UFAM 校園UFAM campus
纜索／網狀橋；有 Saguinus bicolor 使用記錄。Cable / net bridges; use by Saguinus bicolor recorded.
- 接近次數Approaches
- 未報告Not reported
- 完整通過總數Total full crossings
- 未報告Not reported
- 標準化監測時長Standardised monitoring time
- 未報告Not reported
步數Step 0 / 1500種子Seed —已暫停
接近Approaches0
進入Entries0
完整通過Full crossings0
退回Retreats0
改道Reroutes0
等待事件Wait events0
已在對岸On far side0
以上均為目前規則與參數下的模擬計數，用於比較情境，不是真實使用機率，也不代表任何實際橋梁的表現。All figures above are simulated counts under the current rules and parameters, meant for comparing scenarios. They are not real-world use probabilities and do not represent the performance of any actual bridge.
執行Run 空白鍵 / R / SSpace / R / S
相同種子 + 相同參數 = 完全相同的試驗，便於重現與對比。Same seed + same parameters = an identical trial, for reproducibility and comparison.
模擬時標以「步」計，未對應真實秒數或分鐘。Simulation time is counted in steps and is not mapped to real seconds or minutes.
假設Hypothesis
修改後自動重置。群體規模在此僅作情境設定，未依據特定群體資料。Changing this resets automatically. Group size is only a scenario setting here and is not based on data from a specific group.
空白鍵 執行/暫停 · R 重置 · S 單步（焦點在場景或頁面時）Space run/pause · R reset · S single step (when focus is on the scene or page)
情境Scenario 每項標註證據等級evidence level marked per item
物種證據 黑白花狨本身
近緣類群 其他狨科
假設 待驗證
Species evidence pied tamarin itself
Related taxa other callitrichids
Hypothesis to be tested
廊橋Bridge 物種證據Species evidence
已有記錄顯示物種會使用人工橋；但「會用」不等於本設計的成功率。Records show the species does use artificial bridges, but “does use” is not the same as this design’s success rate.
出發端橋口連接Origin-side entrance connection 場景規則 · 未現場測量Scenario rule · not field-measured
到達端橋口連接Arrival-side entrance connection 場景規則 · 未現場測量Scenario rule · not field-measured
斷開時沒有樹冠路徑：個體只能停在枝端觀察，不能跨越空氣。第一版不模擬跳躍。With a gap there is no canopy path: agents can only stop at the branch tip and look; they cannot cross open air. Version one does not simulate leaping.
近緣類群候選因素Related-taxa candidate factor本物種係數未知Species coefficient unknown
百分比是情境強度刻度，不是實測葉面積或遮蔽度。The percentage is a scenario intensity scale, not measured leaf area or canopy cover.
相關研究 · 效應待驗證Related study · effect untested強度為假設Strength hypothesised
百分比是情境強度刻度，不是實測果實量。The percentage is a scenario intensity scale, not a measured fruit quantity.
相關研究 · 效應待驗證Related study · effect untested強度為假設Strength hypothesised
噪聲同時決定車輛事件頻率與個體受驚傾向。情境參數在執行中即時生效；需要嚴格對比時請重置。Noise sets both how often vehicle events occur and how readily agents startle. Scenario parameters take effect immediately while running; reset when you need a strict comparison.
百分比是情境強度刻度，不是實測分貝（dB）。The percentage is a scenario intensity scale, not measured decibels (dB).
權重Weights 全部待驗證All untested
權重組Weight set
三組是互相競爭的假設，不是估計值。手動拖動滑桿即為「自訂」。The three sets are competing hypotheses, not estimates. Moving a slider by hand makes the set “Custom”.
假設Hypothesis
假設Hypothesis
假設Hypothesis
假設Hypothesis
z = −0.8 + wR(2R−1) + wC(2C−1) − wN(2N−1)
− 1.2·wN·[車輛經過] + wS(2·已過比例 + 0.6·[橋上有同伴])
+ 橋口項（連續 +0.3 / 稀疏 −0.9）
每步進入機率 = 0.05 · σ(z) · 斷開時為 0
− 1.2·wN·[車輛經過] + wS(2·已過比例 + 0.6·[橋上有同伴])
+ 橋口項（連續 +0.3 / 稀疏 −0.9）
每步進入機率 = 0.05 · σ(z) · 斷開時為 0
z = −0.8 + wR(2R−1) + wC(2C−1) − wN(2N−1)
− 1.2·wN·[vehicle passing] + wS(2·share already crossed + 0.6·[companion on bridge])
+ entrance term (continuous +0.3 / sparse −0.9)
Per-step entry probability = 0.05 · σ(z) · 0 when there is a gap
− 1.2·wN·[vehicle passing] + wS(2·share already crossed + 0.6·[companion on bridge])
+ entrance term (continuous +0.3 / sparse −0.9)
Per-step entry probability = 0.05 · σ(z) · 0 when there is a gap
進入公式、社會跟隨、A/B/C 權重，以及 0.05、−0.8、1.2 等常數，全部是演示假設，不是估計值。The entry formula, social following, the A/B/C weights and constants such as 0.05, −0.8 and 1.2 are all demonstration assumptions, not estimates.
試驗與對比Trials & comparison 相同種子序列same seed sequence
尚未設定基準。No baseline set yet.
結果會顯示在這裡：平均接近次數、平均完整通過次數，以及出現過通過的試驗數。Results appear here: mean approaches, mean full crossings, and the number of trials with at least one crossing.
從真實記錄，到可檢驗的假設From recorded use to testable hypotheses
這個原型把「接近、觀察、進入、通過、等待、退回、改道」寫成可檢查的規則，讓每一個假設都能被看見、被替換、被現場資料推翻。This prototype writes “approach, observe, enter, cross, wait, retreat, reroute” as inspectable rules, so that every assumption can be seen, swapped out, and overturned by field data.
A · 真實過橋記錄A · Recorded bridge use
以下為已發表文獻中的公開描述（2022 年發表時點），不是本站的現場資料，也未用於校準本模擬。來源：The following are public descriptions from published literature (as of the 2022 publication), not this site’s field data, and are not used to calibrate this simulation. Source: DOI 10.1163/14219980-20211210 · 全文（ResearchGate）Full text (ResearchGate)
Ponta Negra · 兩座尼龍繩梯橋Ponta Negra · two nylon rope-ladder bridges
有機會性觀察到的使用記錄；同時報告了結構問題——使用不等於安全。Opportunistic observations of use recorded; structural problems were also reported — use does not mean safe.
- 接近次數Approaches
- 未報告Not reported
- 完整通過總數Total full crossings
- 未報告Not reported
- 標準化監測時長Standardised monitoring time
- 未報告Not reported
REMAN 道路 · 2011 年安裝三座橋REMAN road · three bridges installed in 2011
有多個黑白花狨群體的使用記錄。「三座」是橋的數量，不是通過次數。Use by several pied tamarin groups recorded. “Three” is the number of bridges, not the number of crossings.
- 接近次數Approaches
- 未報告Not reported
- 完整通過總數Total full crossings
- 未報告Not reported
- 標準化監測時長Standardised monitoring time
- 未報告Not reported
「未報告」表示公開描述沒有提供，不等於 0。橋的數量不能當作通過次數，這些記錄也不能換算為經驗使用率。本站不轉載文獻中的版權照片。“Not reported” means the public description does not provide it; it is not 0. Bridge counts must not be read as crossing counts, and these records cannot be converted into an empirical use rate. This site does not reproduce copyrighted photos from the literature.
B · 證據與假設B · Evidence and assumptions
| 項目 | 證據狀態 | 本模擬如何使用 |
|---|---|---|
| 使用某些人工橋 | 物種證據 本物種有記錄（2022 綜述） | 作為「可能使用」的前提；不推斷任何設計的成功率 |
| 橋口連接幾何（連續／稀疏／斷開） | 場景規則 未現場測量 | 斷開時沒有樹冠路徑，完整通過恆為 0 |
| 橋面遮蔽 | 近緣類群 僅來自其他狨科研究的候選因素；本物種橋口係數未知 | 滑桿百分比是情境刻度，不是葉面積 |
| 對岸資源、道路噪聲 | 相關研究 本物種有移動生態研究（2026）；對過橋的效應方向與大小未校準 | 滑桿百分比不是實測 dB 或果實量 |
| 社會跟隨、進入公式、A/B/C 權重、0.05 等常數 | 演示假設 全部待驗證 | 只用於比較情境，不是估計值 |
| Item | Evidence status | How this simulation uses it |
|---|---|---|
| Use of some artificial bridges | Species evidence recorded for this species (2022 review) | Premise that use is possible; no design success rate is inferred |
| Entrance connection geometry (continuous / sparse / gap) | Scenario rule not field-measured | With a gap there is no canopy path; full crossings stay at 0 |
| Bridge cover | Related taxa candidate factor from other callitrichids only; species coefficient at entrances unknown | Slider percentage is a scenario scale, not leaf area |
| Far-side resources, road noise | Related study movement ecology study of this species (2026); direction and size of effect on crossing not calibrated | Slider percentages are not measured dB or fruit quantity |
| Social following, entry formula, A/B/C weights, constants such as 0.05 | Demonstration assumptions all untested | Used only to compare scenarios; not estimates |
模擬輸出（計數與模擬事件比值）不是文獻記錄，也不是經驗機率；本頁不提供任何可信度評分。Simulation outputs (counts and the simulated event ratio) are not literature records and not empirical probabilities; this page gives no confidence score.
C · 驗證還需什麼資料C · Data needed for validation
每一筆紀錄應是一次獨立的接近事件，至少包含：Each record should be one independent approach event, containing at least:
- 獨立接近事件 ID
approach_event_id - 日期與開始／結束時間
date, time_start, time_end - 橋 ID
bridge_id - 個體／群體 ID，無法辨識填 unknown
individual_or_group_id - 結果：接近／進入／退回／完整通過
outcome - 結構狀態與兩端連接狀態
structure_state, connection_* - 相機有效工作時間
camera_effective_* - 漏檢與遮擋紀錄
missed_detection_notes
- Independent approach event ID
approach_event_id - Date and start / end time
date, time_start, time_end - Bridge ID
bridge_id - Individual / group ID, or unknown
individual_or_group_id - Outcome: approach / enter / retreat / full crossing
outcome - Structure state and connection state at both ends
structure_state, connection_* - Camera effective operating time
camera_effective_* - Missed detections and occlusions
missed_detection_notes
- 同一群體的重複觀測不是獨立的動物，分析時須標示群體並處理重複。
- 「完整通過 ÷ 被偵測到的接近」只有在監測方式一致（相同相機配置、有效時間與漏檢處理）時才可解釋；即使比值高，也不代表族群恢復。
- Repeated observations of the same group are not independent animals; analyses must label groups and handle repeats.
- “Full crossings ÷ detected approaches” can only be interpreted under consistent monitoring (same camera set-up, effective time and missed-detection handling); even a high ratio does not mean population recovery.
空白模板只有欄位名稱，不是真實資料，也不含任何虛構樣本；檔案在你的瀏覽器本地產生。The blank template contains column names only; it is not real data and contains no fabricated samples. The file is generated locally in your browser.
模型卡Model card
規則摘要（按 ODD 結構）Rule summary (ODD structure)
目的：比較橋口連接、遮蔽、資源、噪聲情境對「接近—通過」計數的相對影響，暴露需要現場驗證的假設。
Purpose: compare the relative effects of entrance connection, cover, resources and noise on “approach–crossing” counts, and expose assumptions that need field validation.
實體：個體（位置、狀態、冷卻計時）、樹冠節點、橋面（0–1 位置）、車輛事件。
Entities: agents (position, state, cooldown timer), canopy nodes, bridge deck (position 0–1), vehicle events.
過程與調度：每步先生成或推進車輛事件，再依次更新每個個體。覓食個體以與資源相關的低機率轉為接近；抵達橋口記一次接近；觀察期內按 σ(z) 判定進入；超時則退回（稍後可再接近）或改道（較長時間離開）。
Process & scheduling: each step first spawns or advances vehicle events, then updates each agent in turn. Foraging agents switch to approaching with a low, resource-dependent probability; reaching the bridge entrance counts as one approach; during observation, entry is decided by σ(z); on timeout they retreat (and may approach again later) or reroute (leaving for longer).
橋面：車輛經過時按噪聲與遮蔽決定是否進入等待；等待過久按所處位置決定繼續或退回；到達端斷開時只能退回。
Bridge deck: when a vehicle passes, noise and cover decide whether an agent starts waiting; after waiting too long, its position decides whether it continues or retreats; if the arrival side has a gap, it can only retreat.
初始化：全部個體位於出發樹冠的隨機節點，由種子決定。
Initialisation: all agents start at random nodes in the origin canopy, determined by the seed.
輸出：接近、進入、完整通過、退回、改道、等待事件次數。
Output: counts of approaches, entries, full crossings, retreats, reroutes and wait events.
資料缺口Data gaps
- 橋口逐事件（秒級）行為序列：本站尚未取得，包括接近方向、停留時長、退回時機。
- JEB 2024 公開資料庫已定位，檔案尚未核驗與匯入。
- 聲景 Numbers 資料只測量交流發聲，不含過橋行為，不用於校準過橋。
- 水平跳躍資料：本站尚未取得，因此本版不模擬跳躍，也不給出安全跳距。
- 噪聲與資源的效應依情境而變，無法給出通用係數。
- 10 分鐘間隔 GPS 無法校準橋口秒級行為。
- 近緣類群研究只提供候選變數，不能借用其係數。
- 「遮蔽」與「連接」尚無統一的現場度量定義。
- Per-event (second-scale) behaviour sequences at bridge entrances: not yet obtained by this site, including approach direction, dwell time and retreat timing.
- The JEB 2024 public data repository has been located; its files have not yet been verified or imported.
- The soundscape Numbers data measure communication calls only, contain no bridge crossings, and are not used to calibrate crossing.
- Horizontal leaping data: not yet obtained by this site, so this version does not simulate leaps and gives no safe leap distance.
- Noise and resource effects vary with context; no general coefficients can be given.
- GPS fixes at 10-minute intervals cannot calibrate second-scale behaviour at bridge entrances.
- Studies of related taxa only suggest candidate variables; their coefficients cannot be borrowed.
- “Cover” and “connection” have no agreed field measurement definitions yet.
後續現場驗證計畫Planned field validation
- 參照 Reconecta 的橋面與接近方向監測方式，擬定橋口觀測協議，以秒為單位記錄接近、觀察、進入、通過、等待、退回。Drawing on Reconecta’s monitoring of bridge decks and approach directions, draft a bridge-entrance observation protocol that records approach, observation, entry, crossing, waiting and retreat to the second.
- 核驗 JEB 公開資料庫中的檔案與變數含義，再判斷是否可用於任何參數；不用於推算水平跳距。Verify the files and variable meanings in the JEB public repository before judging whether they can inform any parameter; they will not be used to infer horizontal leap distance.
- 用現場序列同時檢驗 A/B/C 等替代權重，接受或拒絕規則，並報告不確定性。Use field sequences to test alternative weights such as A/B/C side by side, accept or reject rules, and report uncertainty.
- 校準前後的版本分開標註；在通過次數之外，另行評估連通與族群層面的指標。Label pre- and post-calibration versions separately; beyond crossing counts, assess connectivity and population-level indicators separately.
研究依據Research basis
2022 · 巴西樹冠橋綜述2022 · Review of canopy bridges in Brazil 物種證據Species evidence
記錄黑白花狨使用人工橋。支持「物種會使用人工橋」這一前提；不能據此推斷任何設計的成功率。
Records pied tamarins using artificial bridges. Supports the premise that “the species uses artificial bridges”; it cannot be used to infer the success rate of any design.
2024 · JEB 向上跳實驗2024 · JEB upward-leap experiment 物種證據Species evidence
該實驗中黑白花狨僅兩隻，且為向上跳。不能換算為水平安全跳距，因此本版不模擬跳躍。
Only two pied tamarins took part in this experiment, and the leaps were upward. This cannot be converted into a safe horizontal leap distance, so this version does not simulate leaping.
公開資料庫已定位，檔案尚未核驗匯入：Public data repository located; files not yet verified or imported: DOI 10.7924/r47m0hf0g
2026 · 移動生態（九個群體）2026 · Movement ecology (nine groups) 物種證據·依情境Species evidence · context-dependent
噪聲與資源的效應依情境而定，因此本模型只把它們列為方向性變數，強度為假設。10 分鐘間隔 GPS 無法校準橋口秒級行為。
Noise and resource effects depend on context, so this model treats them only as directional variables whose strengths are hypothesised. GPS at 10-minute intervals cannot calibrate second-scale behaviour at bridge entrances.
2024 · 其他狨科類群2024 · Other callitrichid taxa 近緣類群Related taxa
研究對象為其他狨科類群。只支持選取行為候選變數（如遮蔽），不借用其係數。
The study subjects were other callitrichid taxa. It only supports choosing candidate behavioural variables (such as cover); its coefficients are not borrowed.
Reconecta · 人工樹冠橋監測Reconecta · Artificial canopy bridge monitoring 方法參考Method reference
其橋面與接近方向的監測方式可作為觀測協議的參考；不是現成的訓練資料。
Its way of monitoring bridge decks and approach directions can serve as a reference for the observation protocol; it is not ready-made training data.
ODD 協議ODD protocol 規則說明格式Rule description format
用於組織本模型的規則說明（目的、實體、過程、調度、初始化、輸出）。
Used to structure this model’s rule description (purpose, entities, processes, scheduling, initialisation, output).
過橋，不等於族群恢復。Crossing a bridge is not population recovery.
這一版不聲稱的事What this version does not claim
- 不預測真實使用機率；所有數字都是規則模擬的計數。
- 不是已訓練的預測型 world model，未經任何資料校準。
- 不預測水平跳躍極限；沒有樹冠或橋面連接，就無法跨越。
- 不代表任何已建成或計畫建設的橋，也不代表任何合作項目。
- 模擬事件比值不是經驗使用率；文獻中的橋數量也不是通過次數。
- It does not predict real-world use probability; every number is a count from a rule-based simulation.
- It is not a trained predictive world model and has not been calibrated with any data.
- It does not predict horizontal leap limits; without a canopy or bridge connection, there is no crossing.
- It does not represent any built or planned bridge, nor any partnership project.
- The simulated event ratio is not an empirical use rate; bridge counts in the literature are not crossing counts.