Experimental evidence for flexibility in provisioning behaviour in a cooperatively breeding bird

1. Paper citation: María Espinosa-Colín, Carlos de la Cruz, and Jorge S. Gutiérrez (2026) Experimental evidence for flexibility in provisioning behaviour in a cooperatively breeding bird. Behavioral Ecology (in press). 2. Summary: The flexible-investment hypothesis posits that parents should allocate their resources according to offspring needs. The available evidence suggests that this is valid in most avian biparental care systems, yet experiments designed to quantify the degree to which caregivers respond to the demands of offspring are rare in cooperative breeding systems, where offspring are reared by individuals additional to the breeding pair (i.e., helpers). By cross-fostering Iberian magpie (Cyanopica cooki) chicks of different ages, we tested whether breeders and helpers adjust their reproductive investment to meet the demands of their current brood. We show that both breeders and helpers (all males) prolonged their nestling care period when raising chicks younger than their own and shortened it when raising chicks older than their own, with no difference in reproductive success. Moreover, parents and helpers generally adjusted their feeding rates according to chick developmental stage: provisioning increased when broods were experimentally shortened (receiving younger chicks) and decreased when broods were prolonged (receiving older chicks). Planned contrasts indicated that breeders—particularly females in the shortening experiment and males in the prolonging experiment—adjusted their care more strongly than helpers, suggesting that carers of different sexes differ in their flexible responses to brood age. Altogether, these findings support the flexible-investment hypothesis and suggest that helpers’ feeding rules are less flexible than those of breeders. Combined with previous observational work, our results are consistent with the idea that helping behaviour may function as a signal to gain direct benefits through reciprocity. 3. Originator: Jorge S. Gutiérrez. 4. Contact information: Jorge S. Gutiérrez (E-mail addresses: [e-mail]), Department of Anatomy, Cell Biology and Zoology, Faculty of Sciences, University of Extremadura, Badajoz, Spain. 5. Date of data colletion: This experiment was performed during the 1998, 1999 and 2018 breeding seasons (from March to July). 6. Geographic location(s) of data collection: 39º 4' 17.67'' N, 6º 51' 6.39'' W 7. Information about funding sources that supported the collection and curation of the data: NA ACCESS INFORMATION 1. Licenses/restrictions placed on the data Creative Commons Attribution 4.0 International 2. Data derived from other sources: NA DATA FILES AND VARIABLES Data files and the variables that they contains: 1. data_provisioning_clean.csv: contains invididual provisioning (# hourly feeds) data from experimental (prolonged or shortened) and control broods. These data are used in the main analysis on chick provisioning. Variables included in this file: year = year of study nest = nest ID laying_date = laying date of a given nest laying_date_std = standardised laying date (i.e., days after 1 April) eggs = number of eggs chicks = number of chicks fledgings = number of fledgings feed_date = date of a given feed feed_day = days post-hatching when a give feed occurred feed_day day_since_hatching = days after hatching for a given feed status = female, male or helper ring = ring code euring_age = EURING age (see https://www.bto.org/sites/default/files/u17/downloads/about/resources/agecodes.pdf). Note that in the analyses we defined two classes of age: first-year birds and adults helpers = number of helpers help = yes/no feeds = number of feeds/h feeds_per_chick = feeds per chick in a given brood experimental_group = control, prolonged or shortened broods period = before/after swap selected_days = For control broods, we selected the same days to be compared with the respective experimental broods and thus make observations fully comparable group = control, prolonged or shortened broods (accounting for observations fully comparable, see above) nest_val = filter "nest_val = 1" select only one (fixed) value per individual nest; this is just to check whether the number of chicks and fledglings differed among shortened, prolonged and control broods before swapping. 2. data_morphometrics.csv: contains fledgings morphometric measurements from experimental (prolonged or shortened) and control broods. This data are used to assess the effect of treatment on chick development and body condition before fledging. Variables included in this file: year = year of study nest = nest ID laying_date = laying date of a given nest laying_date_std = standardised laying date (i.e., days after 1 April) fledgings = number of fledgings experimental_group = control, prolonged or shortened broods period = before/after swap selected_days = for control broods, we selected the same days to be compared with the respective experimental broods and thus make observations fully comparable group = control, prolonged or shortened broods (accounting for observations fully comparable, see above) help = yes/no chick_ID = chick ring code wing = wing cord length (mm) tarsus = tarsus length (mm), from the middle of the mid-tarsal joint to the distal end of the tarsometatarsus mass = body mass (g) sex = female or male days_measure = days old when measured mass_valid = 0/1 (1 if the individual was measured on days 11-13) 3. data_stay_clean.csv: contains data on rearing period length to test for differences in parental care period between treatments. Variables included in this file: year = year of study nest = nest ID laying_date = laying date of a given nest laying_date_std = standardised laying date (i.e., days after 1 April) eggs = number of eggs chicks = number of chicks fledgings = number of fledgings period = before/after swap selected_days = For control broods, we selected the same days to be compared with the respective experimental broods and thus make observations fully comparable group = control, prolonged or shortened broods (accounting for observations fully comparable, see above) help = yes/no hatching_date = hatching date fledging_date = fledging date stay_at_exp_nest = number of days at the experimental nest hatching_date_original_nest = hatching date at the original nest fledging_age = fledging age (d) CODE SCRIPT code_BEHECO.R: R code to reproduce the analyses presented in the paper cited above. R version 3.6.2 (2019-12-12) Platform: x86_64-apple-darwin15.6.0 (64-bit) Running under: macOS 10.16 Matrix products Attached packages: rstatix_0.7.0; cowplot_1.0.0; EnvStats_2.4.0; viridis_0.6.2; viridisLite_0.4.0; forcats_0.4.0; stringr_1.4.0; dplyr_1.1.4; purrr_0.3.3; readr_2.1.4; tidyr_1.1.3; tibble_3.2.1; tidyverse_1.3.0; ggsignif_0.6.0; car_3.0-6; carData_3.0-3; DHARMa_0.4.5; performance_0.9.1; interactions_1.1.3; emmeans_1.4.6; multcomp_1.4-15; TH.data_1.0-10; MASS_7.3-51.4; survival_3.1-8; mvtnorm_1.1-1; glmmTMB_1.0.2.1; nlme_3.1-142; lmerTest_3.1-1; lme4_1.1-21; Matrix_1.2-18; ggeffects_1.5.0; ggplot2_3.3.6; boot_1.3-23; MuMIn_1.43.15; pscl_1.5.5; reshape_0.8.81. 论文引用：María Espinosa-Colín、Carlos de la Cruz与Jorge S. Gutiérrez（2026），《合作繁殖鸟类育幼行为灵活性的实验证据》，《行为生态学》（已录用待刊）。 2. 摘要：灵活投资假说（flexible-investment hypothesis）提出，亲本应根据后代需求分配抚育资源。现有证据表明，该假说在多数鸟类双亲育幼系统中成立，但在合作繁殖系统——即后代由繁殖对之外的个体（即协助者）共同抚育的系统——中，旨在量化抚育者对后代需求响应程度的实验仍较为匮乏。本研究通过跨窝交换（cross-fostering）不同日龄的伊比利亚灰喜鹊（Cyanopica cooki）雏鸟，验证繁殖个体与协助者是否会根据当前育幼群的需求调整繁殖投资。结果显示，繁殖个体与协助者（均为雄性）在抚育较自身雏鸟更年轻的幼鸟时，会延长育幼时长；而抚育较自身雏鸟更年长的幼鸟时，则会缩短育幼时长，且繁殖成功率无显著差异。此外，亲本与协助者通常会根据雏鸟发育阶段调整投喂速率：当实验性缩小育幼群（接收更年轻雏鸟）时，投喂量增加；当实验性扩大育幼群（接收更年长雏鸟）时，投喂量减少。计划对照分析表明，繁殖个体——尤其是缩短育幼实验中的雌性与延长育幼实验中的雄性——相较于协助者，其育幼行为的调整幅度更为显著，这提示不同性别的抚育者对育幼群日龄的灵活响应存在差异。综合来看，上述研究结果支持灵活投资假说，并表明协助者的投喂规则较繁殖个体更缺乏灵活性。结合此前的观测研究，本结果与“协助行为可作为通过互惠获取直接收益的信号”这一观点一致。 3. 数据发起者：Jorge S. Gutiérrez。 4. 联系方式：Jorge S. Gutiérrez（电子邮箱：[e-mail]），西班牙埃斯特雷马杜拉大学理学院解剖学、细胞生物学与动物学系，巴达霍斯。 5. 数据采集时间：本实验于1998、1999及2018年的繁殖季（3月至7月）开展。 6. 数据采集地理位置：北纬39°4′17.67″，西经6°51′6.39″。 7. 数据采集与整理的资助来源信息：无。 ACCESS INFORMATION 1. 数据许可/限制：知识共享署名4.0国际许可协议（Creative Commons Attribution 4.0 International）。 2. 第三方衍生数据：无。 DATA FILES AND VARIABLES 数据文件及其包含的变量： 1. data_provisioning_clean.csv：包含实验性延长育幼时长、缩短育幼时长组与对照组的个体投喂数据（每小时投喂次数），用于雏鸟投喂行为的核心分析。该文件包含以下变量： year：研究年份 nest：巢箱编号 laying_date：特定巢箱的产卵日期 laying_date_std：标准化产卵日期（即4月1日后的天数） eggs：产卵数 chicks：雏鸟数 fledgings：出飞雏鸟数 feed_date：单次投喂日期 feed_day：某次投喂对应的孵化后天数 feed_day day_since_hatching：单次投喂对应的孵化后天数 status：个体身份，包括雌性、雄性或协助者 ring：环志编码 euring_age：EURING年龄编码（详见https://www.bto.org/sites/default/files/u17/downloads/about/resources/agecodes.pdf）。注：本分析中将年龄划分为两类：当年幼鸟与成鸟 helpers：协助者数量 help：是否存在协助者（是/否） feeds：每小时投喂次数 feeds_per_chick：对应育幼群中每只雏鸟的投喂次数 experimental_group：对照组、延长育幼组或缩短育幼组 period：交换前后 selected_days：对于对照组，选取与对应实验组相匹配的观测日期，以保证观测结果具有完全可比性 group：对照组、延长育幼组或缩短育幼组（考虑到观测结果的完全可比性，详见上文） nest_val：筛选条件"nest_val = 1"将仅选取每个巢箱的一个固定值；该字段用于验证交换前缩短、延长与对照组育幼群的雏鸟数与出飞雏鸟数是否存在差异。 2. data_morphometrics.csv：包含实验性延长、缩短育幼组与对照组出飞雏鸟的形态测量数据，用于评估处理对出飞前雏鸟发育与身体状况的影响。该文件包含以下变量： year：研究年份 nest：巢箱编号 laying_date：特定巢箱的产卵日期 laying_date_std：标准化产卵日期（即4月1日后的天数） fledgings：出飞雏鸟数 experimental_group：对照组、延长育幼组或缩短育幼组 period：交换前后 selected_days：对于对照组，选取与对应实验组相匹配的观测日期，以保证观测结果具有完全可比性 group：对照组、延长育幼组或缩短育幼组（考虑到观测结果的完全可比性，详见上文） help：是否存在协助者（是/否） chick_ID：雏鸟环志编码 wing：翼弦长度（mm） tarsus：跗跖长度（mm），即从跗跖关节中点至跗跖骨远端的长度 mass：身体质量（g） sex：雌性或雄性 days_measure：测量时的雏鸟日龄 mass_valid：0/1（若个体在11-13日龄时被测量，则取值为1）。 3. data_stay_clean.csv：包含育幼时长数据，用于验证不同处理组间亲本育幼时长的差异。该文件包含以下变量： year：研究年份 nest：巢箱编号 laying_date：特定巢箱的产卵日期 laying_date_std：标准化产卵日期（即4月1日后的天数） eggs：产卵数 chicks：雏鸟数 fledgings：出飞雏鸟数 period：交换前后 selected_days：对于对照组，选取与对应实验组相匹配的观测日期，以保证观测结果具有完全可比性 group：对照组、延长育幼组或缩短育幼组（考虑到观测结果的完全可比性，详见上文） help：是否存在协助者（是/否） hatching_date：孵化日期 fledging_date：出飞日期 stay_at_exp_nest：在实验巢箱的停留时长 hatching_date_original_nest：原巢箱的孵化日期 fledging_age：出飞日龄（天）。 CODE SCRIPT code_BEHECO.R：用于复现上述引用论文中分析过程的R语言代码。 R版本：3.6.2（2019-12-12） 运行平台：x86_64-apple-darwin15.6.0 (64-bit) 操作系统：macOS 10.16 加载包：rstatix_0.7.0; cowplot_1.0.0; EnvStats_2.4.0; viridis_0.6.2; viridisLite_0.4.0; forcats_0.4.0; stringr_1.4.0; dplyr_1.1.4; purrr_0.3.3; readr_2.1.4; tidyr_1.1.3; tibble_3.2.1; tidyverse_1.3.0; ggsignif_0.6.0; car_3.0-6; carData_3.0-3; DHARMa_0.4.5; performance_0.9.1; interactions_1.1.3; emmeans_1.4.6; multcomp_1.4-15; TH.data_1.0-10; MASS_7.3-51.4; survival_3.1-8; mvtnorm_1.1-1; glmmTMB_1.0.2.1; nlme_3.1-142; lmerTest_3.1-1; lme4_1.1-21; Matrix_1.2-18; ggeffects_1.5.0; ggplot2_3.3.6; boot_1.3-23; MuMIn_1.43.15; pscl_1.5.5; reshape_0.8.8