Vol. 72: 33–46, 2014 AQUATIC MICROBIAL ECOLOGY Published online March 7 doi: 10.3354/ame01683 Aquat Microb Ecol Differing responses of marine N2 fixers to warming and consequences for future diazotroph community structure Fei-Xue Fu*, Elizabeth Yu, Nathan S. Garcia, Jasmine Gale, Yunsheng Luo, Eric A. Webb, David A. Hutchins Department of Biological Sciences, The University of Southern California, 3616 Trousdale Parkway, Los Angeles, California 90089, USA ABSTRACT: The globally distributed colonial cyanobacterium Trichodesmium and unicellular diazotrophs including Crocosphaera together carry out the majority of marine biological nitrogen (N2) fixation. Future sea surface warming is predicted to influence their abundance and distribu- tion, but temperature reaction norms have been determined for very few representatives of each genus. We compared thermal responses within and between the 2 genera Trichodesmium and Crocosphaera by measuring reaction norms for growth, N2 fixation, carbon fixation, and elemental ratios in 7 strains from a global culture collection. Temperature reaction norms of Trichodesmium and Crocosphaera were remarkably similar for all isolates within each genus, regardless of their geographic origin. Thermal limits of Trichodesmium and Crocosphaera ranged from 18 to 32°C and 24 to 32°C, and optimum growth temperatures were ~26 and ~30°C, respectively. The highest cellular ratios of nitrogen to phosphorus and carbon to nitrogen were found at optimum growth temperatures, and the lowest ratios near their thermal limits. In a mixed competition experiment, Trichodesmium growth rates were ~25% higher than those of Crocosphaera at 24°C, while those of Crocosphaera were ~50% higher at 28°C. Comparison of these results to current and projected seasonal temperature regimes in the subtropical Atlantic and Pacific Oceans suggests that pre- dicted warmer temperatures may favor Crocosphaera over Trichodesmium, but that both genera may be excluded where future temperatures consistently exceed 32°C. Sea surface warming could profoundly alter the community structure and stoichiometry of marine N2-fixing cyanobac- teria, thus fundamentally changing the biogeochemical cycling of this globally significant source of new nitrogen. KEY WORDS: Global change · Warming · Temperature · Nitrogen fixation · Trichodesmium · Crocosphaera Resale or republication not permitted without written consent of the publisher INTRODUCTION which has increased globally an average of ~0.6°C in the past 100 yr (Hoegh-Guldberg & Bruno 2010). Cli- Climate change is predicted to change the biodi- mate change models predict sea surface temperature versity of marine pelagic ecosystems by altering to continue to rise into the future (IPCC 2007, competitive interactions within the biological com- Domingues et al. 2008). munity (Hutchins et al. 2009, Boyd et al. 2010, Boyd & Although concern about anthropogenic climate Hutchins 2012). One major physical parameter that is change is relatively recent, researchers have long changing due to the greenhouse effects of anthro- been interested in how phytoplankton respond to pogenic CO2 emissions is sea surface temperature, changing temperature, and there have been a num- *Corresponding author: [email protected] © Inter-Research 2014 · www.int-res.com 34 Aquat Microb Ecol 72: 33–46, 2014 ber of classic laboratory culture studies on this topic. Crocosphaera watsonii, UCYN-A, and UCYN-C Eppley et al. (1972) demonstrated that algal growth (Zehr 2011). A variety of studies have demonstrated rates and assimilation rates increased with increas- that these unicellular diazotrophic cyanobacteria, ing temperature. Raven & Geider (1988) reported especially Crocosphaera and UCYN-A, are abundant that temperature influences algal genotypic and phe- and contribute substantial amounts of N in many oli- notypic variations, chemical transformations, nutri- gotrophic regions (Langlois et al. 2008, Kitajima et al. ent transport processes, and enzyme kinetics. A re cent 2009, Moisander et al. 2010). Trichodesmium and all paper compiled a large dataset from the literature on of the unicellular groups play key roles in the nitro- the temperature responses of cultured phytoplank- gen cycle, because together they are responsible for ton, and used these data to model their likely distri- the majority of total marine biological N2 fixation bution changes in a future warmer ocean (Thomas et (Sohm et al. 2011a). Since UCYN-A are currently not al. 2012). Another recent study used a community- available in culture, little is known about their physi- wide collective experimental approach to examine ology other than what has been gleaned from envi- the thermal ranges of 25 eukaryotic and prokaryotic ronmental molecular studies. In contrast, Tricho des - species, and concluded that while their optimum mium and Crocosphaera have been extensively temperatures and thermal niche widths were often studied due to the availability of numerous labora- consistent with prior work, maximum growth rates tory culture strains (Falcón et al. 2005, Hutchins et al. can vary widely between studies (Boyd et al. 2013). 2007, 2013, Fu et al. 2008, Webb et al. 2009). The influence of temperature on phytoplankton According to Galloway & Cowling (2002), these species composition has also been seen in numerous groups together fix an average of 110 Tg N yr−1 in field studies across the globe. Banse (1991) high- marine ecosystems. Karl et al. (2002) reported that lighted the role of temperature as a fundamental con- biological N2 fixation accounts for an average of 100 trol on phytoplankton growth rates in the ocean. One to 200 Tg N yr−1. Westberry & Siegel (2006) estimated of the responses of phytoplankton to climate change that globally Trichodesmium fixes 42 Tg N yr−1 dur- may be shifts in assemblage structure (Falkowski et ing bloom conditions, and 20 Tg N yr−1 during non- al. 1998). One study found that as a result of sea sur- bloom conditions. Crocosphaera watsonii and other face warming in Swedish coastal waters, winters unicellular groups like UNCY-A and UCYN-C to - were warmer, ice cover disappeared at a faster rate gether carry out about as much N2 fixation as Tricho - in the spring, and phytoplankton blooms occurred desmium spp. (reviewed in Sohm et al. 2011a). earlier. Furthermore, organisms such as cyanobac- A number of studies have shown that the nutrients teria and chlorophytes appeared sooner in the year, iron and phosphorus can control N2 fixation and and dominated the community over a longer period growth of these 2 marine N2 fixers (Sañudo-Wilhelmy (Weyhenmeyer 2001). et al. 2001, Berman-Frank et al. 2007, Fu et al. 2007, Experimental evidence also suggests that warming Chappell et al. 2012). Temperature has also long causes changes in phytoplankton floristics. In a Bering been recognized as a major factor that controls Tri- Sea shipboard manipulative experiment, as tempera- chodesmium abundance, and 20°C has been consid- ture increased, the community shifted away from ered the minimum temperature for Trichodesmium to diatoms and toward nanophytoplankton (Hare et al. survive (Carpenter 1983). Sea surface temperature 2007). Similarly, Rose et al. (2009) found experimen- has been used to define the geographic extent of this tal warming of Antarctic waters also resulted in as - genus (Capone & Carpenter 1982) and to predict eco- semblage changes, with the diatom genera Cylindro - system N2-fixation rates (Bissett et al. 1999). Temper- theca and Thalassiosira being dominant at elevated ature is predicted to influence the abundance and temperatures. In a North Atlantic spring bloom ex - global distribution of Trichodesmium and Croco s - periment, coccolithophorids outcompeted diatoms at pha era (Breitbarth et al. 2007, Sohm et al. 2011a). warmer temperatures (Feng et al. 2009). It is evident Based on nifH DNA copy abundance, UCYN-A that particular groups will benefit from in creased cyanobacteria (average 103−105 copies l−1, Church et temperature, but warming is also likely to put other al. 2008, Moisander et al. 2010) have been observed organisms at a competitive disadvantage. at temperatures ranging from 15 to 30°C (Needoba et Several groups of nitrogen (N2)-fixing cyanobac- al. 2007, Church et al. 2008, Langlois et al. 2008). teria are globally distributed in tropical and subtrop- Both Trichodesmium and Crocosphaera may benefit ical oligotrophic oceans. These include the non- from rising sea surface temperatures because their heterocystous filamentous genus Trichodesmium, ranges will expand poleward. In particular, Tricho - and unicellular forms including UCYN-B such as desmium spp. has been predicted to increase its Fu et al.: Marine N2 fixer thermal response curves 35 global distribution by 11%. However, this aug- pare the diversity of their responses both within and mented range in the higher latitudes may be offset by between the 2 genera. We also aim to understand temperature increases at low latitudes that exceed whether individual strains of these organisms can optimal growth limits (Breitbarth et al. 2007). Cur- serve as general representations of the thermal toler- rently, sea surface temperatures in these areas reach ances of their genus or species as a whole, and to pro- a maximum of 24 to 26°C in summer (August− vide a framework for making better predictions of September), which coincides with documented tem- how this primary climate change variable will affect peratures for optimum growth rates in Tricho des mi um the community structure of marine