Table 1. Statistically evaluated models and ecological mechanisms.

This table describes initial full models (left column) and the well-documented biological phenomena (right column) that may occur in the fairy pitta and may be responsible for the associations between the explanatory and response variables in the analyses. Due to our small sample size, we could not test all possible effects and interactions. In our observational data analysis, we use the information-theoretic approach, which helps determine mechanisms contributing more strongly to the observed pattern among mutually nonexclusive mechanisms. Please note that Analysis 1 and Analysis 5 are not independent of each other because Analysis 5 uses a subset of data used in Analysis 1, and the number of prey items (Analysis 1) is correlated with the number of earthworms (Analysis 5). The model specification uses variable names defined in the Methods section. The results are shown in Tables S4–S7 and S10–S11.

Initial full statistical model: relationships evaluated in the model	Examples of multiple mutually non-exclusive mechanisms that may contribute to the relationships.
Analysis 1 (Figs. 2B; S2; Table S4): Number of prey items ∼Earthworms present or only earthworms + inter-visit interval + nestling age class + time of day + (1—nest ID) Data set: All visits [n = 200 visits] or YesE visits [n = 192 visits]	1. The number of prey items will be larger in food loads containing earthworms (YesE, compared to NoE visits in the analysis using “earthworms present”, or OnlyE, compared to MIX in the analysis using “only earthworms”) because earthworms may be relatively easier to pack due to their softness compared to other prey with exoskeletons, wings, and legs. 2. The number of prey items may increase with longer inter-visit interval due to greater nestling hunger. Additionally, longer inter-visit intervals may indicate longer foraging trips, potentially leading to more prey collected. 3. Food loads for older nestlings may contain more prey because they require more food. 4. In the morning, food loads may contain more prey due to hungry nestlings, leading to increased foraging activity in adults and resulting in a higher number of prey items in a food load.
Analysis 2 (Fig. S3A; Table S5): Average earthworm length ∼ number of earthworms + nestling age class + time of day + inter-visit interval + (1—nest ID) Data set: OnlyE visits [n = 128 visits]	1. The average earthworm length in a food-load may decrease as the number of earthworms in the food-load increases due to beak size limits, assuming that birds tend to bring a full beak-load of food with each visit. 2-1. The average earthworm length delivered to older nestlings may be longer because they require more food and can swallow large prey. 2-2. The average earthworm length may not be associated with nestling age because parents might primarily follow optimal foraging rules rather than responding to changes in the size of their nestlings. 3. The average earthworm length in the morning may be longer than in the remaining times of the day because nestlings need more food, and birds can more easily find larger ones due to their putative higher availability on the soil surface in the morning compared to later in the day. 4-1. The average earthworm length may increase as the inter-visit interval increases because larger prey may require a longer time to catch and handle prey before bringing them to the nest. 4-2. The average earthworm length may not depend on the inter-visit interval, or it may decrease with an increase in the preceding inter-visit interval. This could happen if a long interval is an indicator of poor general earthworm availability, which in turn requires longer searching and causes parents to accept even smaller earthworms.
Analysis 3 (Figs. 2D, 2E; Table S6): Earthworms present or only earthworms ∼nestling age class + inter-visit interval + *time of day + (1—nest ID) * only used in models where “only earthworms” is the response variable. Data set: All visits [n = 200 visits] or YesE visits [n = 192 visits]	1. Visits with earthworms only (OnlyE type) may be more frequent for younger nestlings than for older ones because older nestlings require more food, and parents may include non-earthworm food items to deliver a greater amount of food to older nestlings. Additionally, very young nestlings may struggle to swallow harder prey in NoE visits, and parents may adjust the food type based on nestling age, delivering mostly earthworms (soft prey) to younger nestlings. 2-1. Longer inter-visit intervals may lead to less frequent OnlyE visits, as longer trips often indicate longer foraging trips, potentially resulting in encountering a more diverse range of prey by parents. 2-2. Feeding visit type may not necessarily be predicted based on the length of the inter-visit interval, especially if longer inter-visit intervals occur due to the extended handling time for longer prey (e.g., sundering) or non-foraging activities. 3. OnlyE visits may be more frequent for nestlings in the morning than in the remaining times of the day because earthworm availability may be higher in the morning compared to later in the day (e.g., due to hot and dry conditions later in the day).
Analysis 4 (Fig. S4; Table S10): Inter-visit interval ∼nestling age class + time of day + only earthworms + rainfall category + number of prey items + (1—nest ID). Data set: YesE visits [n = 192 visits]	1-1. The inter-visit interval may be shorter for older nestlings because they may require food more frequently to maintain their growth rate. 1-2. The inter-visit interval may be longer for older nestlings if parents aim at delivering larger food loads to larger nestlings (Analysis 7 addresses it) and if collecting a larger food load takes more time. 2. The inter-visit interval may be shorter in the morning than at other times of the day because nestlings are hungrier in the morning, necessitating more frequent food deliveries. This pattern is widespread among insectivorous birds. 3-1. The inter-visit interval for OnlyE visits may be shorter compared to MIX visits because the species specializes in hunting earthworms and may be less efficient at foraging for other prey. Other prey may be targeted only when earthworms are scarce, and it is time-consuming for birds to collect enough prey for the full beak-load of food. 3-2. The inter-visit interval for OnlyE visits may be longer because selectively seeking out earthworms could take more time, as they have to search specifically for this prey. 4-1. Heavy rain may increase the inter-visit interval because it may interfere with foraging and flight. 4-2. If heavy rain does not disrupt foraging activities, then rainy weather may shorten the inter-visit interval because earthworm availability is likely higher during rainy weather. 5. The inter-visit interval may be longer for visits involving a larger number of prey items because it takes more time to collect and handle a larger number of prey items.
Analysis 5 (Fig. 3A; Table S11): Number of earthworms ∼rainfall category + inter-visit interval + nestling age class + (1—nest ID). Data set: OnlyE visits [n = 128 visits]	1-1. Heavy rain may decrease the number of earthworms per visit because it may interfere with adult foraging behaviors. 1-2. Heavy rain may increase the number of earthworms per visit because it may increase earthworm activity on the surface, assuming that adult foraging behavior is not severely affected by heavy rain. 2. see no. 2 for Analysis 1 3. see no. 3 for Analysis 1
Analysis 6 (Fig. 3C; Table S11): Biomass of a single earthworm ∼rainfall category + inter-visit interval + nestling age class + (1—nest ID/visit ID). Data set: OnlyE visits [n = 328 earthworms within 128 visits]	1-1. Heavy rain causes more earthworms to emerge, increasing their availability, including larger worms, which are likely preferred by birds. Additionally, birds may rely on sound cues to detect worms, but wet litter reduces the sound created by movements, making larger worms the only ones detectable. 1-2. Heavy rain may hinder foraging and prey handling, especially for larger worms, potentially reducing the size of an average single earthworm in a food-load. 2. Finding a large earthworm may require longer foraging time compared to finding worms of any size, which could result in a longer inter-visit interval and be associated with heavier earthworms brought to the nest. 3. Food-loads brought to older nestlings may contain heavier earthworms than those brought to smaller nestlings because older nestlings require more food and can swallow larger items more easily.
Analysis 7 (Fig. 3E; Table S11): Biomass of all earthworms per visit ∼rainfall category + inter-visit interval + nestling age class + (1—nest ID). Data set: OnlyE visits [n = 128 visits]	If birds maximize the food-load per visit, the biomass of earthworms per OnlyE visit may be not affected by the factors in the model. However, there is also a possibility that: 1. Heavy rain may increase the total biomass per visit because earthworm availability is likely higher during/after rain. 2. With longer inter-visit intervals, parents likely collect more food. Additionally, the longer the nestlings wait for food, the more food they require. 3. The total biomass per visit is expected to be larger for older nestlings because they require more food.