Can 34 years of data estimate 120 years of pattern?
Howard Snell and Solanda Rea
Introduction
The 1997-98 El Niño event missed being the wettest two-year period in the Galápagos Archipelago by less than one millimeter. Between 1 January 1997 and 31 December 1998, 3407.6 mm of rainfall was recorded at the Charles Darwin Research Station on Santa Cruz Island. That is 19% of all the rainfall since 1965, a total surpassed only by the El Niño event of 1982-83, when 3408.2 mm of rain fell (also 19%, Table A.1). Across most of the eastern Tropical Pacific Basin, the 1997-98 El Niño appears to have been the strongest in recorded history (Kerr 1999, McPhaden 1999). While the difference in total rainfall on Santa Cruz between the two El Niños is less than two-thousandths of one percent, most residents and visitors of the Galápagos who experienced both events recall that of 1982-83 as being much more severe. Does memory fail us or are patterns within the Galápagos distinct?
On the large scale, El Niño events are associated with abnormally high and consistent sea-surface temperatures across the tropical Pacific. The lack of the normal west-to-east thermal gradient across the surface of the Pacific is coupled with the weakening and occasional reversal of the easterly tradewinds. A combination including high sea-surface temperatures and weak winds promotes tremendous evaporation of water into the atmosphere, leading to abnormally strong convective storms and great rainfall (McPhaden et al. 1998). Generally El Niño events last between 12 and 18 months (McPhaden et al. 1998). For the last 5000 years these conditions have occurred approximately every 2 to 8.5 years, roughly double the rate during the 10,000 prior years (Robell 1999). The intensity of the rainfall and the thermal anomalies occurring during El Niño events appear to be increasing within the last twenty years (Wolter and Timlin 1998). While global warming may influence that increase, it appears that several recent El Niños, including those of 1982-83 and 1997-98, coincided with other cyclical climatic events of greater temporal amplitude that may have magnified their intensities (Kerr 1999). El Niño events are apparently part of a cycle that often includes a subsequent dry period that can develop into severe droughts. Recently these dry periods have been called La Niña events (McPhaden et al. 1998).
These climatic fluctuations are powerful factors in the ecological interactions of many species within the Galápagos and elsewhere (Robinson and del Pino 1985, Grant 1985, other articles in this volume). The general impression is that terrestrial organisms flourish during the wet Niño conditions and suffer the droughts of La Niñas, while marine species suffer during the Niño periods of high sea-surface temperatures and decreased upwelling of deep, nutrient-rich ocean currents and prosper during "normal" and La Niña conditions. However, several species of terrestrial organisms appeared to suffer higher than usual mortality and low reproductive success during the 1997-98 El Niño event. Additionally, several alien aggressive terrestrial species increased their ranges during the same time. It is possible that the combinations of reduced reproductive success, increased mortality, and spread of alien species could make modern El Niño events damaging for terrestrial organisms. Within this paper we examine patterns of climatic variability within the Galápagos and compare them with patterns of the eastern tropical Pacific and continental Ecuador as a framework for evaluating potential biological consequences.
Methods
Data for this study came from direct measurements, the published literature, and the Internet. Basic meteorological variables have been measured by the Charles Darwin Research Station (CDRS) since 1964. We use mid-day air (Ta) and sea (Ts) temperatures and total daily rainfall (1800-1800, all times are UTC — 6 hr) from 1 January 1965 through 31 December 1998. Air temperature is recorded in shade 2m above the surface, Ts is recorded at the coast in a bucket of water pulled from the sea, and total daily-rainfall is the sum of measurements taken at 0600, 1200, and 1800 from a rain gauge 1.5 m above the surface. The coastal station is on the grounds of the CDRS at 0° 44' 20" S latitude, 90° 18' 25" W longitude. Since 1987, measurements have been recorded from a station in the upper Transition Zone (Bellavista, 194 m altitude) at 0° 42' S latitude, 90° 22' W longitude. We produced monthly averages or totals of these data with pivot tables in Microsoft Excel 97. Accumulated rainfall is the inclusive sum from January to the month represented. Because El Niño events often span two calendar years, we needed an additional set of two-year periods for comparisons. We produced a set of 33 periods using all pairs of successive years. This requires that each year be represented twice, once as the first year of a period and once as the last year of a different period. That set includes all potential El Niño and La Niña events.
Some published data for the tropical eastern Pacific were extracted from figures downloaded from the Internet (McPhaden 1999, Wolter and Timlin 1998, Robell et al. 1999). The figures were imported into ArcView 3.1 (Anonymous 1996), referenced to a coordinate system based on their axis, and the data exported to an Excel 97 spreadsheet. Additional data were downloaded directly from web sites of the National Oceanic and Atmospheric Administration's (NOAA) web pages and the Comprehensive Ocean-Atmosphere Data Set (COADS, ) within the Internet Data Library of the International Research Institute for Climatic Prediction (IRI, www.iri.ideol.columbia.edu).
Results and Discussion
Climatic patterns
Annual rainfall is extremely variable in the arid coastal zone of Galápagos (Figure 1). The coefficient of variation among years is 107.6, and the distribution is highly skewed towards a few years with extremely great rainfall (g = 2.412, Table A1). The wettest year recorded since 1964 is 1983 and the driest 1985 (2769 and 64 mm, respectively, Table A1). Although we have not encountered data demonstrating years with greater rainfall, drier periods have been recorded on San Cristóbal and Daphne (Grant 1985, rainfall measured January through May). However, apparent differences in values from data sets covering the same site and years have led us to revise previously distributed data from the CDRS, and it is possible that there were similar problems in previous data from San Cristóbal. Regardless, a range of total rainfall among years of 2705 mm is great. Unfortunately, rainfall was not recorded at Bellavista during 1983, so we do not know how much rain fell there. Generally, Bellavista receives approximately 500 mm more rain than the CDRS, and that difference is relatively independent of the total amount of rain (Figure 2). Thus, the coastal region does not receive a constant percentage of the rainfall in the upper transition zone, rather it gets, on average, about 13.5% in dry years and 71% in the wettest years. Not only is rainfall greater in the upper transition zone, it is less variable (coefficient of variation = 64.03) and less skewed towards years with great rainfall (g = 1.128, Table A2).
Figure 1. Rainfall and temperature patterns on Santa Cruz Island.
Since 1965, there have been five strong El Niño events in Galápagos: 1975-76, 1982-83, 1986-87, 1993-94, and 1997-98 (Figure 1). Although recorded as a strong El Niño event elsewhere, 1972 and 1973 showed barely abnormal rainfall in Galápagos (Figure 4). The year 1965 was also recognized as an El Niño event (Grant 1985), although, in view of the amounts of rainfall recorded in El Niño events since 1975, the 600 mm recorded then seems slight. The only El Niño recorded in Galápagos that did not result in two years of above-median rainfall was that of 1986-87 (Figure 1). Within Galápagos, it could be appropriate to view that as an event of a single year (1987), but its signal in other parts of the Pacific was that of a typical El Niño event extending into two calendar years (McPhaden 1999). The recorded El Niños of Galápagos are always followed by droughts. If we define a drought as the span of contiguous years with rainfall below the median, the average duration of post-Niño droughts is 2.3 years and the mode is 3 years. The 1980s were a decade of extremes for the islands. The two most severe droughts and the first- and third-ranked El Niños occurred then.
Figure 2. Relationship of coastal and upper transition zone rainfall on Santa Cruz Island.
While the total amounts of rainfall in the 1982-83 and 1997-98 El Niños are effectively identical (Table A1), they differed greatly in medium-term intensity. Although the rainiest day ever recorded at the CDRS was 3 June 1997 (194.6 mm), the monthly totals in 1983 are incredible. During the three months of April, May, and June 1730 mm of rain fell, surpassing the totals of all years except 1998 (Table A1). Apparently a "typical" El Niño does not exist for the Galápagos. Even though strong El Niños in the Galápagos basically accumulate either 3.4 or approximately 1.5 m of rain, the patterns of rainfall-accumulation and the sea-surface temperature anomalies producing them are variable. Some events accumulate rain during the typical rainy seasons, January through April (Table A1) of two years. That pattern was shown in 1975-76 and 1992-93. Other events accumulate rain during the normal rainy season and receive extraordinary rainfall in typically drier months (1982-83, 1986-87, and 1997-98). In general, El Niño events with great rainfall have larger Ts anomalies for longer periods than those with less rainfall (Figure 3). It is difficult to establish the "normal" pattern for climatic variability in the Galápagos Islands when nearly one-third of the years with recorded data are involved in El Niño events and another third are La Niña droughts (Figure 1). Nine months of the year have coefficients of variation for rainfall that exceed 100% (November through July), leaving three months that are predictably dry, at least in the coastal regions (August, September and October, Table A1). The "normal" pattern may simply be that rainfall can occur from January through April in "non-Niño" years and that the misty precipitation of the garúa season will provide little moisture in July through October. The months of November, December, and May to July usually experience little rainfall or garúa except in El Niño years, when they can have from 150 to 660 mm of rain (Table A1).
Figure 3. Ts anomaly and accumulated rainfall during El Niño events in Galápagos.
The frequency of strong oscillations between wet and dry periods (El Niño — La Niña cycles) during the last 34 years is roughly once in seven years. Including weaker oscillations (1965, 1972-73) drops the frequency to once every five years. The periodicity of oscillations appears relatively similar through the 34 years, but the tendency for the oscillations to be extreme is increasing significantly. The annual totals of rainfall in the second half of the period are significantly more variable than those of the first half (1982-98 standard deviation = 757.1, 1965-1981 standard deviation = 219.0, p = 0.02, Levene's Test). This follows a pattern of increasingly strong El Niño events observed across the tropical Pacific region since 1975 (Wolter & Timlin 1998).
Is this increase in Galápagos part of a long-term trend? As of 1972, the conditions in Galápagos were apparently as wet as any period in the preceding 50,000 years and were relatively consistent for the last 3,000 years (Colinvaux 1972, reported in Grant 1985). To examine trends over a shorter period in more detail, we have constructed an estimated history of oscillations in the Galápagos over the last 118 years (Figure 4). This estimated history is based on a record of sea-surface temperatures for the tropical Pacific (Slutz et al. 1985) and compared to a record established from sediments deposited in a montane lake of southern Ecuador (Robell et al. 1999). We estimated the monthly rainfall of Galápagos from a polynomial regression (r = .69, p <0.001) and summed the monthly values into an annual total. The estimated total annual rainfall agrees well with the observed values over the last 34 years (Figure 5).
Figure 4. Estimated pattern of Galápagos rainfall between 1880 and 1998. Squares and triangles indicate, respectively, moderate and severe El Niño events, as determined from historical records. Plus signs indicate El Niño events as determined by a study of organic deposits in an Andean lake.
The frequency of oscillations is relatively constant during the last 118 years and appears similar to the current frequency (12 oscillations in 67 years with sufficient data yields an oscillation every 5.6 years). If we divide the history into three periods of past (1880-1964), recent (1965-1981), and current (1982-1998), we can examine the possible trend. The current total annual rainfall is greater than the recent and past amounts, and the recent and past do not differ significantly (Tukey's HSD: current p <0.0001, recent and past p = 0.325). Variability increases significantly from the past through the recent to the present periods (past standard deviation = 129.2, recent and current values presented above; Levene's Test, p < 0.02). Thus, for the last 118 years, it appears that the amounts of rainfall and the strength of El Niño events remained relatively constant until the current period, when the amounts of rain during El Niño events increased. If these estimated patterns reflect the real trends, the Galápagos Archipelago could be entering one of the wettest periods in its history.
Figure 5. Predictability of annual Galápagos rainfall.
Biological consequences
Many species associated with the marine ecosystem of Galápagos greatly declined in numbers during the 1997-98 El Niño (details of many of the following examples are reported in other papers of this volume). Dead sea lions and marine iguanas became common sights throughout the islands and their numbers may have dropped to less than half of the totals present prior to 1997. Galápagos penguins declined to levels similar to those seen after the 1982-83 El Niño and standardized counts made during 1998 recorded the fewest penguins ever seen. Many other species of seabirds failed to reproduce and may have experienced increased mortality of adults.
Some terrestrial organisms may have responded differently than their marine counterparts, although others appeared to suffer. The population of dark-billed cuckoos on Santa Cruz apparently exploded and may have peaked in July and August of 1998. The population crashed immediately after, and reports of dead cuckoos in the highlands were frequently received at the offices of the CDRS and the Galápagos National Park Service. Several populations of Galápagos tortoises failed to reproduce under natural conditions. In some cases, it appears that increased vegetation and wet soils dropped the nest temperatures to values lower than the embryos could survive (Steve Earsom, pers. comm.). In other cases, increases in populations of introduced fire ants (Solenopsis) attacked and killed hatchling tortoises within their nests. Mortality of adult tortoises increased several fold on Alcedo Volcano, Isabela Island, where more than thirty individuals were killed in floods or by falling down the freshly cut banks of steep ravines (Cruz Márquez, pers. comm.).
Introduced species
The unique biological diversity of the Galápagos Islands is an evolutionary product of the Archipelago's isolation. Prior to the beginning of human activity within the islands in 1535, the barriers that maintained that isolation were the difficulty that most species faced in crossing at least 1000 km of ocean and the problems associated with colonizing an arid set of sparsely vegetated islands. For the last several hundred years the difficulties of dispersal have been removed by the numerous arrivals of boats and airplanes into the archipelago. Thus some 400 to 500 species of plants, some 25 vertebrates, and unknown hundreds of invertebrates have been introduced to compete with and prey upon the natural flora and fauna of the islands. However, many more species have failed to establish themselves once they reached the islands because of the dry conditions usually found in the Archipelago. During the 1997-98 El Niño, it became obvious that the situation is changing.
At least two species of a class of vertebrates that had previously failed to colonize Galápagos became established. Frogs have successfully inhabited the region of Puerto Villamil on Isabela and Puerto Ayora on Santa Cruz Island. The species in Villamil appears to be a member of the family Leptodactylidae and the species in Santa Cruz is Scinax quinquefasciata, a member of the family Hylidae. These families of frogs contain many species adapted for conditions drier than usually associated with frogs. Whether they will spread to the wetter highlands and what their impacts on native flora and fauna will be is hard to say. What we can say is that the combination of human activity and one of the wettest El Niños of the last 118 years has resulted in what was apparently impossible during the previous millions of years of Galápagos history — the establishment of a population of amphibians.
Introduced organisms also spread to new islands within the Archipelago, apparently as a result of the wet conditions. Smooth-billed anis established breeding populations on Genovesa and Fernandina islands. Within the last hundred years but prior to 1980, there were scattered observations of individual smooth-billed anis in Galápagos. Apparently they were either introduced or reached Galápagos on their own, but failed to colonize. Around 1980, a farmer from the highlands of Santa Cruz introduced them in the hopes that they would eat ticks from his cattle. The population remained sparse until the 1982-83 El Niño, when they increased tremendously in numbers and spread to many new islands. Now they have spread to two islands previously free of alien vertebrates.
The rains of the 1997-98 El Niño stopped in May of 1998. A drought developed in the latter half of 1998 and the first months of 1999. Although several experienced farmers predicted that development and reduced their herds of cattle, many farmers on Santa Cruz found that they had greatly exceeded the normal carrying capacity of their farms by building up their herds during El Niño. When the drought developed, these farmers had more cattle than they could support and a poor market in which to sell them. They petitioned the Galápagos National Park Service for grazing rights within the Park and moved several hundred head onto Park land. The region where the cattle were introduced is within the reduced range of Scalesia and part of the tiny remnant of native highland vegetation left on Santa Cruz. The impact of the cattle is both direct and indirect. In grazing and seeking water and shade, they disturb the small native and endemic species growing under the Scalesia canopy, and they injure the roots of the Scalesia, causing trees to fall and the canopy to open. The openings they produce in the Scalesia stands are rapidly colonized by invasive species of introduced plants, many of which are present as seeds in the manure left by the cattle. The result is an acceleration of the spread of invasive plant species and the loss of native humid zone habitat on Santa Cruz.
Potential consequences of stronger El Niño — La Niña Oscillations
In Galápagos, several marine species appear to be on the edges of their distributions. Sea lions reach their southern extent in the islands and penguins are at their northern limits. For the last 17 years, penguins have been unable to recuperate from the mortality experienced in the 1982-83 El Niño. The developing pattern seems to be one of crashes during El Niño events and slow population growth during the subsequent intervals. The overall trend seems to be a reduction in the population that could certainly lead to extinction. The pattern in sea lions is not as well known over the long term, but observations from 1997-98 indicate that a decline could occur. At present, it is impossible to determine whether these potential declines are purely natural events in response to global climatic patterns or if they are exacerbated by the results of human activity such as introduced species or fishing.
Incidents like the colonization of Galápagos by frogs and the movement of cattle into the Park are going to increase as both the climatic oscillations and levels of human activity increase in the Galápagos. The rate of introductions of plants to the Galápagos has increased steadily over the last twenty years as a result of human activity. If wet periods get wetter and farmers continue to exceed the carrying capacity of their farms for drought periods, the islands will experience increasing colonization and invasion by introduced species finding more suitable habitats. We predict an increasingly damaging synergy between human activity and patterns of climatic change in the short term that will present yet more challenges to the conservation biology of the Galápagos Archipelago.
Literature Cited
Anonymous. 1996. ArcView GIS. Environmental Systems Research Institute, Inc., Redlands, California. 340pp.
Colinvaux, P. A. 1972. Climate and the Galápagos Islands. Nature 240:17-20. (reported in Grant 1985).
Grant, P. R. 1985. Climatic fluctuations on the Galapagos Islands and their influence on Darwin's finches. Ornithological Monographs 56: 471-483.
Kerr, R. A. 1999. Big El Niños ride the back of slower climate change. Science 283:1108-1109.
McPhaden, M. J., A. J. Busalacchi, R. Cheney, J. R. Donguy, K. S. Gage, D. Halpern, M. Ji, P. Julian, G. Meyers, G. T. Mitchum, P. P. Niiler, J. Picaut, R. W. Reynolds, N. Smith, and K. Takeuchi. 1998. The Tropical Ocean-Global Atmosphere observing system: a decade of progress. Journal of Geophysical Research 103 (C7):14,169-14,240.
McPhaden, M. J. 1999. Genesis and evolution of the 1997-98 El Niño. Science 283:950-954.
Robell, D. T., G. O. Seltzer, D. M. Anderson, M. B. Abbot, D. B. Enfield, and J. H. Newman. 1999. An ~ 15,000-year record of El Niño-driven alluviation in southwestern Ecuador. Science 283:516-520.
Robinson, G. and E. M. del Pino, eds. 1985. El Niño en las islas Galápagos: el evento de 1982 — 1983. Fundación Charles Darwin, Quito Ecuador, 534 p.
Slutz, R.J., S.J. Lubker, J.D. Hiscox, S.D. Woodruff, R.L. Jenne, P.M. Steurer, and J.D. Elms. 1985. Comprehensive Ocean-Atmosphere Data Set; Release 1'. Climate Research Program, Boulder, Colorado.
Wolter, K. and M. S. Timlin. 1998. Measuring the strength of ENSO — how does 1997/98 rank? Weather 53: 315-324.
Woodruff, S. D., R. J. Slutz, R. L. Jenne, and P. M. Steurer. 1987. A comprehensive ocean-atmosphere data set. Bulletin American Meteorological Society 68: 1239-1250.
Woodruff, S. D., S. J. Lubker, K. Wolter, S. J. Worley, and J. D. Elms. 1993. Comprehensive Ocean-Atmosphere Data Set (COADS) release 1a: 1980-92. Earth System Monitoring 4: 1-8.
Howard L. Snell, Charles Darwin Research Station, Galápagos, Ecuador.
Solanda Rea, Charles Darwin Research Station, Galápagos, Ecuador.