Introduction

Life histories of birds include costly processes such as reproduction, molt, and/or migration (e.g., birds breeding in temperate and polar zones), determining their annual life cycles presumably with minimum of energetic stress (Bridge, 2011; Cornelius et al., 2011; Payne, 1972). Replacement of old and worn feathers through molting processes plays a crucial role in arranging functions for future events such as flights, thermoregulation, species recognition, sexual display, and/or camouflage (Amadon, 1966). In a long-term perspective, climate change such as global warming may influence avian phenological processes with respect to reproduction and migration schedules, including feather molt through various physiological mechanisms (Kiat et al., 2019a; 2019b). However, empirical data and knowledge on avian molt of various species and contexts are relatively insufficient due to limitations in examining variations of avian molt along their life cycles compared to studies of other life history traits (e.g., reproduction) of many species.

The evolution of organisms encompasses the optimization of resource allocation strategies with minimization of fitness-related costs (Roff, 1992; Stearns, 1992). Variation in resource allocation can be shaped by trade-offs among several life history traits, including growth versus molt (Kiat & Izhaki, 2016), growth versus reproduction (Kiat & Sapir, 2018), reproduction versus molt (Svensson & Nilsson, 1997), molt versus migration (Gordo, 2007), and life history contrasts between residents and migrants in a given environment (Kiat et al., 2019a). To date, results of many studies support the adaptive significance of resource allocations where investing too many resources into a certain trait or overlapping two traits are likely to cause current and/or future fitness-related costs. Therefore, the differences between post-juvenile’s and adult’s molt followed by distinction in juvenile-adult feather quality can be explained by resource allocation under trade-offs among growth, reproduction, and molt of birds along their developmental stages (Kiat & Izhaki, 2016; 2018).

Hall and Gwinner (1987) has discovered that the degree of age-related molt is higher in young birds than in adults of white storks (Ciconia ciconia) during the spring-summer period probably due to the cost of overlapping or upcoming reproduction and migration. Adult storks with relatively lighter molt breed at the similar time window of molt and migrate longer distance than juvenile storks with low costs of reproduction and/or migration. However, little is known about characteristics of wing feathers and patterns of oriental storks (Ciconia boyciana) known to be closely related to white storks (Elliott et al., 2020). The purpose of this study was to determine characteristics of black-colored wing feathers’ structures in oriental storks. Results from known wing feathers of two specimens were utilized as references. We then examined variation in age-related wing molt using collection and classification of wing feather losses in captivity. We also discussed molt patterns of oriental storks (i.e., juvenile vs. adult) in a life-history perspective with resource allocation.

Materials|Methods

Results

Averages of all wing feathers (i.e., vane and calamus structures) with 5 known classes and within-class orders (i.e., 77 types) from two specimens were calculated as references (Supplementary Table 1). First, PC score for vane structure was representative for the degree of vane asymmetry (eigenvalue=3.29). PC score increased with decreasing vane asymmetry found in the trailing (wider, more gradual)-leading (narrower, steeper) edge of vanes (Fig. 2A). Here, the degree of vane asymmetry ranged from P, PC, A, SC, S, and T in order. Second, PC score for calamus structure was representative for the degree of rigid attachment of feathers to the wing skin (eigenvalue=4.39). PC score increased as the calamus was longer and thicker with deeper skin level (Fig. 2B). Here, the degree of calamus rigidness ranged from P, S, T, PC, A, and SC in order.

All wing feathers (n=3,807) collected for a year were classified into six types. As shown in Fig. 3, seasonal feather loss (i.e., per capita, the number of feather losses divided by the number of individuals in each cage) appeared to start earlier and become seasonally more expanded in juvenile storks (i.e., hatch year to 2nd year) than adult storks (i.e., after 3rd year) with increasing age. The annual wing feather loss differed by feather type and age (mixed model: FEATHER F5,18=3.33, P=0.03; AGE F1,18=7.80, P=0.01; FEATHER×AGE F5,18=0.40, P=0.84; Fig. 4). Specifically, the proportion of wing feather losses varied for the six types: A (31%), P (41%), PC (29%), S (43%), SC (34%), and T (59%). The percentage of all feather losses significantly decreased with age (slope=–4.45±1.96).

Discussion

We documented that the degree of wing feather molt in young birds was at a higher rate than that in adults presumably due to the cost of reproduction in adults of oriental storks. The present study partially replicated the study of Hall and Gwinner (1987) for white storks using a non-invasive method based on annual wing feather loss for molting inspection (Rohwer & Brom, 2012). Our results indicated that wing feather loss annually decreased with increasing age (Fig. 4) where juveniles (i.e., hatch year to 3rd year) tended to molt earlier and temporally elongated compared to adults (i.e., after 3rd year) with respect to seasonal variation in wing feather molt (Fig. 3). Our results were comparable to those of a previous study (Hall & Gwinner, 1987), suggesting that differences in age-specific molt found in oriental storks might be associated with the cost of reproduction in adults.

In a life history perspective, resource allocation strategies are necessary under various trade-offs among self-maintenance (e.g., growth and molt), reproduction, and other demands depending upon the ontogenic stage of early life in birds (Roff, 1992; Stearns, 1992). Here, fitness-related costs of molt-breeding overlap were well studied for adults through experimental approaches (e.g., enlarged brood size or simulated molt) to increase the cost of reproduction, specifically in pied flycatchers (Ficedula hypoleuca) and blue tits (Cyanistes caeruleus). Parents with enlarged broods tended to delay molting compared to those with control broods, where the degree of molt was lighter and briefer in males and young birds than in females or old birds during the nestling period (Siikamäki et al., 1994; Svensson & Nilsson, 1997). Also, molt-breeding overlap simulated by removing wing feathers is known to lower parents’ body condition, recruitment rate, and return rate to next year, resulting in current and/or future fitness costs (Hemborg & Lundberg, 1998).

Compared to costs of molt-breeding overlap found in adults, juveniles might need another resource allocation strategy under a trade-off between growth and molt at the early developmental stage after hatching. The amount of post-juvenile molt can be either partial or complete (Kiat & Izhaki, 2016). Juveniles should favor the former strategy under high costs of survival, while others should allocate limited energetic resources into not only development, but also molt simultaneously. Furthermore, juvenile feathers can also exhibit relatively lower quality in feather length, mass, and/or melanization measurements compared to feathers of adults (Kiat & Sapir, 2018). The amount of nest-grown feathers is likely to be replaced depending on the timing of developmental completion. Therefore, patterns of age-related wing molt found in this study might be partial under two different trade-offs (i.e., growth-molt-breeding) in the puberty-maturity spectrum (i.e., juvenile to adult stages) of oriental storks.

Prior to 1950s, the distribution and life histories of oriental storks might be continuous from Russia to China, Korean Peninsula, and Japan, meaning that their life forms were likely to range from long-distance migrants towards the most northern breeding sites to short-distance migrants or residents towards the most southern breeding areas (e.g., Republic of Korea and Japan) along the latitudinal gradient. Currently, the northern population lives as long-distance migrants, while the recently establishing populations from reintroduced storks in Korea and Japan seem to live as residents because they lack the regular migratory behavior (J. Yoon, personal observation). It is still unknown whether migratory and released storks exhibit distinct life history traits shaped by various trade-offs among growth, molt, and reproduction at the most northern and southern parts of their distribution (Kiat et al., 2019a; Wingfield, 2008). Although age-related wing feather loss is found at low-intermediate rates, future study should investigate the degree of molt-breeding overlap and its fitness-related costs of northern and newly establishing southern populations of oriental storks.

Supplemental Materials