LENOIR J.*, BERTRAND R., GEGOUT J.C., SVENNING J.C., and DECOCQ G. The 55th Symposium of the IAVS (Mokpo, Korea, 23-28 July 2012) Special Session 3 *jonathan.lenoir@u-picardie.fr Impact of climate change on vegetation in Europe: Greater contemporary-time lags for long-lived organisms and lowland biotas LENOIR J.*, BERTRAND R., GEGOUT J.C., SVENNING J.C., and DECOCQ G. Good afternoon everyone, I’m going to give you a European perspective of the Impact of climate change on vegetation. During my talk I will focus on the concept of contemporary-time lags in the biotic responses of forest plants to climate change

What do we mean by contemporary-time lags? Differences between: What are the observed changes in species distribution or community composition under contemporary climate change What would be the expected changes in species distribution or community composition to perfectly match climatic changes And… Wiens et al. (2010) Implied: the niche conservatism assumption By contemporary-time lag, I mean the differences between observed and expected changes in species distribution or community composition to perfectly match contemporary climate change. Talking about expected changes, those are changes in species distribution or community composition that we expect under the niche conservatism assumption Definitions  Expectations  Observations  Conclusions

What is niche conservatism? Perfect niche conservatism Perfect niche adaptation Colwell & Rangel (2009) Geographical space Ecological space 12°C 13°C 14°C 15°C 11°C 16°C 12°C 13°C 14°C 15°C 11°C 16°C 14°C 15°C 16°C 17°C 13°C 12°C 14°C 15°C 16°C 17°C 13°C 12°C Refering to the duality of the niche concept, there is two exteremes we can expect as temperature increases. First, we can assume perfect niche conservatism which means that species will track their climatic niche by shifting their distributions or else, we can assume perfect niche adaptation which means that species will adapt their climatic niche to the new set of temperature conditions keeping their distributions unchanged Definitions  Expectations  Observations  Conclusions

What do we need to measure potential time lags? Visser & Both (2005) To measure time lags between observed and expected changes in species distribution or community composition we need a yardstick Definitions  Expectations  Observations  Conclusions

Definitions  Expectations  Observations  Conclusions A yardstick for changes in species altitudinal distribution in response to increased temperatures Yardstick: +80 m/decade 1256 m 700 m 1000 m 800 m 900 m 1100 m 600 m + T* Assuming perfect niche conservatism and an adiabatic lapse rate of +0.6°C/100 m, species altitudinal distribution will shift about +80 m/decade 1256 m 700 m 1000 m 800 m 900 m 1100 m 600 m + T* Let’s consider an increase in temperature conditions of 1.05°C in our study region during 1987-2008: +0.48°C/decade 1256 m 700 m 1000 m 800 m 900 m 1100 m 600 m + T* Species niche (T) Species distribution Species occurrence About our expectations, here, I represented a given species and its distribution across the altitudinal gradient. The orange colors show the temperature enveloppe for this specific species. If we consider an increase of 1.05 degrees in our study region between 1987 and 2008 and if we assume both perfect niche conservatism and an adiabatic lapse rate of 0.6 degrees per 100 meters, then this given species should track its temperature enveloppe by shifting 80 meters per decade. This is our expectations in a very simple world and we can use it as a yardstick Definitions  Expectations  Observations  Conclusions

Definitions  Expectations  Observations  Conclusions A yardstick for changes in community composition in response to increased temperatures Temperature trend over time T1 = T0 + 1.05 Warming starts (1987) Let’s consider an increase in temperature conditions of 1.05°C in our study region during 1987-2008 Yardstick: +1.05C Community composition Time (yr) T (°C) T0 T(°C) Species niches (T) Species 100% 0% Full recovery Partial recovery Absence of recovery Assuming perfect niche conservatism Let’s now consider several species and their niche width along the temperature gradient. We can use the composition of a given community at a given location and a given time to infer temperature conditions for that particular location and this particular moment in time. Considering an increase in temperature conditions of 1.05 degrees in the study area between 1987 and 2008, this is what we expect the community will look like under the hypothesis of perfect niche conservatism: species composition should be totally different and reflect the new temperature conditions. Indeed, under the assumption of perfect niche conservatism, temperature conditions estimated from the community composition should perfectly match temperature conditions observed by meteorological stations. If not, species composition may be lagging behind climate change. Therefore our yardstick here is that floristically reconstructed temperature should predict an increase of 1.05 degrees during 1987-2008 Definitions  Expectations  Observations  Conclusions

Definitions  Expectations  Observations  Conclusions Contemporary shifts in plant species optimum elevation across mountain forests in France +37 m/decade*** (n = 115 species) Lenoir et al. (2008) Elevation (m) Probability of presence Yardstick: +80 m/decade +9 m/decade (n = 56 species) Lags Let’s now look at the observations. We looked at altitudinal range shifts of 171 forest plant species in the French mountain forests and we found a significant trend of species shifting their optimum elevation upward. However this trend was not as imporatnt as expected under perfect niche conservatism. Indeed, our yardstick was that species should shift about 80 meters per decade upwards so as to perfectly match increasing temperatures in our study region, but we found that woody species (trees and shrubs) shifted only 9 meters per decade whereas herbaceous species (ferns, forbs, grasses, sedges) shifted only 37 meters per decade. Therefore, there was important lags compared to the expected trends given by our yardstick Definitions  Expectations  Observations  Conclusions

Definitions  Expectations  Observations  Conclusions Contemporary changes in plant community composition across the French forests Bertrand et al. (2011) 1970 1980 1990 2000 2010 7 8 9 10 11 12 13 Highland Lowland Yardstick: +1.05C FrT = +0.02C FrT = +0.54C Lags We also looked at changes in plant community composition within french forest communities. This was the work of Romain Bertarnd’s thesis. Romain looked at these changes in community composition separately for lowland and highland areas using plant community composition to reconstruct the temperature trends over time. Here are the temperature trends given by meteorological stations in France (climatically reconstructed temperatures) before the start of climate warming in 1987. As you can see temperature in lowland areas are higher than in highland areas and quite stable over this time period. If we now look at the temperature trends from floristically reconstructed temperatures, these are also quite stable over the time period anterior to climate warming. Our yardstick said that floristically reconstructed temperatures should increase about 1.05 degrees between 1965-1986 and 1987-2008. However, we found that floristically reconstructed temperatures increased only about 0.54 degrees in the highland areas between the two periods. In lowland areas, the changes in floristically reconstructed temperatures between the two periods was not significant and only reach 0.02 degrees. Therefore, we found a large lag between our observations and the expectations, this lag being three times larger in lowland areas compared to highland areas T(°C) climatically reconstructed (CrT) T(°C) floristically reconstructed (FrT) Definitions  Expectations  Observations  Conclusions

Definitions  Expectations  Observations  Conclusions Take-home message Contemporary-time lags are much greater: For trees and shrubs than for grasses, sedges, forbs, and ferns > In lowland forests than in highland forests > To conclude, contemporary-time lags are much greater for long-lived species like trees and shrubs and also within isolated forest habitats in lowland areas. What are the potential determinants explaining these lags? What are the potential determinants explaining these lags? Definitions  Expectations  Observations  Conclusions

Bullet points for discussion Identifying biotic and abiotic determinants of contemporary-time lags in the biotic responses of forest plants to climate change in Europe: Niche adaptation Life span Local persistence Dispersal ability Habitat fragmentation Spatial closeness between isotherms Velocity of temperature changes Mitigating effects due to other changes in abiotic conditions I have listed here several potential biotic and abiotic determinants to explain these contemporary-time lags in the biotic responses of forest plants to climate change. We can discuss these points each by each. How much time do I have left to discuss these points? Definitions  Expectations  Observations  Conclusions

The 55th Symposium of the IAVS (Mokpo, Korea, 23-28 July 2012) Special Session 3 *jonathan.lenoir@u-picardie.fr Acknowledgements Nobuyuki TANAKA and Ikutaro TSUYAMA for organizing and inviting me to contribute in this Special Session All my co-authors and especially Romain BERTRAND for contributing to this talk All of you for your kind attention I whish to thank Nobuyuki Tanaka and Ikutaro Tsuyama for organizing and inviting me to contribute in this Special Session

The 55th Symposium of the IAVS (Mokpo, Korea, 23-28 July 2012) Special Session 3 *jonathan.lenoir@u-picardie.fr Questions?

Bullet points for discussion Identifying biotic and abiotic determinants of contemporary-time lags in the biotic responses of forest plants to climate change in Europe: Niche adaptation Life span Local persistence Dispersal ability Habitat fragmentation Spatial closeness between isotherms Velocity of temperature changes Mitigating effects due to other changes in abiotic conditions Definitions  Expectations  Observations  Conclusions

Niche conservatism, niche adaptation or something in between? Phenotypic plasticity and micro-evolutionary processes may contribute to niche adaptation thus explaining the lag between our observations and our expectations 12°C 13°C 14°C 15°C 11°C 16°C Geographical space Ecological space 14°C 15°C 16°C 17°C 13°C 12°C Definitions  Expectations  Observations  Conclusions

Definitions  Expectations  Observations  Conclusions Life span Devictor et al. (2011) Lagging 212 km behind climate change Lagging 135 km behind climate change Long-lived species reaching maturation after a longer time period than short-lived species are more likely to lag behind climate change Definitions  Expectations  Observations  Conclusions

Adult survival & clonal reproduction Germination & establishment Local persistence Seed mortality Juvenile survival Growth & maturation Seed production Adult survival & clonal reproduction Seed banking Germination & establishment Adult mortality Juvenile mortality Local demography Seed dispersal No dispersal limitations Full dispersal limitations Dullinger et al. (2012) Niche model Occupied area in the Alps (km2) Year High DD parameters Low DD parameters DD: Demography & Dispersal Hybrid model Definitions  Expectations  Observations  Conclusions

Definitions  Expectations  Observations  Conclusions Dispersal ability Many forest plant species are limited by their poor dispersal abilities which may contribute to the observed lags in the biotic responses of forest plants to climate change Definitions  Expectations  Observations  Conclusions

Habitat fragmentation Fragmentation of forest habitats is much greater in lowland than in highland areas: Bertrand et al. (2011) 67% of highland areas are covered by forest patches exceeding 5 km2 and being highly connected (PI = 3814) 29% of lowland areas are covered by forest patches exceeding 5 km2 and being poorly connected (PI = 32) Definitions  Expectations  Observations  Conclusions

Spatial closeness between isotherms 1965-1986 1987-2008 To track their climatic niches according to temperature increases between 1965-1986 and 1987-2008 in France, forest plant species should shift their distributions: Bertrand et al. (2011) +1.1 km upward in highland forests +35.6 km northward in lowland forests Definitions  Expectations  Observations  Conclusions

Velocity of climate change Loarie et al. (2009) Velocity of climate change is much greater in lowland areas than in highland areas Definitions  Expectations  Observations  Conclusions

Definitions  Expectations  Observations  Conclusions Mitigating effects Crimmins et al. (2011) Increasing precipitations mitigates the effect of temperature increase on water balance: species do not only track temperatures Definitions  Expectations  Observations  Conclusions