新植和宿根都可按茎群、叶位和绿色比例表达LAI,但建立时已有根系、芽库及采收后环境不同。 Q117研究中的某些组成关系相近,不能代替对其他品种、多代宿根和受胁迫地块的检验。 [R01、R08]
Both plant and ratoon crops can be represented by stalk cohorts, leaf ranks and green fractions. Initial roots, bud banks and postharvest environments differ. Similar component relations in Q117 do not replace tests across other cultivars, ratoon ages and stressed fields. [R01, R08]
4.3 研究“世界差异”应保留条件,而非只列国家
Global comparison requires conditions, not country labels alone
跨地区整理至少保留品种、作物类型、起始日期、物候、降雨与灌溉、行距、采样面积和测量方法。 相同日期或相同株高不等于相同发育阶段。 把叶形系数、光学修正或LAI阈值移植到新环境前,需要独立配对校验。
Cross-region records should retain genotype, crop class, start date, stage, water supply, row geometry, sampled area and method. Equal dates or heights do not establish equal developmental stages. Revalidate shape coefficients, optical corrections and LAI thresholds before transferring them to new environments.
4.4 密度变化的计算例
An example of changing density
12 × 0.25 = 3.0;9 × 0.40 = 3.6
教学假设EX03:茎密度从12降至9茎/m²,但平均单茎绿叶面积从0.25增至0.40 m²,群体LAI仍上升。 此计算不证明每根原有茎都扩大了叶面积,也可能含有选择性死蘖。
Illustration EX03: density falls from 12 to 9 stalks/m² while mean green area rises from 0.25 to 0.40 m²/stalk, so LAI increases. This does not prove that every original stalk expanded; selective mortality may change the mean.
016.05 原始试验与相反证据
Original experiments and conflicting evidence
以下按研究而非论文数量评价证据。 全文、摘要的可读性单独记录;能够阅读全文不等于机制已经强证实。 表中“未取得”表示本章没有获得,不能补写为零或推定重复数。
Evidence is evaluated by study, not by paper count. Access level is recorded separately from evidential strength. Reading a full text does not establish a mechanism conclusively; missing methods or replication are not filled with zeros.
表016-03 叶片与群体发育研究
Table 016-03. Leaf and canopy-development studies
来源 / Source | 本章核查对象 / Verified focus | 支持与限制 / Support and limits |
|---|---|---|
R01 · 1998 | 温度、出叶、叶片大小与分蘖 / Temperature, appearance, size and tillers | 摘要;贡献按差值推得;无胁迫模型 / Abstract; inferred tiller contribution; unstressed model |
R02 · 2004 | 4品种、2年、3地点 / Four cultivars, two years, three sites | 机构技术摘要;0.72不是全球常数 / Institutional abstract; 0.72 not universal |
R03 · 2010 | 3克隆、4区组、逐叶追踪 / Three clones, four blocks, leaf tracking | 全文;播期与年份混杂;面积用假定系数 / Full HTML; year/date confounding; assumed factor |
R08 · 1994 | 2品种、季节宿根起始期 / Two cultivars, seasonal ratoon starts | 出版商摘要;多日期不等于独立试验 / Publisher abstract; dates not independent trials |
5.1 重复来源与重复验证的区别
Repeated sources versus independent verification
R03引用R02的0.72系数,本章仅把R02记录为该数值的来源之一;R03是使用实例。 R01还使用模型与外部数据比较,完整重复关系待阅读全文后再登记,不能把每张图或每个阶段算成独立验证。
R03 reuses R02’s 0.72 coefficient: R02 provides its empirical source and R03 demonstrates use, not another calibration. R01 includes model comparisons with additional data; their independence remains to be checked from full methods. Figures or stages are not separate validations.
5.2 原文出现不一致时怎样处理
Handling inconsistencies in source text
R03有一处叶间积温不等号与文字解释不一致。 本章不据此给出统一早晚叶速率排序,而保留原文疑点。 需要该细节时,应回看原表及试验资料后再修订,而不是凭经验改字。
R03 contains an inequality inconsistent with its verbal phyllochron interpretation. This chapter retains the discrepancy rather than converting it into a universal early/late-leaf ranking. Resolve it against original tables and data before revising the record.
表016-04 测量、校准与水分研究
Table 016-04. Measurement, calibration and water-response studies
来源 / Source | 本章核读 / Access and design | 结论边界 / Boundary |
|---|---|---|
R04 · 2020 | 全文;甘蔗、玉米、大豆校准 / Full HTML; three-crop calibration | 汇总偏差不等于甘蔗专属精度 / Pooled bias not cane-only accuracy |
R05 · 2018 | PDF文字;巴西4地块数据描述 / PDF text; four Brazilian fields | 图表未视觉核实,数据文件未取得 / Figures unverified; raw files not acquired |
R06 · 2023 | 全文;136个ESU、8日期 / Full HTML; 136 ESUs, eight dates | 随机划分不等于跨地区外推 / Random split not external regional validation |
R07 · 2004 | 出版商摘要与预览片段;2试验 / Indexed abstract/preview; two experiments | 只提取定性水分响应;重复数未知 / Qualitative water response; replication unknown |
5.3 不把混合作物误差写成甘蔗误差
Do not relabel pooled errors as sugarcane errors
R04报告三种作物合并口径下,校正前平均高估约20%,校正后约4%,RMSE为0.35 m²/m²。 这说明校准值得检验,但不是“甘蔗LAI仪器误差固定20%”,也不是新环境可直接采用的统一修正比例。 [R04]
R04 reports pooled overestimation of about 20% before correction and 4% afterwards, with RMSE 0.35 m²/m² across three crops. This motivates calibration; it does not establish a fixed 20% sugarcane instrument error or a correction transferable to any new environment. [R04]
5.4 样本量冲突与负校正值
Sample-size conflict and negative corrected estimates
R04方法写有四重复和“五株取样”的叙述,又写甘蔗合计25株。 核实后的株数保留为空。 R05的加性校正产生少量负估计值,绘图时显示为零;数据库应同时保留原始值、校正值和展示规则,不能把负估计解释为负的实体叶面积。 [R04、R05]
R04 reports four replicates and five sampled plants alongside a cane total of 25; the resolved sample count remains null. R05’s additive correction produced some negative estimates displayed as zero. Retain raw, corrected and display-rule fields rather than interpreting these as negative physical area. [R04, R05]
016.06 观测指标与测量方法
Variables and measurement methods
6.1 直接面积法:首先固定地面分母
Direct area: fix the ground denominator first
确定采样区边界,包含分配给蔗行的完整水平地面面积,登记所有茎或按预先定义的层抽样。 宽窄行不能只取窄行宽度作为分母。 对坡地需区分水平投影面积和斜面面积,不能混在同一LAI变量中。
Define the sampling boundary and complete horizontal ground area assigned to the rows. Census stalks or sample predefined strata. In paired rows, do not use only the narrow gap as the denominator. On slopes, distinguish horizontal projected area from surface area.
6.2 扫描、摄影与面积仪的参照链
A reference chain for scanning, imaging and area meters
叶片平铺但不拉伸,减少重叠;用已知面积标尺检查像素换算,分别保存原图、分割掩膜和面积结果。 断叶或缺叶记录为损失,不能用完整叶模板填补为实测。 独立叶片和独立小区应分别编号,重复扫描只估计技术误差。
Flatten blades without stretching, avoid overlap and calibrate pixel area with a known-area reference. Preserve original images, segmentation masks and results. Record missing tissue rather than filling it from an intact-blade template as if observed. Separate leaf and plot identifiers; repeated scans quantify technical error only.
6.3 长宽法校准:系数与截距都要检验
Calibrating length–width models
拟合a=kLW时,检查是否需要截距、系数是否随叶位变化,并将训练与验证按植株或小区分开。 若同一茎的叶片随机分到训练和验证中,验证误差可能低估转移到新材料时的不确定性。
When fitting a=kLW, test the intercept and rank dependence. Split calibration and validation by plant or plot. Randomly assigning leaves from the same stalk to both sets may understate the uncertainty of transfer to new material.
EX01:1.2 × 0.05 × 0.72 = 0.0432 m²/叶
教学假设:每茎8片完全相同的绿叶、10茎/m²时,LAI为3.456;改用k=0.75得到3.600,相差4.17%。 这只是系数敏感性演算,不是某品种的实测表现。
Illustration: with eight identical green blades per stalk and ten stalks/m², LAI is 3.456. Using k=0.75 gives 3.600, a 4.17% change. This is sensitivity arithmetic, not an observed cultivar response.
6.4 部分绿叶的面积折算
Area weighting for partially green blades
EX02:4 × 0.04 + 2 × 0.04 × 0.25 = 0.18 m²/茎
教学假设:6片叶中4片全绿、2片仅25%为绿,每片0.04 m²。 10茎/m²对应绿叶LAI=1.8;若把6片都计为全绿,则得到2.4,高估33.3%。
Illustration: six blades of 0.04 m² include four fully green and two 25%-green blades. At ten stalks/m², green LAI is 1.8; counting all blades as fully green yields 2.4, an overestimate of 33.3%.
若研究采用“超过50%绿色才算绿叶”等二值规则,保留该规则与数量;不要在没有面积数据时将数量自动改写成连续绿叶面积。
If a study counts only leaves exceeding a 50%-green threshold, preserve that count and rule. Do not silently convert a binary count into continuous green area without area information.
6.5 光学法测量的是透过率,而不是逐叶清点
Optical methods measure transmittance, not a leaf census
LAI-2200C对视场中阻挡光的结构均敏感,并依赖叶片空间分布等假设。 倒伏、成行聚集、枯叶和非叶器官可能改变结果。 保存传感器高度、视场限制、天况、测量方向和校正版本;同日期采用直接方法配对核验。 [M01]
LAI-2200C responds to light-blocking structures and assumes aspects of foliage distribution. Lodging, row clumping, dead leaves and non-leaf organs can change its estimate. Record sensor height, view restriction, sky conditions, direction and correction version, and pair readings with direct reference measurements. [M01]
6.6 遥感反演必须保留地面真值定义
Remote retrievals must retain the reference definition
R06在每个5.49 m² ESU中收获3株,测单面叶面积并按ESU植株数量放大。 其LAI反演RMSE中位数0.48 m²/m²来自同一区域观测的随机划分,并不是跨国或跨年的外部验证精度。 [R06]
R06 harvested three plants within each 5.49 m² ESU, measured one-sided leaf area and scaled it using ESU plant counts. Its median LAI-retrieval RMSE of 0.48 m²/m² came from random partitions within one regional campaign, not external validation across countries or years. [R06]
6.7 曲线形状与绿叶面积持续期
Curve shape and green leaf area duration
LAD ≈ Σ [(LAIᵢ + LAIᵢ₊₁)/2] × (tᵢ₊₁ − tᵢ)
EX04教学序列:第0、20、40天的LAI分别为2、4、3,则LAD=130 m²·d/m²。 区间采用线性近似;若长时间缺测,不应把插值当作实测。
EX04 uses invented observations at days 0, 20 and 40 with LAI 2, 4 and 3; LAD is 130 m²·d/m². The trapezoidal approximation assumes linear change between dates. Interpolation across missing intervals is not observation.
LAI曲线通常需要允许增加、平台和下降;单一单调增长模型不能表达衰老或收割。 拟合结果要检查边界、负值与不合理波动,并把采收、倒伏和剥叶事件作为可追溯的断点。
A trajectory must allow increase, plateaus and decline; a single monotonic growth curve cannot represent senescence or harvest. Check boundaries, negatives and implausible oscillation, and retain harvest, lodging or removal events as traceable breakpoints.
6.8 LAI与截光:有条件的递减收益
Conditional diminishing gains in interception
fintercepted = 1 − exp(−k × LAI)
这是固定消光系数的简化表达,不是本章验证的新模型。 它说明增加叶面积可能带来递减的截光增量;没有同时计算呼吸、耗水、茎糖分配与成熟结果时,不能从该式求得经济最优LAI。
This fixed-extinction expression is a simplified model, not a newly validated model here. It illustrates diminishing interception gains with added area. Without respiration, water use, allocation and maturity outcomes, it cannot supply an economic optimum.
6.9 误差与协方差
Errors and covariance
乘积LAI=n×ā的误差来自密度、单茎面积、叶形系数和绿色比例。 大茎少的地块与小茎多的地块存在协变,不能总把各项误差视为相互独立。 可按小区或蔗蔸重抽样,重复计算全流程,并另外报告系统性校准误差。
Uncertainty in LAI=n×ā comes from density, stalk area, shape factors and green fractions. Density and size may covary, so independent-error assumptions are not always valid. Resample by plot or stool and recompute the full chain, while separately reporting systematic calibration uncertainty.
6.10 推荐的连续观测设计
A proposed longitudinal observation design
以下是GAKI建议的研究流程,不是已经实施的试验:按土壤和地形分区随机布置独立小区;固定长期观测茎,并另设破坏性取样带;建立期、快速扩展期和衰老期提高采样频率;每次同时记录茎密度、叶片面积、绿色比例、物候和光学读数。 频率及重复数由先导变异和研究问题确定。
Proposed GAKI workflow, not a completed experiment: block by soil and terrain, randomize independent plots, maintain longitudinal stalks and separate destructive strips, and intensify sampling around establishment, rapid expansion and senescence. At each date record density, blade area, green fraction, stage and optical readings. Sampling frequency and replication follow pilot variability and the research question.
6.11 最小数据字段与检查规则
Minimum data fields and checks
表016-05 结构化记录字段
Table 016-05. Structured recording fields
字段 / Field | 记录内容 / Content | 检查 / Check |
|---|---|---|
trial_id; plot_id | 试验及独立小区 / Trial and independent plot | 重复层级不能省 / Preserve hierarchy |
stalk_id; cohort_id | 永久茎号与分蘖世代 / Stalk and tiller cohort | 死蘖不删历史 / Retain deaths |
leaf_id; rank_basis | 永久叶号、相对叶位口径 / Permanent ID and rank convention | +1变化不换永久ID / Keep identity |
date; crop_age; stage | 日期、作龄与物候 / Date, age and stage | 新植、宿根起点注明 / Define zero |
ground_area_m2 | 水平取样地面面积 / Horizontal support | 含行间与缺株 / Include gaps |
blade_area; green_fraction | 单面面积与绿色比例 / Area and green fraction | 单位一致;g在0—1 / Consistent units |
method; calibration_id | 测量方法与校准版本 / Method and calibration | 直接/间接不混名 / Keep target |
value_kind; qc_flag | 实测、汇总、推算、缺失 / Observation, summary, estimate, missing | 缺失≠0 / Missing is not zero |
source_id; locator | 论文/数据及定位 / Source and locator | 保留原始值 / Retain original |
重复测量时,分析单位以处理分配的小区为核心;多片叶和多次扫描不能人为增加独立重复数。 遥感验证宜另设留出地块、年份或地点,防止同一冠层信息进入训练和验证两边。
For repeated measurements, the plot receiving treatment remains the main experimental unit. Additional leaves or scans do not increase treatment replication. Remote-sensing validation should hold out fields, years or sites to prevent closely related canopy information entering both sets.
016.07 真实图像与图像证据
Authentic images and image evidence
本章复用已保存原件,不生成伪实拍图,也不从背景照片提取不存在的LAI。 原作者、日期、许可与文件校验值随证据记录保存。
This chapter reuses an existing original image, generates no synthetic evidence photographs and extracts no unmeasured LAI from contextual imagery. Creator, date, license and file hash remain in the evidence record.

网页预览经过缩放或格式转换,未裁切;原比例尺与标签保留。 / Web preview resized or re-encoded without cropping; original scales and labels retained.
图016-01 甘蔗田缘与冠层实景。 该图只说明空间观测背景,不能据照片亮度、覆盖程度或叶片重叠直接计算LAI。 摄影:Kerry Raymond,2022年1月26日;CC BY 4.0。
Figure 016-01. Field-edge sugarcane canopy as spatial context. Image brightness, cover or overlapping foliage cannot directly establish LAI. Photo: Kerry Raymond, 26 January 2022; CC BY 4.0.
7.1 可用于定量面积的图像应具备什么
Requirements for quantitative area images
优先采集同一叶片的平铺标尺图、绿色/坏死分割图、对应永久叶号与时间;冠层照片应有小区边界、拍摄方向、相机高度与同步直接面积。 图像文件需保留原件,分割、裁剪或标注均作为衍生版本登记。
Prioritize scaled flat-blade images, green/necrotic masks and matching permanent leaf IDs and dates. Canopy photos require plot boundaries, viewing direction, camera height and synchronized area references. Preserve originals and register segmentation, cropping and labels as derived versions.
7.2 本章图像缺口
Image gaps in the source review
检索到2025年甘蔗杂草层与LAI反演研究的田间装置图版线索,但原图下载失败后已停止尝试。 该线索未计为已取得图片,图版许可及第三方材料权属仍需核实。 当前缺少标尺明确的逐叶展开、部分黄化和连续冠层对照照片。
A 2025 sugarcane weed-layer/LAI study provided a lead to a field-equipment figure. The original download failed and was not retried. It is not counted as acquired; figure-specific licensing and third-party credits remain unverified. Scaled blade, partial-senescence and longitudinal canopy comparison photographs remain missing.
016.08 适用边界与研究缺口
Applicability and research gaps
8.1 不设置没有环境条件的“最佳LAI”
No unconditional optimal LAI
LAI提高可以增加光截获,也可能增加维持成本、耗水和自遮阴。 是否有利取决于建立速度、保持时长、叶片功能、气象、密度和最终糖产量。 没有同一试验下的产量及投入比较,不应给出通用于世界蔗区的目标值。
Higher LAI can increase interception while adding maintenance costs, water demand and self-shading. Its value depends on establishment, duration, leaf function, weather, density and harvested sugar. A universal target is unwarranted without yield and input comparisons under matched conditions.
8.2 必须公开的证据缺口
Evidence gaps that remain explicit
四篇原始研究仅核读摘要或出版商片段;R05可读PDF文字,但页面图未核验、修订说明未取得。 R04的样本数冲突未解决。 尚未取得可重新分析的逐叶或逐小区原始数据;只有一幅背景实拍。 它们不影响保存本章研究初稿,但阻止把本章标成出版定稿。
Four primary studies were reviewed at abstract or indexed-excerpt level. R05’s PDF text was available, but page images and amendment details were not verified. R04’s sample-count conflict remains. No reanalysable leaf- or plot-level dataset was obtained, and imagery is contextual only. These do not prevent draft preservation, but preclude publication-ready status.
8.3 三项可证伪的后续问题
Three testable follow-up questions
第一,按叶位分层的长宽模型能否在未参与拟合的品种中降低误差? 第二,光学LAI与直接绿叶LAI的偏差是否随枯叶和茎秆比例改变? 第三,相同峰值LAI但不同绿叶持续期的群体,在同等供水和品种条件下是否具有不同实收糖产量? 每个问题都需要预先定义验证集与结果指标。
First, does rank-stratified length–width calibration reduce error in held-out cultivars? Second, does optical-versus-direct green-LAI bias vary with dead-leaf and stalk fractions? Third, do crops sharing peak LAI but differing in green-area duration differ in harvested sugar under matched water and genotype? Each requires predefined validation sets and outcomes.
8.4 与相邻章节的衔接
Links to neighbouring chapters
第015章提供叶角和空间分布基础;第017章继续研究透光、消光与辐射截获;第022章讨论辐射利用效率。 本章的面积口径、来源ID和校准规则应沿用,不应在后续章节把预测、观察和教学示例重新混合。
Chapter 015 supplies leaf-angle and spatial structure; Chapter 017 continues with transmission, extinction and interception; Chapter 022 addresses radiation-use efficiency. Carry forward the area definitions, source IDs and calibration rules without merging predictions, observations and teaching examples.
本章结论:LAI是叶片、茎群与时间共同形成的状态变量。 可靠研究必须把面积增加、衰老损失、取样分母和测量方法放在同一条可追溯记录链中。
Conclusion: LAI is a state variable jointly formed by blades, stalk populations and time. Reliable work connects area gain, senescent loss, sampled ground support and measurement method in one auditable record chain.
附录A 双语术语
Appendix A. Bilingual terminology
表016-06 术语对应表
Table 016-06. Terminology crosswalk
中文 | English | 中文 | English |
|---|---|---|---|
叶片面积 | Lamina area | 单面叶面积 | One-sided leaf area |
绿叶面积 | Green leaf area | 绿色比例 | Green fraction |
叶面积指数 | Leaf area index, LAI | 绿叶面积指数 | Green leaf area index, GLAI |
植物面积指数 | Plant area index, PAI | 有效叶面积指数 | Effective leaf area index |
叶面积密度 | Leaf area density | 冠层覆盖率 | Canopy cover |
叶面积持续期 | Leaf area duration | 叶形系数 | Leaf shape factor |
最大叶宽 | Maximum blade width | 叶片扩展 | Leaf expansion |
出叶 | Leaf appearance | 叶衰老 | Leaf senescence |
叶领 | Leaf collar / ligule region | 永久叶号 | Permanent leaf identifier |
相对叶位 | Relative leaf rank | 分蘖世代 | Tiller cohort |
茎密度 | Stalk density | 主茎 | Main stalk |
宿根 | Ratoon crop | 水平地面面积 | Horizontal ground area |
破坏性取样 | Destructive sampling | 亚样本 | Subsample |
独立实验单位 | Independent experimental unit | 光隙率 | Gap fraction |
聚集效应 | Clumping effect | 校准 | Calibration |
外部验证 | External validation | 面积分割掩膜 | Area segmentation mask |
重复测量 | Repeated measures | 系统误差 | Systematic error |
原始观测 | Raw observation | 发表汇总值 | Published summary value |
面积密度的分母是体积;LAI的分母是地面面积。 缩写“LAD”在不同论文中可能表示leaf area density或leaf area duration,入库时必须写全称及单位。 本章用LAD表示面积持续期。
Leaf area density has a volume denominator, whereas LAI has a ground-area denominator. “LAD” may denote density or duration in different papers; store the full name and units. Here it denotes leaf area duration.
附录B 原始研究来源与核读范围
Appendix B. Primary sources and review scope
R01 Temperature and leaf area expansion of sugarcane: integration of controlled-environment, field and model studies
Robertson, M.J.; Bonnett, G.D.; Hughes, R.M.; Muchow, R.C.; Campbell, J.A. (1998). DOI: 10.1071/PP98042
出版商摘要;Q117、控温与灌溉田间研究。 未获得完整方法,不据摘要填写独立重复数。
Publisher abstract; Q117 controlled-environment and irrigated field work. Full methods and independent replication are unresolved.
R02 Sugarcane leaf area development under field conditions in Florida, USA
Sinclair, T.R.; Gilbert, R.A.; Perdomo, R.E.; Shine, J.M. Jr.; Powell, G.; Montes, G. (2004). DOI: 10.1016/j.fcr.2003.12.005
USDA原作者机构技术摘要;四品种两年多地点研究。 引用原论文DOI,不把机构摘要另算一项研究。
Author-institution technical abstract for a four-cultivar, two-year field study. The institutional summary is not counted as another study.
R03 Leaf development and growth of selected sugarcane clones in a subtropical environment
Streck, N.A.; Hanauer, J.G.; Gabriel, L.F.; Buske, T.C.; Langner, J.A. (2010). DOI: 10.1590/S0100-204X2010001000001
已核读网页方法和结果。 叶形系数来自引用;年份与播期混杂;叶间积温表述疑点保留。
Methods and results read in HTML. Shape coefficient inherited; year/date confounded; phyllochron wording discrepancy retained.
R04 Improving indirect measurements of the leaf area index using canopy height
Gonçalves, A.O.; Silva, E.H.F.M.; Gasparotto, L.G.; Rosa, J.M.; Carmo, S.; Fattori Júnior, I.M.; Marin, F.R. (2020). DOI: 10.1590/S1678-3921.pab2020.v55.01894
已核读网页方法和结果。 混合作物汇总、样本数冲突与校准外推限制分别登记。
Methods and results read in HTML. Pooled-crop summaries, sample-count conflict and transfer limits recorded separately.
访问层级与证据类型独立记录。 本文没有把“全文可读”命名为强证据A级,也没有把摘要自动当作无效证据;每项主张按方法、材料和适用范围判断。
Access and evidence type are separate. Full-text availability is not labelled strong “Grade A” evidence, nor is every abstract dismissed. Individual claims are judged against methods, material and scope.
原始研究来源(续)
Primary sources, continued
R05 Ground reference data for sugarcane biomass estimation in São Paulo state, Brazil
Molijn, R.A.; Iannini, L.; Rocha, J.V.; Hanssen, R.F. (2018). DOI: 10.1038/sdata.2018.150
已读PDF文字;截图核验与本地全文下载未成功,未从图表读数。 数据登记号D01与本文为同一研究。
PDF text read; screenshots and local download failed. No figure/table extraction. D01 belongs to this study, not an independent validation.
R06 Simultaneous retrieval of sugarcane variables from Sentinel-2 data using Bayesian regularized neural network
Hajeb, M.; Hamzeh, S.; Alavipanah, S.K.; Neissi, L.; Verrelst, J. (2023). DOI: 10.1016/j.jag.2022.103168
已核读公开作者稿方法与结果。 随机划分验证;原始数据按请求提供,本章未取得。
Public author manuscript methods/results read. Random-split validation; original data are available on request and were not acquired here.
R07 Sugarcane water stress criteria for irrigation and drying off
Inman-Bamber, N.G. (2004). DOI: 10.1016/j.fcr.2004.01.018
只核读出版商索引摘要及预览片段。 使用定性生长响应,不转写成全球水分阈值。
Publisher-indexed abstract and preview only. Qualitative growth response used; no global water-stress threshold inferred.
R08 Temperature and seasonal effects on canopy development and light interception of sugarcane
Inman-Bamber, N.G. (1994). DOI: 10.1016/0378-4290(94)90051-5
只核读出版商索引摘要。 季节起始日期保留为设计条件,不等同于独立试验重复。
Publisher-indexed abstract only. Seasonal crop starts are design conditions, not automatically independent trials.
本章统计8篇DOI不同的原始研究,不能据此声称有8次独立验证。 研究间数据复用、处理单位和重复层级尚未全部核实。
Eight DOI-distinct primary papers are recorded, not eight confirmed independent validations. Data reuse, treatment units and replication are not fully resolved.
附录C 方法、数据与交付记录
Appendix C. Methods, data and delivery record
M01 LI-COR — LAI-2200C Theory of Operation
制造商官方测量理论;用于定义单面叶面积、视场遮挡与假设条件,不作为甘蔗增产试验证据。
Official manufacturer theory for area definitions, occlusion and assumptions; not sugarcane yield-trial evidence.
https://www.licor.com/support/LAI-2200C/topics/theory-of-operation.html
D01 关联数据登记 / Linked data registration
登记号:10.4121/uuid:37112e18-f794-4d66-a8cd-7f1e92af09fc。 由R05引用定位;落地页未成功核读,数据文件、许可与修订版本仍待获取。
Identifier: 10.4121/uuid:37112e18-f794-4d66-a8cd-7f1e92af09fc. Located in R05; landing page not read successfully. Dataset bytes, license and revision details remain unresolved.
M02 复用实拍照片 / Reused authentic photograph
摄影:Kerry Raymond;2022年1月26日;Alberton,Queensland,Australia。 许可CC BY 4.0。 沿用已核验出处并重新检查本地文件;不声称本章重新核对在线元数据。
Photo: Kerry Raymond; 26 January 2022; Alberton, Queensland, Australia. CC BY 4.0. Inherited verified provenance with local file recheck; no claim of fresh online metadata verification.
原图SHA-256:fc287e4f805052bee0d1a15c308152b87cc5bd804f4ddf208f4b908e72381fb2
许可 / License: https://creativecommons.org/licenses/by/4.0/
表016-07 数据性质与交付状态
Table 016-07. Data types and delivery status
项目 / Item | 状态 / Status |
|---|---|
原始研究 / Primary studies | 8篇:3篇HTML全文;1篇PDF文字;4篇摘要/片段 / 8 papers: 3 HTML, 1 PDF text, 4 abstract/excerpt |
结构化数值 / Structured numerical records | 45条;其中4条空值并附原因 / 45 records with explicit missing-value reasons |
原始观测文件 / Raw observation files | 0;未取得不等于没有数据 / 0 acquired; not evidence of absence |
教学计算 / Illustrations | 4组,全部明确标注假设 / Four, all explicitly hypothetical |
图像 / Images | 1幅复用实拍;1条未取得线索 / One reused photo; one unacquired lead |
出版与网站 / Publication and website |
Master merging requires a verified current version and preservation of existing work.