There is now a substantial number of observational and experimental studies that demonstrate the link between climate and biological or physical processes in ecosystems. The authors of this chapter assembled a database of more than 2,500 studies that address climate and either a physical process (e.g., melting of ice on lakes) or a biological factor (e.g., spring arrival time) of an animal or plant. Most of these studies address experiments that are valuable primarily in helping to understand the biological mechanisms prompting the responses of plants and animals to climate but are not helpful in detecting patterns of change. Many of these studies were conducted over a period of shorter than 10 years. Because at least 10 years of data are needed to show a possible trend, this narrowed the number of studies to approximately 500. Because temperature is the variable that can most reliably be predicted with increasing GHGs, only studies that addressed temperature as the climatic variable were examined, leaving approximately 250 studies. All 250 studies then were examined to determine if they met at least two of the following criteria:

• The authors found a statistically significant correlation between temperature and a species trait (e.g., egg-laying date, location of range boundaries) or physical process.
• The authors found a statistically significant change in the species trait over time.
• The authors found a statistically significant change in temperature over time.

In some cases, the criteria were met by two companion papers rather than a single paper. These criteria narrowed the qualifying studies to 60; seven were companion papers. Among the 60 studies, 16 look at physical processes, 10 examine vegetation changes, eight look at invertebrates, six investigate amphibians and reptiles, 26 examine birds, and one addresses mammals. Some of these studies investigate multiple taxa (e.g., bird and insect) in the same paper. A total of 39 physical processes, 117 plants, 65 insects, 63 amphibians and reptiles, 209 birds, and 10 mammal species were examined in the 43 studies (summarized in Table 5-3). Approximately 39% of these species showed no change. Changes in the other 61% included earlier ice-off and later freeze dates in inland lakes and streams, earlier breeding times, shifting to higher elevations or latitudes, and changes in densities, development, morphologies, and genetics.

Several lines of evidence indicate lengthening of the vegetative growing season by 1.2-3.6 days per decade in the Northern Hemisphere, particularly at higher latitudes where temperature rise also has been greatest. This lengthening involved earlier onset of spring and later onset of fall. Summer photosynthetic activity [based on Normalized Differential Vegetation Index (NDVI) estimates from satellite data] increased from 1981 to 1991 (Myneni et al., 1997), concurrent with an advance (by 7 days) and an increase in amplitude of the annual CO₂ cycle since the 1960s, most intensely during the 1980s (Keeling et al., 1996). Phenological/climate models for Finland indicate an overall increase in growing season length since 1900 (Carter, 1998). These physical measures are in accord with observations on organisms. In controlled, mixed-species gardens across Europe, a lengthening of the growing season by 10.8 days occurred from 1959 to 1993 (Menzel and Fabian, 1999). Likewise, a study of 36 species in the central United States documented advances in flowering dates by an average of 7.3 days from 1936 to 1998 (Bradley et al., 1999).

Responses to increased atmospheric CO₂ have been detected in increased stomatal densities in the leaves of temperate woodland plants (Beerling and Kelly, 1997). Recent changes (over 9- to 30-year periods) in community composition have occurred at protected sites in the lower United States and Alaska, concurrent with local warming trends (Chapin et al., 1995; Brown et al., 1997a; Alward et al., 1999). Results of warming experiments coupled with previous knowledge of species' habitat requirements implicate climate as one factor in these community reorganizations, but additional effects of multiple pressures have led to complex responses that were not always predicted by bioclimatic theory (Schneider and Root, 1996).

Multiple studies of treelines at high latitudes in the northern hemisphere have shown 20th century poleward shifts, often measured as increased growth at northern boundaries and decreased growth at southern boundaries. Interpretation of these trends is not straightforward because most change occurred during the early 20th century warming and the trends have been less pronounced or absent in recent warm decades (Kullman, 1986, 1990; Hamburg and Cogbill, 1988; Innes, 1991; Lescop-Sinclair and Payette, 1995; Jacoby and D'Arrigo, 1995; Briffa et al., 1998). These authors have hypothesized that the general lack of response to recent warming is a result of increases in water stress, severity of insect attack, and UV radiation and trends toward earlier snowmelt or to sunlight becoming a limiting growth factor. In addition, some localities that showed warming and increased growth in the early 20th century have shown cooling and stable growth since the1970s (Kullman, 1991, 1993). In contrast, simple predictions of range shifts have been fulfilled in alpine herbs, which have moved to higher altitudes concurrent with warming in Switzerland (Grabherr et al., 1994), and loss of low-elevation pine forests in Florida as sea-level rise has caused toxic levels of salination near coastal areas (Ross et al., 1994).

Table 5-3: The number of species and processes in each region that were found in each particular study to be significantly associated with regional temperature change. For inclusion in the table, each study had to meet two of the following three criteria: species or processes changing over time; regional temperature changing over time; and significant association between how the temperature and species or processes were changing. The first number indicates the number of species or processes changing in the manner predicted with global warming. The second number is the number of species or processes changing in a manner opposite to that predicted with a warming planet. When considering those species that have shown a change, 80% are changing in the manner expected with global warming, while 20% are changing in the opposite direction. Note that about 61 of all species examined did not show a statistically significant change. References for each cell are located below the table and collated by row number and column number (e.g., references for European birds are under E,5—row E, column 5). "—" indicates that no studies were found for this region and category.
Region	Column 1: Lake and Stream Ice		Column 2: Vegetation		Column 3: Invertebrate		Column 4: Amphibians and Reptiles		Column 5: Birds		Column 6: Mammals
Row A: Africa	—	—	—	—	—	—	—	—	—	—	—	—
Row B: Antarctica	—	—	2	0	—	—	—	—	2	0	—	—
Row C: Asia	3	0	—	—	—	—	—	—	—	—	—	—
Row D: Australia	—	—	—	—	—	—	—	—	—	—	—	—
Row E: Europe	8	0	13	1	46	1	7	0	258	92	7	0
Row F: North America	18	0	32	11	—	—	—	—	17	4	3	0
Row G: Latin America	—	—	—	—	—	—	22	0	15	0	—	—
Total	29	0	47	12	46	1	29	0	292	96	10	0
Notes: B,2. Smith (1994); B,5. Fraser et al. (1992), Cunningham and Moors (1994), Smith et al. (1999); C,1. Magnuson et al. (2000); E,1. Magnuson et al. (2000); E,2. Grabherr et al. (1994), Ross et al. (1994), Hasenauer et al. (1999), Menzel and Fabian (1999); E,3. Fleming and Tatchell (1995), Zhou et al. (1995), de Jong and Brakefield (1998), Rodriguez-Trelles and Rodriguez (1998), Visser et al. (1998), Parmesan et al. (1999), Parmesan et al. (2000); E,4. Beebee (1995), Reading and Clarke (1995), Reading (1998), Sparks (1999); E,5. Jarvinen (1989, 1994), Gatter (1992), Bezzel and Jetz (1995), Mason (1995), Winkel and Hudde (1996, 1997), Crick et al. (1997), Ludwichowski (1997), Forchhammer et al. (1998), McCleery and Parrins (1998), Prop et al. (1998), Visser et al. (1998), Bergmann (1999), Crick and Sparks (1999), Slater (1999), Sparks (1999), Thomas and Lennon (1999); E,6. Post and Stenseth (1999); F,1. Magnuson et al. (2000); F,2. Barber et al. (2000), Bradley et al. (1999); F,5. Bradley et al. (1999), Brown et al. (1999), Dunn and Winkler (1999); F,6. Post and Stenseth (1999); G,4. Pounds et al. (1999); G,5. Pounds et al. (1999).

IPCC Homepage