text in 
English (pdf)
Article in xml format
Article references
Automatic translation
Send this article by e-mail
Cited by SciELO 
Cited by Google 
Similars in
SciELO 
Similars in Google 
Permalink
INTRODUCTION
Continuous seabird monitoring efforts allow, among other things, to relate population trends to variation in oceanographic conditions (Gaston et al., 2009; Grémillet and Boulinier, 2009; Humphries, 2015). For example, it has been established that changes in the conditions of sea surface temperature (SST) can affect the reproduction and survival of seabirds, by changing the trophic structure of the pelagic environment, due to the alteration in food availability and quality of their prey (Schreiber, 2002; Frederiksen et al., 2006; Tompkins et al., 2017; Champagnon et al., 2018). However, the response to climate change appears to be related to locality, species, sex, and developmental stage (Anderson, 1989; Maness and Anderson, 2013; Champagnon et al., 2018). For example, the survival of juveniles of the Blue-footed Booby (Sula nebouxii) is reduced during the warm phase of the ENSO (El Niño-Southern Oscillation) (Oro et al., 2010), while the number of immature individuals of the Black-browed Albatross (Thalassarche melanophris) increases with increasing SST (Pardo et al., 2017).
The Gorgona National Natural Park (GNNP) has been developing the longest-running seabird monitoring program in Colombia, started in 2002 (Estela et al., 2010). During this monitoring, information on the presence, abundance, and reproductive status of four species of seabirds recurrently sighted in the GNNP (Suliformes and Pelecaniformes) has been continuously recorded (Cadena-López and Naranjo, 2010; Perlaza-Gamboa et al., 2020). Among them is the Brown Booby (Sula leucogaster etesiaca), an endemic subspecies of the American humid tropical Pacific (Ospina-Álvarez, 2004), resident on Gorgona Island where the largest population is located with a colony of approximately 300 individuals and 100 reproductive pairs (Cadena-López and Naranjo, 2010; Estela et al., 2016). This subspecies is considered as threatened in the country (Renjifo et al., 2016).
Considering that the variability of oceanographic conditions associated with GNNP are influenced by ENSO (Blanco, 2009) and this has been identified as a relevant factor for the survival and reproductive aspects of other booby species in the Eastern Tropical Pacific (Anderson, 1989; Ribic et al., 1992; Mauck and Grubb, 1995; Ancona et al., 2012) and on Gorgona Island (Perlaza-Gamboa et al., 2020), we evaluate the temporal dynamics of the Brown Booby colony in this locality, in order to answer the following research question “Does the sea surface temperature affect the temporal variation in the abundance of the different development stages of the Brown Booby (Sula leucogaster etesiaca) on the Gorgona Island?”. To do this, we start with the analysis of the records of the abundance of adults, juveniles and chicks of this species carried out during the monitoring of seabirds in the GNNP from 2002 to 2018, as well as the sea surface temperature (SST) and thermal anomaly (ANOM) in this locality.
STUDY AREA
Within the delimited conservation area for the Gorgona National Natural Park (GNNP), are the Gorgona and Gorgonilla islands (2˚ 55′ 45″-3˚ 00′ 55″ N and 78˚ 09′ 00″-78˚ 14′ 30″ W), located approximately 30 km from the continent (Diaz et al., 2001; Giraldo et al., 2014a). In the extreme north of Gorgona and around Gorgonilla there are a series of emerged rocky islets and promontories, in which the Brown Booby builds its nests (Ospina-Álvarez, 2004; Cadena-López and Naranjo, 2010). These islets have variable shapes and inclinations. The boobies are capable of nesting on land with slopes of less than 10° and up to 90° (Ospina-Álvarez, 2004). The islets may present some vegetation cover, mainly herbaceous, although the nesting sites can be established in parts of the islet devoid of vegetation at an average height of 8.9 m.a.s.l. (Ospina-Álvarez, 2004).
The tide is semi-diurnal, with a height range between 4 and 5 m (Diaz et al., 2001; IDEAM, 2019). Precipitation levels greater than 6000 mm/year are recorded, with higher records between May and November and lower from December to April (Blanco, 2009). Regarding the characteristics of the water column of the pelagic environment, a cold period of high salinity has been reported between January and April, during which the thermocline is located at 7 m depth, and a warm period of low salinity between May and December, during which the thermocline is more than 40 m deep (Giraldo et al., 2008).
MATERIALS AND METHODS
The information used for this research corresponded to the monthly abundance records of Brown Booby carried out by trained officials of the GNNP, between 2002 and 2018, through visual censuses along a predetermined route within the framework of the seabird monitoring program. of this protected area (Figure 1). These records are made on two consecutive days of the first week of each month. Detailed information on the methodology associated with this monitoring program can be obtained in Payán (2016) and Perlaza-Gamboa et al. (2020).
Figure 1 Generalities of the monitoring of seabirds of the Gorgona NPN, Colombian Pacific. A) Route taken during the monitoring. B) Example of an islet in the northern sector of the Island. C) Female and chick of the Brown Booby (Sula leucogaster etesiaca). Photographs: Felipe A. Estela.
Each individual observed was assigned to one of three stages of development defined according to the characteristics of their plumage: Chicks were identified by their characteristic white down, juveniles by presenting a general grayish-brown coloration with dark brown spots, and adults by the contrast between the dark brown color of the head, neck, and back, with the white of the lower chest and belly (Hilty and Brown, 2001; Ospina-Álvarez, 2004). The abundances were standardized by dividing the number of individuals observed by the total distance traveled in each of the samplings. To give uniformity to the abundance data, only the sampling events in which the two observation days were carried out were considered for the numerical analysis. The chicks abundances recorded since 2002 were used since their identification was always possible. However, during the first years of sampling, it was not discerned between adult and juvenile individuals, so to analyze these two stages of development, the data obtained from 2011 were used, the year from which it began to differentiate between these two categories. Data from individuals with unidentified developmental status were not used. For the descriptive analysis of these data, the median was used as the central tendency parameter and the maximum and minimum abundance as dispersion parameters, to reduce the influence of atypical data and because they were discrete data.
The Gorgona Island SST data were obtained from the monthly satellite images recorded by the MODIS-Aqua sensor of the Ocean Biology Processing Group (OBPG), at the CPC16 station of the ERFEN sampling mesh (3° 0′ 0″ N-78° 0′ 0″ W) (OBPG, 2015). The thermal anomaly (ANOM) was calculated by subtracting the monthly historical average for the period 2002-2018 from the monthly time series of the SST.
To establish the time lag of the possible association between temperature and the abundance of chicks, adults, and juveniles, a cross-correlation was implemented between the monthly abundance records with the SST and the ANOM of Gorgona Island. The greatest lag (Lag) allowed for chicks was four months since it is the period that includes the incubation period plus the time it takes for the chicks to grow until they are visible in the nest (Nelson, 1978; Ospina-Álvarez, 2004). For the juveniles, the greatest lag allowed was seven months, which corresponds, on average for the species, to the time from incubation to the feeding of fledglings by the parents (Nelson, 1978; Ospina-Álvarez, 2004). Finally, for adults, the maximum lag allowed was 12 months, since this period completely covers the variation in their abundance. The strength of the correlation was measured with the autocorrelation function (ACF), which for this study acquires a value of -1 for the highest negative correlation and 1 for the highest positive correlation.
A generalized linear model (GLM) was developed for each stage of development independently, to explain the possible relationship between abundance and SST and ANOM, that presented significant correlations, taking into account the time lag found in the cross-correlation. The quasipoisson distribution was used because count data were evaluated and there was no overdispersion. The analyzes and Figures were made in the R program (R Core Team, 2018). The ggplot2 package was used for the Figures (Wickham, 2016), astsa for the cross-correlation (Stoffer, 2017) and MASS for the models (Venables and Ripley, 2002), in addition to other functions to organize the data (Wickham, 2007; Fox and Weisberg, 2011; Trapletti and Hornik, 2018).
RESULTS
During the study period, 184 monthly records of Brown Booby chicks and 92 monthly records of adults and juveniles were made. The abundance of adults presented the greatest variation, with a monthly median of 5.70 ind/km and a maximum and minimum of 11.02 and 2.00 ind/km respectively. Chicks presented a monthly median of 0.22 ind/km (minimum = 0 ind/km and maximum = 2.62 ind/km) and the juveniles 0.42 ind/km (minimum = 0 ind/km and maximum = 1.54 ind/km), with higher records, in both cases, during the last semester (Figure 2).
Figure 2 Monthly variation of the abundance of A) adults, B) chicks and C) juveniles of the Brown Booby (Sula leucogaster etesiaca) on Gorgona Island, Colombian Pacific. The box represents the median and 25% and 75% percentiles. The bars represent the non-outlier maximum and minimum value and the points correspond to outliers.
The sea surface temperature of Gorgona Island presented a median equal to 27.26 °C, with records between 25.47 °C and 28.05 °C. The adults did not present any significant correlation with SST or ANOM. In contrast, the abundance of juveniles showed a significant correlation with ANOM and SST, with a time lag of two and one months, respectively. This suggests that the number of juveniles is associated with the ANOM registered two months before and with the SST of the previous month. Finally, the abundance of chicks was correlated with the ANOM without any time lag, and with the SST reported four months earlier (Table 1).
Table 1 Cross-correlation between the monthly abundance of chicks, juveniles, and adults of the Brown Booby (Sula leucogaster etesiaca) with the sea surface temperature (SST) of Gorgona Island (Gor) and its thermal anomaly (ANOM). ACF = strength of correlation. ANOM = thermal anomaly. Lag = time lag of correlation measured in months. NS = nonsignificant correlation.
According to the generalized linear model (GLM) developed between the abundance of juveniles with the ANOM (coefficient = 0.33; P < 0.001) and the SST (coefficient = 0.52; P < 0.001), considering the respective lag (Lag), the number of juveniles tends to decrease with the increase of both thermal variables (Figure 3A and 3B). On the other hand, the abundance of chicks tends to decrease with the increase in ANOM (coefficient = 0.58; P = 0.017) (Figure 3C), but increases with increasing SST (coefficient = 2.08; P < 0.001) (Figure 3D).
DISCUSSION
The abundance of chicks and juveniles Brown Boobies associated with Gorgona Island was higher in the last months of the year, probably as a consequence of the concentration of reproductive effort between July and September, a condition that was reported by Cadena-López and Naranjo (2010). However, this species on Gorgona Island exhibits reproductive activity throughout the year, with a low intra-annual variation in population size (Ospina-Álvarez, 2004; Cadena-López and Naranjo, 2010; Perlaza-Gamboa et al., 2020). In this reproductive regime, egg laying mainly occurs between June and August, with an incubation period between 40 and 44 days, followed by a period between one and three months in which the chicks grow until they are visible and acquire the characteristic white down, that at the end of this period they begin to molt to a brown plumage (Nelson, 1978). Once these individuals make the first flight, at approximately 96 days of age (Nelson, 1978; Ospina-Álvarez, 2004), they are fed by the parents for the next three to eight weeks, although this period can be extended depending on prey availability (Schreiber and Norton, 2002). This system encourages individuals to acquire independence from their parents during the first months of the year, during which time greater productivity has been reported in the pelagic environment near Gorgona Island (Giraldo et al., 2014b).
This reproductive strategy has been reported for other species of seabirds, which begin to reproduce when there is less productivity in the ecosystem and their prey is less available, but it allows the juveniles to achieve independence from their parents when there is the greatest availability of prey. (Passuni et al., 2016, 2018). In general, this strategy is carried out, mainly, by birds whose food is widely distributed, such as seabirds (Nelson, 1978; Furness and Monaghan, 1987; Vilchis et al., 2006). In addition, it is worth mentioning that the four-month lag in the positive correlation between the abundance of chicks and the SST suggests that the greater abundance observed in October and November is influenced by the temperatures registered in the middle of the year, at which time the transition between the period of highest to lowest productivity occurs in the Eastern Tropical Pacific (Pennington et al., 2006) and probably in the surroundings of Gorgona Island (Giraldo et al., 2014b). This pattern is similar to that observed in the Mexican tropical Pacific, where the reproductive period was mainly associated with the warmest and least productive time of the year, but coincides with the increase in the availability of migratory warm-water prey (Hernández-Vázquez et al., 2017). For these reasons, to understand the reproductive regime of the Brown Booby on Gorgona Island, it is necessary to deepen the knowledge about the intra-annual population variation of its prey resources, as well as to determine the period in which most of the offspring acquire independence from parents.
It is likely that the lower productivity associated with warmer waters, and greater precipitation between May and December (Diaz et al., 2001; Giraldo et al., 2014b), causes greater loss of chicks, as has been reported for other seabirds in the Eastern Pacific (Jaksic and Fariña, 2010). This could trigger more frequent second clutches by reproductive pairs (Nelson, 1978), resulting in a greater number of chicks observed in recent months. On the other hand, the negative correlation found between the abundance of chicks and the thermal anomaly suggests that these individuals are affected by changes in sea temperature, possibly through the decrease in the food provided by their parents. For example, for some seabirds, increases in sea temperature and decreases in reproductive success have been associated with altered sea productivity and prey availability (Ancona et al., 2011; Furness, 2016). The lower availability of prey resources related to warmer waters may explain the decrease in the abundance of juvenile Brown Booby when SST and ANOM increase in the pelagic environment of Gorgona Island.
During adverse conditions, such as low prey availability, some long-lived species, such as the Brown Booby (Hennicke et al., 2012) generally tend to minimize individual costs of reproduction (Mauck and Grubb, 1995). For example, adults reduce the effort invested in parental care and the supply of food to dependent juveniles or chicks, to increase their survival, but decrease reproductive success (Maness and Anderson, 2013). This strategy may explain why no significant correlation was observed between adults and SST or ANOM, but it was observed in chicks and juveniles. Furthermore, the abundance of this booby described by Perlaza-Gamboa et al. (2020) could be mainly due to the decrease in the number of juveniles, possibly because the food supply was lower when the SST increased. However, this effect is likely less in species that have parental care before and after the chicks become fledglings (Stienen and Brenninkmeijer, 2002; Maness and Anderson, 2013), as is the case of the Brown Booby (Ospina-Álvarez, 2004). Nevertheless, these hypotheses must be confirmed by a study that evaluates the strategies carried out by parents during stressful conditions, and their consequences on the development and survival of their offspring before and after having achieved independence.
CONCLUSIONS
The SST and ANOM of Gorgona Island are associated with the trends observed in the population variation of chicks and juveniles of the Brown Booby, possibly through an effect on the supply of pelagic fish provided by the parents, since it is probable that these Individuals cannot meet the needs of the young, but they do manage to stay alive during times of food shortage. This is evidenced in a lower number of juveniles and chicks during the months with the greatest anomalies, which suggests that the variation in SST and ANOM may affect the reproduction of this population. Also, possibly the reproductive regime ensures that the young achieve independence from the parents when there is greater availability of prey resources.
