Climate_NN1 Change_NN1 :_: A_ZZ1 Controlled_JJ Experiment_NN1 Scientists_NN2 have_VH0 carefully_RR manipulated_VVN grasslands_NN2 and_CC forests_NN2 to_TO see_VVI howprecipitation_NN1 ,_, carbon_NN1 dioxide_NN1 and_CC temperature_NN1 changes_NN2 affect_VV0 the_AT biosphere_NN1 ,_, allowing_VVG them_PPHO2 to_TO forecast_VVI the_AT future_NN1 Thirty_MC years_NNT2 ago_RA Charles_NP1 F._NP1 Baes_NP1 ,_, Jr._NNA ,_, a_AT1 chemist_NN1 at_II the_AT U.S._NP1 Department_NN1 of_IO Energy_NN1 's_GE Oak_NN1 Ridge_NN1 National_JJ Laboratory_NN1 ,_, wrote_VVD that_CST the_AT earth_NN1 was_VBDZ undergoing_VVG a_AT1 great_JJ "_" uncontrolled_JJ experiment_NN1 ,_, "_" one_PN1 that_CST would_VM soon_RR reveal_VVI the_AT global_JJ consequences_NN2 of_IO rising_VVG greenhouse_NN1 gas_NN1 concentrations_NN2 ._. 
Today_RT scientists_NN2 know_VV0 that_DD1 deforestation_NN1 ,_, land_NN1 use_NN1 and_CC the_AT burning_JJ of_IO fossil_NN1 fuels_NN2 are_VBR warming_VVG our_APPGE planet_NN1 ._. 
We_PPIS2 are_VBR less_RGR certain_JJ ,_, however_RR ,_, about_II how_RRQ climate_NN1 change_NN1 will_VM alter_VVI forests_NN2 and_CC grasslands_NN2 ,_, as_II31 well_II32 as_II33 the_AT goods_NN2 and_CC services_NN2 these_DD2 ecosystems_NN2 provide_VV0 society_NN1 ._. 
Much_DA1 of_IO the_AT climate_NN1 change_NN1 news_NN1 in_II the_AT mass_JJ media_NN comes_VVZ not_XX from_II experiments_NN2 but_CCB observations_NN2 ._. 
Scientists_NN2 monitor_VV0 Arctic_JJ sea_NN1 ice_NN1 ,_, glaciers_NN2 and_CC natural_JJ events_NN2 such_II21 as_II22 the_AT timing_NN1 of_IO leaf_NN1 appearance_NN1 and_CC inform_VV0 the_AT public_NN1 when_RRQ changes_NN2 fall_VV0 outside_II normal_JJ expectations_NN2 ._. 
Recording_VVG this_DD1 kind_NN1 of_IO information_NN1 over_II time_NNT1 is_VBZ important_JJ ._. 
But_CCB rather_II21 than_II22 waiting_VVG to_TO see_VVI how_RRQ an_AT1 evolving_JJ climate_NN1 slowly_RR alters_VVZ the_AT biosphere_NN1 ,_, climate_NN1 change_NN1 biologists_NN2 are_VBR conducting_VVG field_NN1 experiments_NN2 ,_, often_RR at_II large_JJ scales_NN2 ,_, to_TO see_VVI how_RRQ ecosystems_NN2 will_VM respond_VVI to_II more_DAR or_CC less_DAR precipitation_NN1 ,_, rising_VVG concentrations_NN2 of_IO carbon_NN1 dioxide_NN1 (_( CO2_FO )_) and_CC warming_VVG temperatures_NN2 ._. 
Experimental_JJ data_NN are_VBR key_JJ to_II determining_VVG if_CS and_CC to_II what_DDQ extent_NN1 ecosystems_NN2 will_VM be_VBI affected_VVN by_II climate_NN1 change_NN1 in_II 10_MC ,_, 50_MC or_CC 100_MC years_NNT2 and_CC how_RRQ those_DD2 changes_NN2 might_VM feed_VVI back_RP to_II further_JJR advance_NN1 change_NN1 ._. 
The_AT results_NN2 can_VM help_VVI separate_JJ fact_NN1 from_II fiction_NN1 in_II the_AT climate_NN1 debate_NN1 ,_, which_DDQ is_VBZ charged_VVN with_IW emotion_NN1 ._. 
For_IF years_NNT2 researchers_NN2 investigated_VVD how_RGQ single_JJ plants_NN2 --_NN1 typically_RR grown_VVN for_IF several_DA2 months_NNT2 inside_II climate-controlled_JJ chambers_NN2 --_NN1 responded_VVN to_II varied_JJ conditions_NN2 ._. 
Understanding_VVG mechanisms_NN2 at_II this_DD1 scale_NN1 is_VBZ necessary_JJ ._. 
But_CCB we_PPIS2 must_VM also_RR study_VVI plants_NN2 in_II their_APPGE proper_JJ context_NN1 :_: actual_JJ ecosystems_NN2 ._. 
Largely_RR unknown_JJ to_II the_AT public_NN1 ,_, several_DA2 sizable_JJ outdoor_JJ experiments_NN2 involving_VVG altered_JJ precipitation_NN1 and_CC CO2_FO concentrations_NN2 have_VH0 been_VBN under_RR21 way_RR22 for_IF more_DAR than_CSN a_AT1 decade_NNT1 ,_, including_II those_DD2 that_CST are_VBR described_VVN in_II boxes_NN2 on_II the_AT following_JJ pages_NN2 ._. 
Temperature_NN1 experiments_NN2 have_VH0 begun_VVN as_RR21 well_RR22 ._. 
Enough_DD data_NN have_VH0 now_RT been_VBN generated_VVN to_TO improve_VVI models_NN2 that_CST predict_VV0 climate_NN1 and_CC vegetation_NN1 changes_NN2 ,_, providing_VVG a_AT1 more_RGR accurate_JJ picture_NN1 of_IO how_RRQ woodlands_NN2 ,_, prairies_NN2 and_CC agricultural_JJ crops_NN2 may_VM change_VVI in_II an_AT1 increasingly_RR warmer_JJR world_NN1 that_CST is_VBZ subject_II21 to_II22 different_JJ precipitation_NN1 patterns_NN2 and_CC blanketed_VVD in_II more_DAR CO2_FO ._. 
Issue_VV0 :_: Although_CS temperature_NN1 and_CC CO2_FO levels_NN2 will_VM generally_RR rise_VVI worldwide_RL ,_, climate_NN1 models_NN2 predict_VV0 that_DD1 precipitation_NN1 increases_NN2 and_CC decreases_NN2 will_VM vary_VVI much_DA1 more_DAR from_II place_NN1 to_II place_NN1 in_II coming_VVG decades_NNT2 ._. 
Experiment_NN1 :_: Scientists_NN2 have_VH0 built_VVN a_AT1 variety_NN1 of_IO structures_NN2 to_TO lower_VVI or_CC raise_VVI the_AT amount_NN1 of_IO water_NN1 that_CST reaches_VVZ plants_NN2 in_II grasslands_NN2 ,_, forests_NN2 and_CC croplands_NN2 ,_, as_II31 well_II32 as_II33 the_AT treeless_JJ tundra_NN1 in_II northern_JJ latitudes_NN2 ._. 
Domed_JJ canopies_NN2 or_CC troughs_NN2 are_VBR most_RGT often_RR used_VVN ._. 
The_AT water_NN1 can_VM be_VBI sent_VVN off-site_JJ or_CC redistributed_VVD across_II a_AT1 nearby_JJ site_NN1 to_TO test_VVI greater_JJR precipitation_NN1 there_RL ._. 
Some_DD shelters_NN2 can_VM be_VBI moved_VVN or_CC retracted_VVD ._. 
Barriers_NN2 or_CC trenches_NN2 can_VM be_VBI built_VVN into_II the_AT soil_NN1 to_TO prevent_VVI surface_NN1 water_NN1 from_II creeping_VVG into_II the_AT study_NN1 plots_NN2 and_CC to_TO prevent_VVI plant_NN1 roots_NN2 from_II accessing_VVG water_NN1 outside_II the_AT plots_NN2 ._. 
Projects_NN2 such_II21 as_II22 the_AT Throughfall_NP1 Displacement_NN1 Experiment_NN1 (_( TDE_NP1 )_) near_II Oak_NN1 Ridge_NN1 ,_, Tenn._NP1 ,_, employ_VV0 elaborate_JJ trough_NN1 and_CC gutter_NN1 systems_NN2 in_II the_AT understory_JJ of_IO a_AT1 forest_NN1 to_TO create_VVI dry_JJ and_CC wet_JJ soil_NN1 conditions_NN2 (_( photograph_NN1 and_CC illustration_NN1 )_) ._. 
As_RG many_DA2 as_CSA 1,900_MC troughs_NN2 may_VM be_VBI distributed_VVN across_II zones_NN2 the_AT size_NN1 of_IO football_NN1 fields_NN2 ._. 
Similar_JJ designs_NN2 can_VM be_VBI placed_VVN between_II trees_NN2 that_CST are_VBR widely_RR spaced_VVN ,_, for_REX21 example_REX22 ,_, in_II woodlands_NN2 of_IO New_NP1 Mexico_NP1 ,_, where_CS Nathan_NP1 McDowell_NP1 of_IO Los_NP1 Alamos_NP1 National_JJ Laboratory_NN1 is_VBZ studying_VVG the_AT role_NN1 of_IO drought_NN1 and_CC insects_NN2 in_II tree_NN1 mortality_NN1 ._. 
Results_NN2 :_: On_II the_AT Konza_NP1 Prairie_NN1 ,_, studied_VVN by_II Kansas_NP1 State_NN1 University_NN1 ,_, some_DD grasses_NN2 have_VH0 tolerated_VVN shifting_JJ precipitation_NN1 patterns_NN2 better_RRR than_CSN others_NN2 ._. 
Competition_NN1 between_II plants_NN2 for_IF water_NN1 could_VM rise_VVI in_II a_AT1 warming_JJ world_NN1 ._. 
In_II temperate_JJ forests_NN2 ,_, such_II21 as_II22 the_AT one_PN1 being_VBG studied_VVN by_II Paul_NP1 Hanson_NP1 in_II his_APPGE 13-year-long_JJ TDE_NN1 project_NN1 ,_, mature_JJ trees_NN2 with_IW deep_JJ roots_NN2 have_VH0 withstood_VVN sustained_JJ reductions_NN2 in_II precipitation_NN1 ._. 
But_CCB many_DA2 saplings_NN2 and_CC seedlings_NN2 with_IW shallow_JJ roots_NN2 have_VH0 died_VVN ._. 
This_DD1 experiment_NN1 has_VHZ also_RR revealed_VVN that_CST trees_NN2 are_VBR harmed_VVN more_RRR by_II excessively_RR dry_JJ intervals_NN2 during_II certain_JJ seasons_NNT2 ;_; reduced_JJ rainfall_NN1 during_II active_JJ stem_NN1 expansion_NN1 in_II early_JJ spring_NN1 retarded_JJ growth_NN1 more_RGR dramatically_RR than_CSN reduced_JJ rainfall_NN1 did_VDD at_II other_JJ times_NNT2 ._. 
Late-season_JJ droughts_NN2 occurring_VVG after_II trees_NN2 had_VHD stopped_VVN growing_JJ were_VBDR of_IO little_JJ consequence_NN1 ,_, as_CS31 long_CS32 as_CS33 soil-water_JJ supplies_NN2 were_VBDR fully_RR recharged_VVN before_II the_AT subsequent_JJ growing_JJ season_NNT1 ._. 
In_II contrast_NN1 ,_, some_DD large_JJ trees_NN2 in_II the_AT Amazon_NP1 rain_NN1 forests_NN2 of_IO Brazil_NP1 perished_VVD during_II the_AT fourth_MD year_NNT1 of_IO a_AT1 drought_NN1 created_VVN by_II researchers_NN2 from_II the_AT Woods_NN2 Hole_NN1 Research_NN1 Center_NN1 in_II Massachusetts_NP1 ._. 
Yet_RR saplings_NN2 and_CC small-diameter_JJ trees_NN2 were_VBDR less_RRR affected_VVN ._. 
Blocking_VVG 60_MC percent_NNU of_IO the_AT rainfall_NN1 dried_VVD deep_JJ soils_NN2 ,_, but_CCB surface_NN1 soils_NN2 remained_VVD fairly_RR moist_JJ ,_, the_AT opposite_NN1 of_IO TDE_NN1 results_NN2 ._. 
It_PPH1 is_VBZ clear_JJ that_CST complex_JJ interactions_NN2 must_VM be_VBI understood_VVN before_II models_NN2 can_VM faithfully_RR represent_VVI climate_NN1 change_NN1 effects_NN2 ._. 
Issue_VV0 :_: Scientists_NN2 estimate_VV0 that_CST the_AT oceans_NN2 and_CC terrestrial_JJ ecosystems_NN2 soak_VV0 up_RP at_RR21 least_RR22 half_DB of_IO the_AT CO2_FO released_VVN by_II burning_JJ fossil_NN1 fuels_NN2 ._. 
Plants_NN2 do_VD0 it_PPH1 by_II using_VVG the_AT gas_NN1 to_TO produce_VVI carbohydrates_NN2 during_II photosynthesis_NN1 ._. 
But_CCB will_VM that_DD1 conversion_NN1 continue_VVI at_II higher_JJR CO2_FO concentrations_NN2 ?_? 
And_CC will_VM more_RRR CO2_FO alter_VV0 the_AT sugars_NN2 ,_, carbohydrates_NN2 and_CC protective_JJ compounds_NN2 in_II plants_NN2 ,_, in_II turn_NN1 helping_NN1 or_CC hindering_VVG insects_NN2 and_CC pathogens_NN2 ?_? 
Experiment_NN1 :_: The_AT Free-Air_NP1 CO2_FO Enrichment_NP1 (_( FACE_NN1 )_) experiment_NN1 has_VHZ been_VBN running_VVG at_II Oak_NN1 Ridge_NN1 National_JJ Laboratory_NN1 for_IF more_DAR than_CSN 10_MC years_NNT2 ,_, under_II Richard_NP1 Norby_NP1 ._. 
It_PPH1 consists_VVZ of_IO four_MC study_NN1 areas_NN2 ringed_VVD by_II vent_NN1 pipes_NN2 suspended_VVN from_II towers_NN2 (_( photograph_NN1 and_CC illustration_NN1 )_) ._. 
The_AT pipes_NN2 release_VV0 CO2_FO such_CS21 that_CS22 the_AT trees_NN2 all_DB receive_VV0 a_AT1 selected_JJ amount_NN1 ._. 
Similar_JJ FACE_NN1 experiments_NN2 operate_VV0 at_II almost_RR 35_MC other_JJ natural_JJ and_CC managed_VVD ecosystems_NN2 worldwide_RL ,_, ranging_VVG from_II one-meter-diameter_JJ plots_NN2 in_II bogs_NN2 ,_, 23-meter_JJ plots_NN2 on_II croplands_NN2 and_CC 30-meter_JJ plots_NN2 on_II forest_NN1 plantations_NN2 ._. 
Results_NN2 :_: The_AT data_NN confirm_VV0 that_CST higher_JJR CO2_FO levels_NN2 stimulate_VV0 photosynthesis_NN1 ,_, which_DDQ incorporates_VVZ more_DAR carbon_NN1 into_II plant_NN1 tissues_NN2 ._. 
This_DD1 net_JJ primary_JJ production_NN1 (_( NPP_NP1 )_) is_VBZ also_RR sustained_VVN over_II multiple_JJ growing_JJ seasons_NNT2 ._. 
NPP_NP1 in_II forest_NN1 experiments_NN2 in_II Wisconsin_NP1 ,_, North_ND1 Carolina_NP1 ,_, Tennessee_NP1 and_CC Italy_NP1 has_VHZ increased_VVN 23_MC percent_NNU annually_RR when_CS CO2_FO is_VBZ raised_VVN from_II the_AT ambient_JJ level_NN1 of_IO 388_MC parts_NN2 per_II million_NNO (_( ppm_NNU )_) to_II 550_MC ppm_NNU --_NN1 a_AT1 level_NN1 that_CST could_VM arise_VVI within_II 100_MC years_NNT2 if_CS nations_NN2 do_VD0 nothing_PN1 to_TO curb_VVI emissions_NN2 ._. 
Recent_JJ modeling_NN1 results_NN2 suggest_VV0 that_CST plants_NN2 will_VM respond_VVI positively_RR to_II elevated_JJ CO2_FO levels_NN2 ,_, although_CS gains_NN2 may_VM be_VBI tempered_VVN where_CS soil_NN1 contains_VVZ insufficient_JJ nutrients_NN2 such_II21 as_II22 nitrogen_NN1 ._. 
Increased_JJ NPP_NP1 has_VHZ been_VBN consistent_JJ across_II the_AT FACE_NN1 plots_NN2 worldwide_RL ._. 
But_CCB NPP_NP1 indicates_VVZ only_RR the_AT amount_NN1 of_IO carbon_NN1 added_VVN to_II a_AT1 plant_NN1 ;_; it_PPH1 does_VDZ not_XX reflect_VVI the_AT long-term_JJ fate_NN1 of_IO that_DD1 carbon_NN1 ._. 
In_II the_AT North_ND1 Carolina_NP1 loblolly_JJ pine_NN1 forest_NN1 ,_, the_AT additional_JJ carbon_NN1 was_VBDZ stored_VVN primarily_RR in_II stems_NN2 and_CC branches_NN2 ,_, where_CS it_PPH1 might_VM persist_VVI for_IF decades_NNT2 ._. 
In_II Tennessee_NP1 's_GE sweetgum_NN1 forest_NN1 ,_, however_RR ,_, the_AT carbon_NN1 appeared_VVD mostly_RR in_II new_JJ ,_, small_JJ roots_NN2 ._. 
Such_DA roots_NN2 are_VBR advantageous_JJ ,_, of_RR21 course_RR22 ,_, but_CCB live_VV0 for_IF only_RR a_AT1 few_DA2 weeks_NNT2 to_II a_AT1 year_NNT1 ;_; much_DA1 of_IO the_AT carbon_NN1 returns_VVZ to_II the_AT atmosphere_NN1 as_CSA microbes_NN2 decompose_VV0 the_AT roots_NN2 ._. 
Scientists_NN2 are_VBR trying_VVG to_TO understand_VVI what_DDQ drives_VVZ carbon_NN1 to_II one_MC1 destination_NN1 or_CC another_DD1 ;_; we_PPIS2 will_VM learn_VVI more_RRR as_II trees_NN2 are_VBR harvested_VVN and_CC soils_NN2 are_VBR excavated_VVN in_II the_AT coming_JJ months_NNT2 at_II various_JJ sites_NN2 ._. 
FACE_VV0 experiments_NN2 have_VH0 already_RR paid_VVN off_RP ._. 
James_NP1 Randerson_NP1 of_IO the_AT University_NN1 of_IO California_NP1 ,_, Irvine_NP1 ,_, and_CC scientists_NN2 at_II Oak_NN1 Ridge_NN1 ,_, the_AT National_JJ Center_NN1 for_IF Atmospheric_JJ Research_NN1 in_II Boulder_NP1 ,_, Colo._NP1 ,_, and_CC other_JJ institutions_NN2 have_VH0 used_VVN the_AT data_NN to_TO evaluate_VVI and_CC improve_VVI the_AT Community_NN1 Climate_NN1 System_NN1 Model_NN1 ,_, which_DDQ simulates_VVZ the_AT physical_JJ ,_, chemical_JJ and_CC biological_JJ processes_NN2 that_CST drive_VV0 the_AT earth_NN1 's_GE climate_NN1 system_NN1 ._. 
Temperature_NN1 :_: Highs_NN2 and_CC Lows_NP2 Issue_VV0 :_: Future_JJ warming_NN1 will_VM vary_VVI by_II geographic_JJ location_NN1 ._. 
By_II 2100_MC North_ND1 America_NP1 will_VM be_VBI 3.8_MC to_II 5.9_MC degrees_NN2 Celsius_NP1 warmer_JJR in_II winter_NNT1 and_CC 2.8_MC to_II 3.3_MC degrees_NN2 C_RG warmer_JJR in_II summer_NNT1 ._. 
The_AT changes_NN2 will_VM affect_VVI plant_NN1 metabolism_NN1 as_II31 well_II32 as_II33 water_NN1 and_CC nutrient_NN1 availability_NN1 in_II soil_NN1 ,_, competition_NN1 among_II plants_NN2 ,_, and_CC the_AT voracity_NN1 of_IO herbivores_NN2 ,_, insects_NN2 and_CC pathogens_NN2 ._. 
Experiment_NN1 :_: Researchers_NN2 have_VH0 tried_VVN various_JJ ways_NN2 to_TO warm_VVI very_RG small_JJ plots_NN2 ,_, including_II infrared-emitting_JJ lamps_NN2 ,_, electrical_JJ heating_NN1 tapes_NN2 in_II the_AT soil_NN1 and_CC open-top_JJ chambers_NN2 --_JJ cylindrical_JJ frames_NN2 wrapped_VVN in_II transparent_JJ plastic_NN1 and_CC fitted_VVN with_IW warmair_NN1 blowers_NN2 ._. 
These_DD2 approaches_NN2 have_VH0 proved_VVN useful_JJ but_CCB also_RR have_VH0 drawbacks_NN2 ._. 
Most_DAT can_VM heat_VVI only_RR a_AT1 small_JJ area_NN1 ,_, and_CC many_DA2 have_VH0 heated_VVN only_JJ parts_NN2 of_IO the_AT ecosystem_NN1 ._. 
Heating_NN1 tapes_NN2 create_VV0 unnatural_JJ hot_JJ spots_NN2 in_II soils_NN2 ._. 
Passively_RR warmed_VVN chambers_NN2 depend_VV0 on_II time_NNT1 of_IO day_NNT1 and_CC season_NNT1 ,_, and_CC they_PPHS2 influence_VV0 rainfall_NN1 ,_, wind_NN1 and_CC sunlight_NN1 in_II ways_NN2 that_CST complicate_VV0 interpretation_NN1 of_IO outcomes_NN2 ._. 
Results_NN2 :_: Arctic_JJ ecosystems_NN2 and_CC the_AT boreal_JJ regions_NN2 to_II their_APPGE immediate_JJ south_ND1 are_VBR particularly_RR vulnerable_JJ to_II temperature_NN1 changes_NN2 ._. 
The_AT International_JJ Tundra_NN1 Experiment_NN1 ,_, led_VVN by_II Greg_NP1 Henry_NP1 of_IO the_AT University_NN1 of_IO British_NP1 Columbia_NP1 ,_, uses_VVZ passive_JJ chambers_NN2 to_TO warm_VVI small_JJ plots_NN2 at_II more_DAR than_CSN a_AT1 dozen_NNO sites_NN2 in_II various_JJ countries_NN2 ._. 
Results_NN2 thus_RR far_RR show_VV0 that_CST a_AT1 one-_NN1 to_II three-degree_JJ C_NP1 increase_NN1 enhances_VVZ growth_NN1 and_CC ground_NN1 cover_NN1 of_IO deciduous_JJ shrubs_NN2 and_CC grasses_NN2 as_CSA compared_VVN with_IW mosses_NN2 and_CC lichens_NN2 ._. 
This_DD1 differential_JJ response_NN1 supports_VVZ the_AT hypothesis_NN1 that_CST warming_NN1 will_VM cause_VVI a_AT1 decline_NN1 in_II biodiversity_NN1 across_II high-latitude_JJ ecosystems_NN2 ._. 
A_ZZ1 shift_VV0 from_II herbaceous_JJ to_II woody_JJ vegetation_NN1 could_VM also_RR raise_VVI the_AT energy_NN1 absorbed_VVN by_II the_AT earth_NN1 versus_II that_DD1 reflected_VVD back_RP into_II space_NN1 ,_, further_RRR increasing_JJ global_JJ temperature_NN1 ._. 
Experiments_NN2 at_II other_JJ latitudes_NN2 are_VBR offering_VVG clues_NN2 about_II local_JJ extinctions_NN2 ,_, range_NN1 migrations_NN2 and_CC altered_JJ species_NN composition_NN1 ._. 
At_II Oak_NN1 Ridge_NN1 National_JJ Laboratory_NN1 ,_, Carla_NP1 Gunderson_NP1 has_VHZ exposed_VVN four_MC species_NN of_IO deciduous_JJ trees_NN2 to_II temperatures_NN2 up_RG21 to_RG22 four_MC degrees_NN2 C_ZZ1 above_II ambient_NN1 (_( photograph_NN1 and_CC illustration_NN1 )_) ._. 
Seedlings_NN2 and_CC saplings_NN2 physiologically_RR adjust_VV0 ,_, and_CC more_RGR often_RR than_CSN not_XX they_PPHS2 show_VV0 enhanced_JJ growth_NN1 ._. 
Trees_NN2 produced_VVD leaves_NN2 six_MC to_II 14_MC days_NNT2 earlier_RRR in_II the_AT spring_NN1 ,_, and_CC they_PPHS2 retained_VVD green_NN1 leaves_NN2 later_RRR into_II autumn_NNT1 ,_, lengthening_VVG the_AT growing_JJ season_NNT1 by_II up_RG21 to_RG22 three_MC weeks_NNT2 ._. 
Anecdotal_JJ evidence_NN1 suggests_VVZ ,_, however_RR ,_, that_DD1 earlier_JJR spring_NN1 growth_NN1 might_VM more_RGR frequently_RR expose_VVI plants_NN2 to_II a_AT1 late_JJ ,_, damaging_JJ frost_NN1 ._. 
Despite_II useful_JJ results_NN2 ,_, it_PPH1 is_VBZ hard_JJ to_TO extrapolate_VVI data_NN from_II small_JJ plots_NN2 to_II ecosystems_NN2 ._. 
New_JJ ways_NN2 to_TO warm_VVI larger_JJR areas_NN2 are_VBR needed_VVN ._. 
Techniques_NN2 have_VH0 primarily_RR been_VBN powered_VVN by_II electricity_NN1 ,_, but_CCB natural_JJ gas_NN1 or_CC geothermal_JJ energy_NN1 may_VM be_VBI better_JJR options_NN2 for_IF remote_JJ locations_NN2 ._. 
imposed_VVN uniformly_RR and_CC that_CST the_AT infrastructure_NN1 is_VBZ robust_JJ enough_RR to_TO last_VVI for_IF years_NNT2 ._. 
Biologists_NN2 must_VM also_RR build_VVI installations_NN2 that_CST not_XX only_RR alter_VV0 CO_FO 2concentration_FO ,_, temperature_NN1 or_CC precipitation_NN1 patterns_NN2 ,_, but_CCB all_DB three_MC factors_NN2 in_II combination_NN1 ._. 
We_PPIS2 have_VH0 so_RG far_RR only_RR scratched_VVD the_AT surface_NN1 ._. 
A_AT1 new_JJ experiment_NN1 near_II Cheyenne_NP1 ,_, Wyo._NP1 ,_, is_VBZ evaluating_VVG how_RRQ plants_NN2 in_II a_AT1 northern_JJ mixed-grass_JJ prairie_NN1 will_VM fare_VVI given_JJ simultaneous_JJ changes_NN2 in_II CO2_FO concentration_NN1 and_CC temperature_NN1 ._. 
In_II the_AT first_MD year_NNT1 of_IO the_AT Prairie_NN1 Heating_NN1 and_CC CO_FO 2_MC Enrichment_NP1 experiment_NN1 ,_, Jack_NP1 Morgan_NP1 of_IO the_AT U.S._NP1 Department_NN1 of_IO Agriculture_NN1 's_GE Agricultural_JJ Research_NN1 Service_NN1 has_VHZ found_VVN indications_NN2 that_CST warming_VVG in_II combination_NN1 with_IW higher_JJR CO_FO 2_MC concentration_NN1 may_VM enhance_VVI the_AT abundance_NN1 of_IO warm-season_JJ grasses_NN2 in_II the_AT Great_JJ Plains_NN2 ,_, at_II the_AT expense_NN1 of_IO cool-season_JJ grasses_NN2 ._. 
How_RRQ best_RRT to_TO manipulate_VVI multiple_JJ factors_NN2 and_CC how_RRQ to_TO account_VVI for_IF such_DA combinations_NN2 ,_, as_CSA well_RR as_CSA possible_JJ feedbacks_NN2 ,_, in_II models_NN2 are_VBR complex_JJ questions_NN2 ._. 
We_PPIS2 will_VM need_VVI experimentally_RR derived_VVN data_NN very_RG soon_RR if_CS we_PPIS2 are_VBR to_TO help_VVI society_NN1 anticipate_VVI ,_, plan_NN1 and_CC adapt_VV0 to_II a_AT1 climate_NN1 that_CST is_VBZ already_RR changing_VVG rapidly_RR ._. 
Although_CS the_AT invasion_NN1 of_IO woody_JJ plants_NN2 in_II world_NN1 grasslands_NN2 over_II the_AT past_JJ 200_MC years_NNT2 has_VHZ resulted_VVN mainly_RR from_II overgrazing_NN1 and_CC from_II fire_NN1 suppression_NN1 ,_, rising_VVG atmospheric_JJ CO_FO 2_MC concentrations_NN2 may_VM be_VBI contributing_VVG to_II the_AT encroachment_NN1 of_IO trees_NN2 and_CC shrubs_NN2 across_II the_AT western_JJ U.S._NP1 Future_NN1 CO_FO 2_MC concentrations_NN2 will_VM affect_VVI plants_NN2 in_II ways_NN2 that_CST could_VM impact_NN1 public_NN1 health_NN1 ,_, including_II greater_JJR production_NN1 of_IO pollens_NN2 that_CST trigger_NN1 allergies_NN2 and_CC greater_JJR growth_NN1 and_CC toxicity_NN1 of_IO poison_NN1 ivy_NN1 and_CC other_JJ invasive_JJ species_NN ._. 
The_AT results_NN2 of_IO large_JJ outdoor_JJ experiments_NN2 are_VBR telling_VVG ,_, but_CCB most_DAT investigations_NN2 have_VH0 been_VBN conducted_VVN at_II middle_JJ latitudes_NN2 and_CC mostly_RR in_II the_AT U.S._NP1 and_CC Europe_NP1 ._. 
New_JJ experiments_NN2 at_II a_AT1 wider_JJR range_NN1 of_IO latitudes_NN2 are_VBR needed_VVN to_TO clearly_RR predict_VVI the_AT response_NN1 of_IO boreal_JJ ,_, tundra_NN1 and_CC tropical_JJ plants_NN2 and_CC ecosystems_NN2 ._. 
Several_DA2 years_NNT2 will_VM be_VBI needed_VVN to_TO prepare_VVI such_DA experiments_NN2 because_CS they_PPHS2 are_VBR likely_JJ to_TO be_VBI scientifically_RR complicated_JJ and_CC located_VVN in_II remote_JJ regions_NN2 ._. 
They_PPHS2 will_VM require_VVI significant_JJ engineering_NN1 to_TO ensure_VVI that_DD1 altered_JJ conditions_NN2 are_VBR Next_MD Generation_NN1 of_IO Elevated_JJ CO_FO 2_MC Experiments_NN2 with_IW Crops_NN2 :_: A_ZZ1 Critical_JJ Investment_NN1 for_IF Feeding_VVG the_AT Future_JJ World_NN1 ._. 
Elizabeth_NP1 A._NP1 Consequences_NN2 of_IO More_RGR Extreme_JJ Precipitation_NN1 Regimes_NN2 for_IF Terrestrial_JJ ._. 
