I aim to investigate these question with a combination of observational and experimental studies. Specifically, I make use of a frequently used model organism, the Eurasian blue tit (Cyanistes caeruleus). The blue tit is a small cavity nesting, passerine bird occurring in the Western Palearctic (Föger and Pegoraro 2004). The IUCN redlist lists the blue tit as a least concern conservation status with large, even increasing populations sizes (BirdLife International 2016). It is a 12g passerine, blue and yellow in colour, with small sexual dimorphism (Föger and Pegoraro 2004). Preferred habitats for breeding are mixed deciduous forests as opposed to coniferous stands. Generally this species occurs in lowlands, but the record for breeding has been set at 3500 meters above sea level in the Caucasus mountain range (Cramp and Perrins 1993; Föger and Pegoraro 2004). This socially monogamous species has been thoroughly studied since the 1850s due to its wide distribution and accessible breeding in artificial nest boxes and large, manipulable clutch sizes (5-15 eggs; Krüper, 1853; Nur, 1986). In this species only the female builds nests and incubates the eggs, though both parents provision the chicks (Cramp and Perrins 1993). Thus, conveniently I can manipulate maternal investment strategies at the early stages in this model organism and also investigate parental costs for both parents at the rearing stage, plus test for partner responses.
Within my fieldwork, I utilise a novel 1000m altitudinal study system, located in the French Pyrenees mountain range. The French Pyrenees are characterised by relatively short, though steep valleys with mixed forests gradually turning into beech and fir stands above 900 m (Ninot et al. 2017). The tree line and the transition to mountain pastures is situated at ca. 1500 m, depending on the geological profile (Prodon et al. 2002). Additionally, since the second half of the 20th century, abandonment of land and farming practice is leading to increases in forested areas (Gibon and Balent 2005; Mottet et al. 2006). The focus population breeds in an established nest box population (N = ca. 640) across a 450-1500 m altitudinal gradient. I have aimed to distribute the nest boxes evenly across the altitudinal gradient, however due to characteristics of the terrain (e.g. steep slopes) some irregularities and minor gaps exist (Fig. 1.1). Woodcrete nest boxes were installed before the first breeding season in 2012 with a distance of more than 50 m between neighbouring boxes. In addition, handcrafted bamboo poles are used to lift down nest boxes. Nest boxes are shared with other passerine species; mainly great tits (Parus major), coal tits (Periparus ater), marsh tits (Poecile palustris), and occasionally nuthatches (Sitta europaea). A full characterisation of the study system can be found in Chapter Two.
The first data chapter (Chapter Two) will investigate general breeding parameters of our blue tit system. Specifically, I will investigate the associations among altitude, phenology, fecundity, productivity and nestling mass, from egg laying until fledging. Purely observational, climatic and reproductive data will be collated across six breeding seasons; including average daily temperature, clutch size, hatching and fledging numbers, fledging mass, and reproductive timing; i.e. first egg lay date. As part of characterisation of the altitudinal gradient, I will first investigate if average daily temperature shifts with elevation using temperature logger data. A gradual altitudinal decline in temperature is predicted (Körner 2007). Furthermore, lay date has been found to be closely linked to temperature, and further to phenology, i.e. tree and caterpillar development, which has consequences on later chick survival (McCleery and Perrins 1998; Sanz 2002). Thus, I predict that a delayed start of reproduction will be observed with increasing altitude, which should consequently affect later breeding parameters. Further, as the productive period is shorter at high altitude (Rolland 2003; Körner 2007), I predict that the reproductive output such as fledging numbers should be negatively affected. Specifically, I investigate phenological plasticity in response to altitude and year in this population. Second, I then investigate the effects of lay date and altitude on clutch size and hatching success, as a means of quantifying the phenotypic correlation between lay date and clutch size across the altitudinal gradient, and its effects on hatchability. Finally, I test the effects of lay date on fledging success and nestling mass to provide insights into phenological mismatch in this population, and whether such metrics of success are modified by phenology-fecundity associations.
The third chapter will look at if environmental cues such as budburst are used to differing degrees along the altitudinal gradient. As aforementioned, lay date has been found to be linked to phenology, i.e. tree development, which has consequences on later chick survival (McCleery and Perrins 1998; Sanz 2002). To investigate if females can predict optimal prey availability and thus if hatch date is correlated with this, budburst will be used as a proxy. As temperature is lower at higher altitudes (Körner 2007; Chapter Two), I predict budburst to be delayed compared to lowlands. This budburst should be tracked more closely by higher elevation birds as the season for reproducing in shortened, resulting in fewer reproductive opportunities. Thus, correct reproductive timing to exploit maximum environmental resources is crucial for reproductive success at high altitudes. To decipher this relationship, I will look at observational phenological and reproductive data, specifically at how well budburst and lay date is matched with altitude and whether this temporal relationship affects reproductive output such as fledgling numbers and mass. Additionally, I will investigate if strategies are used to improve the association between budburst and hatching after laying.
The fourth chapter will focus in on how parental investment choices are linked across different phases of a single breeding attempt. I will investigate how experimentally manipulated investment choices in early breeding phases (the number of eggs laid) will affect later investment levels at the rearing stage. The rationale behind this experiment is that most studies have ignored the costs of egg laying and incubation to females (Monaghan and Nager 1997). Both can contain a cost in various bird species, in particular for future fitness of the female (Reid et al. 2000; Visser and Lessells 2001; Nager et al. 2001). These early costs should impact later investment in offspring (Savage et al. 2013), which should shift the care contributions of the parent directly affected (females), and their partner (males). However, these costs may also affect future reproductive abilities and especially in short-live species such as blue tits. Hypothetically, high early investment may lead to high investment at the rearing stage, as residual reproductive value is reduced if key resources are depleted faster and future survival is reduced (Stearns 1992). Thus, I predict that heightened, early invest by females will affect later investment choices in the rearing phase, however decreased and increased investment are possible. To test these two predictions, females are made to lay additional eggs, though incubation and rearing costs are kept constant, as in control groups, which were not made to lay additional eggs. This is achieved by a cross-fostering approach. Later investment in the rearing stage is investigated by observational data on provisioning of both parents.
The fifth chapter will concentrate on the rearing stage and highlight how different environmental drivers influencing parental care. In particular, I will look at if contributions of females and males change depending on altitude, year, caterpillar availability and intrinsic nest characteristics such as brood age and size. I predict that if environmental harshness (altitude) increases it will be harder for parents to provision at equivalent levels to their lowland counterparts. On the other hand, I predict that parents may respond in line with the pace of life framework, with high altitude individuals shifting to a slower pace resulting in higher parental care in fewer offspring (Hille and Cooper 2015; Boyle et al. 2016). Additionally, traditional theory does not consider differential task division between the sexes, such as nest sanitisation by the female and predator defence task by the male (Maynard Smith 1977; Klug et al. 2013). This should result in differential investment strategies. As part of this, I will first explore natural nestling provisioning. As brood size is a key fitness trait, it must be a crucial factor for investment choices. However, the sexes may have different optimal brood sizes, depending on previous and future investment choices. Thus secondarily, a temporary brood manipulation is performed to reveal possible underlying differences in provisioning strategies between the sexes, in response to artificially increased or decreased brood sizes, compared to controls, and if responses change with altitude. I will test classic models of bi-parental care predicting: (a) comparable provisioning contributions of males and females independently of ecology; (b) partial compensation response rules by both sexes; and (c) these partial response rules to be manifest as overall increases in nestling mass.
Globally these observational and experimental data chapters aim to investigate underlying drivers and mechanisms of bi-parental care in birds. Reproductive costs of both parents during the pre- and post-hatching stages will be manipulated naturally with use of the altitudinal gradient and through directed experiments to investigate underlying reproductive strategies. To conclude, the sixth chapter will constitute an overall discussion aiming to tie all the results together found during this PhD. I aim to highlight the novelty these results for the field of parental care. I will be indicating overarching parental investment (care) strategies for this particular system. Impacts on potential species’ evolutionary processes and endurance under climate change prognoses will be discussed.
There is a pressing need to understand whether and how populations respond to changing climates. To date, much of our understanding stems from longitudinal studies of sufficient duration to encapsulate climate shifts. While such studies provide essential insights, they obviously require significant time, and the magnitude of any effect measured is contingent upon the magnitude of inter-annual variation in climate; which is often modest. Here I use a 1000 m altitudinal gradient in the French Pyrenees to generate representative 2-3 °C differences in temperature faced by breeders in a population of blue tits (Cyanistes caeruleus). During the six years of study, I found that breeding phenology typically varied by ca. nine days within altitudinal zones, but was on average 11 days earlier at low versus high altitudes. Early breeding was generally associated with larger clutch sizes, which in conjunction with reduced nestling mortality, led to more young being fledged. However, compared with birds breeding at low elevation, those breeding at high elevations also laid larger clutches than expected for their lay dates. As a consequence, despite low elevation birds showing reduced probability of hatching failure compared with high elevation birds, they did not enjoy greater breeding success; and high elevation broods were also in better condition. Our results suggest that constraints on plasticity are unlikely to explain phenological mismatches; and lead us to hypothesise that the answer lies with the relative quickening of development of ectothermic prey with warming springs, compounded by current selection on negative phenology-fecundity associations of endothermic predators.
Recent meta-analyses indicate that organisms of diverse taxonomy are responding to climate change by advancing the timing of key life events, particularly reproduction (Thackeray et al. 2010, 2016). Phenological responses within populations appear to be largely plastic (Phillimore et al. 2010), and such plasticity is suggested to play a significant role in allowing populations to adapt in real time to changing climate (Parmesan 2006; Both et al. 2006; Visser 2008; Both et al. 2009a; Visser et al. 2012; Gienapp et al. 2013). Nevertheless, whether plastic advances in breeding phenology are sufficient or adaptive will depend additionally on associated changes to reproductive investment, including fecundity and any subsequent levels of care. Despite this, less is known about potential constraints to plasticity or climatic impacts on adaptive associations among phenology, key life history traits and metrics of success (Visser et al. 2015; Visser 2016). In order to address these shortcomings, the obvious general association between the location of a population and its climate will often need to be de-coupled. There are two potential ways of achieving such decoupling in natural systems: intensive longitudinal study encapsulating sufficient climatic variation; and the use of altitudinal gradients to generate representative levels of climatic variation in the short term and to test responses by individuals from the same population in conjunction with their downstream consequences for investment and success.
Testing adaptive responses to climatic variation for fecundity and subsequent levels of care is more challenging than testing impacts on breeding phenology because fewer taxon are amenable to quantitative assessment of such measures. Birds offer an important model in this regard because fecundity and subsequent care is variable and easily measured. Current evidence from longitudinal studies in such taxon, often spanning several decades, suggests that advancing lay date is generally often associated with increased clutch size (Potti 2009; Dunn & Møller 2014). This might be construed adaptive because the ability to advance breeding more in response to warming springs is likely to generate improved match with peaks in prey availability (Visser et al. 2006; Charmantier et al. 2008). On the other hand, higher fecundity generally leads to reductions in per capita prey acquisition rates, potentially compounding any effects of mismatches between phenology and prey availability. Interestingly, quantitative genetic approaches suggest a negative genetic correlation between phenology and fecundity (Sheldon et al. 2003), suggesting that an advance in lay date might often be associated with an incidental increase in clutch size. Compensating for increased clutch size as a consequence of advanced phenology would require increased parental effort, but whether or not this is the case is not well known (Dunn and Winkler 2010). Thus, it is currently unclear whether or not commonly reported negative associations between phenology and fecundity are adaptive, or contribute to documented detrimental effects of climate change (Dunn and Møller 2014).
While longitudinal studies are unquestionably invaluable, opportunities to establish such studies are now more limited and the time taken to do so is prohibitive with respect to the need for answers. A potentially viable alternative approach is to use altitudinal gradients to generate representative variation in climate among individuals within a single population. Altitudinal gradients have been commonly used to test for ecological impacts on key fitness-related traits. For example, a recent meta-analysis of bird species breeding across altitudinal gradients showed that breeding phenology was considerably later at higher elevations, and that clutch sizes tended to be smaller (57 % of 98 species); with average reductions of ca. 6 % (Boyle et al. 2016). These finding mirror the results of longitudinal studies: that warmer weather leads to both advanced phenology and fecundity. However, almost all previous altitudinal studies have conducted comparisons of the same species across different populations, meaning that varying degrees of local adaptation could cloud assessment of plastic responses to climatic variation. In order to provide a more realistic analogy of climate change impacts, associations between phenology, fecundity, levels of care and productivity need to be investigated across altitudinal gradients within the same population.
Table of contents :
Chapter 1: General introduction
1.1 General framework
1.1.1 Reproductive investment
1.1.2 Parental care
1.1.3 Systems of parental care
1.1.4 Environmental variation
1.2 PhD aims
1.3 Study species
1.4 Study system
1.5 Thesis outline
Chapter Two: Extreme plasticity in breeding phenology across an altitudinal gradient: implications for understanding phenological mismatch
2.3.1 Study population and habitat
2.3.2 Phenology, investment and success
2.3.3 Statistical analysis
2.4.1 Breeding phenology
2.4.2 Contributors to reproductive success: clutch size and hatching success
2.4.3 Reproductive Output
2.4.4 Nestling mass
Chapter Three: Testing the use of budburst as a reliable cue to breeding phenology in a population of blue tits breeding along an altitudinal gradient
3.3.1 Statistical analysis
3.4.1 Phenology of budburst and lay date
3.4.2 Timing parameters post-laying
3.4.3 Reproductive success
Chapter Four: Inducing females to lay more eggs leads to increased per capita provisioning rates of nestlings in blue tits
4.3.1. Experimental design
4.3.2 Treatment effects pre-provisioning
4.3.3 Provisioning behaviour
4.3.4 Statistical analysis
4.4.1 Treatment effects pre-provisioning
4.4.2 Provisioning behaviour
4.4.3 Nestling mass
Chapter Five: Brood size manipulations across an altitudinal gradient shed new light on investment strategies in a bi-parental care system
5.3.1 Sex differences in provisioning natural brood sizes
5.3.2 Responses to brood size manipulations
5.3.3 Treatment effects on brood mass
5.4.1 Sex differences in provisioning natural brood sizes
5.4.2 Responses to brood size manipulations
5.4.3 Treatment effects on brood mass
Chapter Six: General discussion
6.1 PhD findings
6.2 General conclusions
6.2.1 Altitudinal effects on parental investment
6.2.2 Sexual conflict
6.3 Global vision