1: \documentclass[12pt,preprint]{aastex}
2:
3: \newcommand{\etal}{\emph{et al. }}
4:
5: \begin{document}
6:
7: \shorttitle{Transition Dwarf Oxygen Abundances}
8: \title{HII Region Oxygen Abundances in Starbursting Transition Dwarf Galaxies}
9:
10: \author{Kate E. Dellenbusch\altaffilmark{1}\altaffiltext{1}{Visiting Astronomer, Kitt Peak National Observatory,
11: National Optical Astronomy Observatory, NOAO is operated by the Association of Universities for Research
12: in Astronomy, Inc. (AURA), under cooperative agreement with the National Science Foundation}}
13: \affil{Department of Astronomy, University of Wisconsin, 475 N. Charter Street, Madison, WI 53706-1582}
14: \email{dellenb@astro.wisc.edu}
15: \author{John S. Gallagher, III}
16: \affil{Department of Astronomy, University of Wisconsin, 475 N. Charter Street, Madison, WI 53706-1582}
17: \email{jsg@astro.wisc.edu}
18: \and
19: \author{Patricia M. Knezek\altaffilmark{1}}
20: \affil{WIYN Consortium, Inc., 950 North Cherry Avenue, Tucson, AZ 85726-6732}
21: \email{knezek@noao.edu}
22:
23:
24: \begin{abstract}
25:
26: We present empirical HII region oxygen abundances for a sample of low-luminosity starburst galaxies which are in
27: a short lived evolutionary state. All five galaxies are characterized by centrally concentrated star formation,
28: which is embedded in smooth stellar envelopes resembling dE-like systems. The galaxies also have small gas contents
29: with typical M$_{HI}$/L$_B$ $\lesssim$ 0.1 resulting in gas exhaustion timescales less than 1 Gyr, even when molecular
30: gas is considered. We find, compared to other morphologically similar systems, the galaxies of this sample have surprisingly high
31: oxygen abundances with 12 + log(O/H) $\sim$ 9.0. We propose that these objects are a subclass of evolved blue compact dwarfs,
32: which have exhausted most of their gas supply while retaining their metals. We further propose that we are seeing these objects during
33: a short phase in which they are nearing the end of their starburst activity, and could become early-type dwarfs.
34:
35: \end{abstract}
36: \keywords{galaxies: abundances --- galaxies: dwarf --- galaxies: starburst}
37:
38: \section{Introduction}
39:
40:
41: Although dwarf galaxies are the structurally simplest and most numerous type of galaxy in
42: the Universe, the specifics of their formation and evolution remain poorly understood.
43: One of the main features of dwarf galaxy populations is a grand division between the early-type
44: spheroidal systems and late-type irregulars; e.g. van den Bergh (1977) emphasized that dwarf
45: galaxies are typically either very actively forming stars or are inactive, with few dwarfs
46: between these states. This raises the question of whether there is an evolutionary link
47: between the two main dwarf galaxy families: are there processes which drive late-type dwarfs
48: to evolve into early-type systems or are these differences ingrained at the time of formation?
49:
50: The possibility of evolutionary links between early- and late-type dwarf galaxies has been widely
51: discussed, especially in the context of the large population of dE systems found in clusters of
52: galaxies (e.g. Ferguson \& Binggeli 1994; Conselice et al. 2003; Lisker, Grebel \& Binggeli 2006).
53: In their morphological study Sandage \& Hoffman (1991) proposed that some galaxies in the field
54: and Virgo cluster should be considered as ``transition dwarfs'' with evolutionary states between
55: those of the two main dwarf galaxy families. The Knezek, Sembach, \& Gallagher (1999)
56: photometric study of candidate transition dwarfs found that the systems in the Sandage \& Hoffman (1991) sample cover a wide range of
57: evolutionary states; their status as evolutionary bridges thus remains unclear.
58: The situation for the higher luminosity end of the dwarf classification therefore could resemble that for the low luminosity
59: dSph/dI galaxies, which share traits of both faint dwarf galaxy structural classes (Grebel, Gallagher \& Harbeck 1993).
60:
61: In this paper we present HII region oxygen abundances for a sample of moderate luminosity galaxies
62: which may be a ``missing link" in dwarf galaxy evolution. These galaxies seem to be evolving via starburts
63: from a gas-rich, actively star forming state to a gas-poor inactive state. We use the term ``transition" galaxy in the
64: spirit of Sandage \& Hoffman (1991) and Knezek, Sembach, \& Gallagher (1999): transition dwarf
65: galaxies show a mixture of early and late-type characteristics and here, specifically
66: are experiencing star formation, but with very little HI gas (See Table \ref{obsprop}). It is not clear, however,
67: if these ``transition" dwarfs represent an actual evolutionary transition from Im to dE
68: systems, or just one of gas content and level of star formation in the most slowly evolving
69: part of the dE structural branch. For example, an alternate model is for these objects to
70: be highly evolved versions of long lived blue compact dwarfs (BCDs) that eventually may fade to form dE galaxies
71: (e.g. Bothun et al. 1986; Drinkwater \& Hardy 1991; Marlowe et al. 1992).
72:
73:
74: Several characteristics of the sample of galaxies presented here indicate that they are in
75: phases of rapid evolution. (1) They are actively forming stars with star formation rates on the order of
76: a few solar masses per year, but the starburts are fueled by very little HI gas (see Table \ref{obsprop}).
77: Star formation can continue in these galaxies
78: at the current rate for only about 10$^8$ yrs based on HI content. Although these galaxies may
79: have significant amounts of molecular gas, it is unlikely to be enough to significantly increase this gas
80: depletion timescale (e.g. Gordon 1991; Li et al. 1994). (2) The star formation is currently centrally
81: concentrated; much of these galaxies are already stellar fossils. (3) They have smooth outer
82: isophotes, which look more like those from an
83: early rather than late-type galaxy (Dellenbusch, Noble, \& Gallagher, in preperation).
84: In addition, they are fairly isolated, located in loose group
85: environments, and show no signs of a recent merger; any dynamical effects are already well in the past.
86: More details about the structures and star
87: formation rates of this sample will also be given in Knezek et al. (in preparation).
88:
89:
90:
91:
92: \section{Observations and Data Reduction}
93:
94:
95: The data presented in this paper were obtained with the GoldCam spectrograph at the
96: 2.1-m telescope on Kitt Peak, Arizona. The observations were obtained during six nights
97: from 2006 January 31 to February 5. Grating 09 with decker 3 was used. This yielded
98: a wavelength range from about 3500 to 7500 \AA\ with a typical FWHM spectral resolution of 8.1 \AA.
99: We used the blocking filter WG345 to remove possible overlap between orders.
100:
101: Typical exposure times were 30 minutes, except for one night of increased cloud cover,
102: in which case 20 minute exposures were used. For all but one galaxy, six 30 minute exposures
103: were combined for a total integration time of 3 hours. For NGC 3353, due to
104: poor weather conditions and variable cloud cover, only one 20 minute exposure is used in the
105: analysis. The slit was centered on the central HII region of each galaxy.
106: For NGC 3353 a position angle of 45$^{\circ}$ was used; for the other four galaxies,
107: the central HII region is sufficiently round that the position angle was left at 90$^{\circ}$.
108:
109: Spectrophotometric standards from Massey et al. (1988) were observed periodically throughout
110: each night so flux calibration could be performed. HeNeAr comparison lamps were observed after
111: each target exposure to calibrate the wavelength solution.
112:
113:
114: The data were reduced following standard methods in IRAF\footnote{IFAF is distributed by the
115: National Optical Astronomy Observatories, which are operated by the Association of Universities
116: for Research in Astronomy, Inc., under cooperative agreement with the National Science
117: Foundation.}. The bias level was determined and corrected for each frame using the overscan
118: region. Separate zero-second exposures were not used to construct an average bias frame
119: because of large variations in bias level from exposure to exposure. Flat-fielding was
120: accomplished by combining and normalizing a series of quartz-lamp flats. Once standard CCD
121: reductions were complete, including the fixing of bad columns, cosmic rays were removed
122: from the frames using L.A. COSMIC, a rejection routine which utilizes Laplacian edge detection
123: (van Dokkum, 2001). The IRAF package DOSLIT was used to calibrate and extract the spectra. An
124: example of one of the extracted long exposure spectra is shown in Figure~\ref{spectrum}.
125:
126:
127:
128: \section{Results and Analysis}
129:
130: Relative emission-line fluxes were measured from the extracted spectra and corrected for underlying
131: stellar Balmer absorption and interstellar reddening. Underlying stellar Balmer
132: absorption with an equivalent width of 4 \AA\ was assumed based on the strength of the H$\gamma$ absoprtion
133: with an approximate correction for the emission line contribution. Case B recombination with
134: T$_e$=10,000 K is adopted along with the SMC extinction curve from Gordon et al. (2003)
135: to correct for interstellar reddening. We estimate the errors on flux ratios to be $\sim$ 30$\%$.
136:
137: Oxygen abundances were estimated indirectly from our corrected HII region spectra using various
138: abundance diagnostics. We use the `optimized abundance determination' method described
139: in Kewley \& Dopita (2002). This combined method utilizes various indicators to
140: cover a range of abundances. Under this method, the [NII]/[OII] ratio is used to
141: estimate abundances in the high metallicity regime with 12 + log (O/H) $\geq$ 8.6.
142: This estimator is good for all but one of our galaxies (NGC 3353).
143:
144: The R23 oxygen abundance indicator was also used and produced consistent
145: results (see Table \ref{abund}). This method was suggested by Pagel et al. (1979) and has seen wide use.
146: The R23 indicator is double valued, having high- and low-abundance branches, and is therefore
147: a poor estimator of oxygen abundance near the turnover region between the branches. This turnover
148: occurs 8.0 $<$ 12 + log(O/H) $<$ 8.3 (Pilyugin \& Thuan 2005). On the basis of the strength of [NII] emission,
149: we estimate abundances for our galaxies
150: using the upper branch of the R23 indicator from the models of Pilyugin (2000) and Pilyugin \& Thuan
151: (2005). Because NGC 3353 lies near this turnover region, we have a less robust estimate for it's
152: oxygen abundance than for our other galaxies.
153:
154:
155:
156: \section{Discussion}
157:
158:
159: The galaxies presented here have surprisingly high HII region oxygen abundances, as compared to
160: similar luminosity dI and BCD galaxies with equally strong HII emission lines. To examine this further,
161: we compare these galaxies with other low luminosity
162: systems by placing them on a luminosity-metallicity plot, shown in Figure \ref{Lz}.
163: To put these objects in context with other dwarf galaxies, luminosity-metallicity
164: relations for dIrrs (H. Lee et al. 2006) and HII galaxies (J. Lee et al. 2004) are shown as
165: solid and dashed lines respectively. In addition, three other star forming dwarf galaxies are
166: also shown; the SMC, LMC, and NGC 5253. NGC 5253 is included because it is a
167: nearby starbursting
168: dwarf galaxy that exhibits several characteristics similar to the transition dwarfs presented
169: here, indicating it may be the same class of object.
170:
171: The transition dwarfs studied here have high oxygen abundances and lie well above the L-Z$_O$
172: relations for most other types of small starburst galaxies. In other words, they are more metal rich than their
173: similar luminosity HII/BCD/dI counterparts. It should be noted that there is typically a significant
174: amount of scatter of about 0.3 dex (e.g. Lee et al. 2006) in the plots from which these mean relations
175: are drawn. Nevertheless, the galaxies
176: in our sample clearly do not follow the same L-Z$_O$ relation as most other dwarf
177: galaxies with strong HII emission. This indicates a difference in the past evolution of the galaxies in our sample,
178: which could be connected to their low gas content and mixed morphologies. Such high abundances in dwarf elliptical
179: galaxies is not unprecedented, however. Gu et al. (2006) recently found the dE galaxy IC 225 to have a blue core and
180: an oxygen abundance of 12 + log(O/H) = 8.98. We also observed this galaxy and found an oxygen abundance consistent
181: with their result.
182:
183: Because approximate oxygen abundance and luminosity are easily obtained, L-Z$_O$ scaling relations
184: have been derived for large samples of galaxies (e.g. Lee et al. 2004; Tremonti et al. 2004).
185: The underlying physical processes that lead to these relations however, are more closely related
186: to stellar mass and metallicity than to luminosity. To examine this more direct physical relation we estimate
187: stellar masses from archival (2MASS) K-band magnitudes\footnote{This
188: publication makes use of data products from the Two Micron All Sky Survey, which is a joint project
189: of the University of Massachusetts and the Infrared Processing and Analysis Center/California
190: Institute of Technology, funded by the National Aeronautics and Space Administration and the
191: National Science Foundation.}. Our stellar estimates assume a stellar M/L
192: ratio calculated from the models of Bell \& De Jong (2001). We utilize the model coefficients
193: for B-R colors, assuming a closed box chemical evolution model and Salpeter IMF. We find
194: the galaxies to have log(M$_*$/M$_\odot$) $\sim$ 10. Metallicity vs. stellar mass is plotted in
195: Figure \ref{Mz}. Although our galaxies again primarily lie above the M$_{\ast}$ - Z$_O$ relation for dwarf
196: galaxies (dashed line - e.g. Lee et al. 2006), they do scatter around the mean relation found by Tremonti et al.
197: (2004) - solid line for star forming SDSS galaxies. This is explained perhaps in part by the fact that the galaxies
198: presented here are at the more massive end of the dwarf galaxy spectrum,and therefore in terms of mass, are
199: more like the Tremonti et al.(2004) sample than the Lee et al.(2006) sample. These combined results indicate
200: that although these systems are unlike other emission line galaxies in L-Z$_O$ space, they are like typical galaxies
201: of similar stellar mass.
202:
203: Can we place these objects in an evolutionary context? An important characteristic that defines these
204: galaxies is a dense concentration of ISM at their centers.
205: It is not obvious how galaxies come to have centrally concentrated ISM, but it is clearly common
206: in the Universe as BCD's routinely share this characteristic (e.g. Simpson \& Gottesman 2000; Roye \& Hunter 2000).
207: Is there an evolutionary path that leads to this
208: centralized star formation or are some galaxies born with a centralized gas concentration? If these
209: galaxies began with extended gas disks, like a dI, then a trigger would be needed to explain the gas
210: being accreted into the center (Tajiri \& Kamaya, 2002). Interactions would provide such a trigger if the gas loses angular
211: momentum and is therefore able to accrete into the galaxy centers fueling the central starburst. Alternatively, an
212: instability associated with low angular momentum could be at work in these galaxies, as suggested by Hoffman et al.
213: (2003) for Virgo cluster BCDs (see also van Zee et al. 2001).
214:
215: We postulate that the galaxies presented here are a subclass of low gas-content BCD's. We further suggest
216: that we may be seeing BCD's that have retained their metals and are running out of gas and nearing the end of
217: their starburst phase. This is a short lived, as evidenced by the fact that more galaxies of this type are not
218: observed, but important evolutionary phase. The end result could be a dE-like system with a metal-rich, younger
219: stellar core, perhaps related to the objects found in the Virgo Cluster by Lisker et al. (in press).
220:
221:
222:
223: \acknowledgments
224:
225: We would like to thank P. Mucciarelli for assisting with the observations.
226: This research is supported in part by NSF grant AST 98-03018 and the University of Wisconsin Graduate
227: School. JSG also thanks Polichronis Papaderos and Thorsten Lisker for useful discussions of dwarf
228: galaxy evolutionary processes.
229:
230:
231: \begin{thebibliography}{}
232:
233: \bibitem[Bell \& de Jong (2001)]{BdJ01}
234: Bell, E. F. \& de Jong, R. S. \ 2001, \apj, 550, 212
235:
236: \bibitem[Bothun et al. (1986)]{BMCM86}
237: Bothun, G. D., Mould, J.R., Caldwell, N., \& MacGillvray, H. T. \ 1986, \aj, 92, 1007
238:
239: \bibitem[Conselice et al. (2003)]{COGW03}
240: Conselice, C. J., O'Neil, K., Gallagher, J. S., Wyse, R. F. G. \ 2003, \apj, 591, 167
241:
242: \bibitem[Dellenbusch et al. (2007)]{DNG07}
243: Dellenbusch, K. E., Noble, A., \& Gallagher, J. S., \ in prep.
244:
245: \bibitem[Drinkwater \& Hardy 1991]{DrinkHard91}
246: Drinkwater, M. \& Hardy, E. 1991, AJ, 101, 94
247:
248: \bibitem[Ferguson \& Binggeli (1994)]{FB94}
249: Ferguson, H. C. \& Binggeli, B. \ 1994, \aapr, 6, 67
250:
251: \bibitem[Gordon (1991)]{Gordon91}
252: Gordon, M. A. \ 1991, \apj, 371, 563
253:
254: \bibitem[Gordon et al. (2003)]{Gordon03}
255: Gordon, K. D., Clayton, G. C., Misselt, K. A., Landolt, A. U., \& Wolff, M. J. \ 2003,
256: \apj, 594, 273
257:
258: \bibitem[Grebel et al. (2003)]{Grebel03}
259: Grebel, E. K., Gallagher, III, J. S., \& Harbeck, D. \ 2003, \aj, 125, 1926
260:
261: \bibitem[Gu et al. (2006)]{Gu06}
262: Gu, Q., Zhao, Y., Shi, L., Peng, Z., \& Luo, X. \ 2006, \aj, 131, 806
263:
264: \bibitem[Hoffman et al. (2003)]{Hoffman03}
265: Hoffman, G. L., Brosch, H., Salpeter, E. E., \& Carle, N. J. \ 2003, \aj, 126, 2774
266:
267: \bibitem[Kewley \& Dopita (2002)]{KD02}
268: Kewley, L. J. \& Dopita, M. A. \ 2002, \apjs, 142, 35
269:
270: \bibitem[Knezek et al. (2007)]{Knezek07}
271: Knezek, P. M., Dellenbusch, K. E., \& Gallagher, J. S. \ in prep.
272:
273: \bibitem[Knezek et al. (1999)]{Knezek99}
274: Knezek, P. M., Sembach, K. R., \& Gallagher, III, J. S. \ 1999, \apj, 514, 119
275:
276: \bibitem[Lee et al. (2006)]{Lee06}
277: Lee, H., Skillman, E. D., Cannon, J. M, Jackson, D. C., Gehrz, R. D., Polomski, E. F.,
278: Woodward, C. E. \ 2006, \apj, 647, 970
279:
280: \bibitem[Lee et al. (2004)]{LSM04}
281: Lee, J. C., Salzer, J. J., \& Melbourne, J. \ 2004, \apj, 616, 752
282:
283: \bibitem[Li et al. (1994)]{Li94}
284: Li, J. G., Seaquist, E. R., Wang, Z., \& Sage, L. J. \ 1994, \aj, 107, 90
285:
286: \bibitem[Lisker et al. (2006)]{LGWG06}
287: Lisker, T., Glatt, K., Westera, P., \& Grebel, E. K. \ 2006, \aj, in press (astro-ph/0608294)
288:
289: \bibitem[Lisker et al. (2006)]{LGB06}
290: Lisker, T., Grebel, E. K., \& Binggeli, B. \ 2006, \aj, 132, 497
291:
292: \bibitem[Marlowe et al. (1999)]{MMH99}
293: Marlowe, A. T., Meurer, G. R., \& Heckman, T. M. \ 1999, \apj, 552, 183
294:
295: \bibitem[Massey et al. (1988)]{Massey88}
296: Massey, P., Strobel, K., Barnes, J. V., \& Anderson, E. \ 1988, \apj, 328, 315
297:
298: \bibitem[Pagel et al. (1979)]{Pagel79}
299: Pagel, B. E. J., Edmunds, M. G., Blackwell, D. E., Chun, M. S., \& Smith, G. \ 1979,
300: \mnras, 189, 95
301:
302: \bibitem[Pilyugin (2000)]{Pilyugin00}
303: Pilyugin, L. S. \ 2000, \aap, 362, 325
304:
305: \bibitem[Pilyugin \& Thuan (2005)]{PT05}
306: Pilyugin, L. S. \& Thuan, T. X. \ 2005, \apj, 631, 231
307:
308: \bibitem[Roye \& Hunter (2000)]{Roye00}
309: Roye, E. W. \& Hunter, D. A. \ 2000, \aj, 119, 1145
310:
311: \bibitem[Sandage \& Hoffman (1991)]{Sandage91}
312: Sandage, A. \& Hoffman, G. L. \ 1991, \apj, 379, L45
313:
314: \bibitem[Simpson \& Gottesman (2000)]{Simpson00}
315: Simpson, C. E. \& Gottesman, S. T. \ 2000, \aj, 120, 2975
316:
317: \bibitem[Tajiri \& Kamaya (2002)]{TK02}
318: Tajiri, Y. Y. \& Kamaya, H. \ 2002, \aap, 367, 373
319:
320: \bibitem[Tremonti et al. (2004)]{Tremonti04}
321: Tremonti, C. A. et al. \ 2004, \apj, 613, 898
322:
323: \bibitem[van den Bergh (1977)]{vandenBergh77}
324: van den Bergh, S. \ 1977, {\it Evolution of Galaxies and Stellar Populations, Proceeding of
325: a Conference at Yale University}, 19
326:
327: \bibitem[van Dokkum (2001)]{vanDokkum01}
328: van Dokkum, P. G. \ 2001, \pasp, 113, 1420
329:
330: \bibitem[van Zee et al. (2001)]{vanZee01}
331: van Zee, L., Salzer, J. J., \& Skillman, E. D. \ 2001, \aj, 122, 121
332:
333: \end{thebibliography}
334:
335: \clearpage
336:
337: \begin{deluxetable}{lccc}
338: \tablewidth{0pt}
339: \tablecaption{Observed Galaxy Properties \label{obsprop}}
340: \tablehead{
341: \colhead{Galaxy} &
342: \colhead{EW$_{H\beta}$ (\AA)\tablenotemark{a}} &
343: \colhead{SFR (M$_{\odot}$yr$^{-1}$)} &
344: \colhead{M$_{HI}$/L$_B$}
345: }
346: \startdata
347: NGC 3265 & 14 & 2.1 & 0.10 \\
348: NGC 3353 & 23 & 6.0 & 0.30 \\
349: NGC 3928 & 10 & 1.6 & 0.33 \\
350: NGC 3773 & 12 & 2.1 & 0.08 \\
351: IC 745 & 18 & 1.8 & 0.05 \\
352: \enddata
353: \tablenotetext{a}{includes 4\AA\ correction for underlying stellar balmer absorption}
354: \end{deluxetable}
355:
356: \clearpage
357:
358: \begin{deluxetable}{lcccc}
359: \tablewidth{0pt}
360: \tablecaption{Oxygen Abundance Estimators \label{abund}}
361: \tablehead{
362: \colhead{Galaxy} &
363: \colhead{R$_{23}$} &
364: \colhead{12 + log(O/H)\tablenotemark{a}} &
365: \colhead{[NII]/[OII]} &
366: \colhead{12 + log(O/H)\tablenotemark{b}}
367: }
368:
369: \startdata
370: NGC 3265 & 2.0 & 9.1 & 1.5 & 9.2 \\
371: NGC 3353 & 8.8 & 8.2 & 0.7 & 8.4 \\
372: NGC 3928 & 2.0 & 9.1 & 0.8 & 9.1 \\
373: NGC 3773 & 4.9 & 8.5 & 0.2 & 8.8 \\
374: IC 745 & 2.0 & 9.1 & 0.5 & 9.0 \\
375: \enddata
376: \tablenotetext{a}{oxygen abundance estimated from R$_{23}$}
377: \tablenotetext{b}{oxygen abundance estimated from [NII]/[OII]}
378: \end{deluxetable}
379:
380: \clearpage
381:
382: \begin{figure}
383: \plotone{f1.eps}
384: \caption{Spectrum in log Flux of NGC~3265, one of our sample galaxies. Key lines are labeled.\label{spectrum}}
385: \end{figure}
386:
387: \clearpage
388:
389: \begin{figure}
390: \plotone{f2.eps}
391: \caption{Nebular oxygen abundances vs. absolute magnitude in B. Our sample galaxies are shown by filled circles, with
392: three other dwarfs indicated with symbols as described in the key. For a wider comparison, the dashed line shows
393: the best fit for HII galaxies from Lee et al. (2004) and the solid line is the best fit to dIrrs from Lee
394: et al. (2006). Note, all of our galaxies lie above both of these L-z relations. We plot [NII]/[OII] derived
395: abundances for the transition dwarfs. \label{Lz}}
396: \end{figure}
397:
398: \clearpage
399:
400: \begin{figure}
401: \plotone{f3.eps}
402: \caption{Oxygen abundance vs. stellar mass. Our sample galaxies are shown by filled circles. The solid line is
403: the fit found by Tremonti et al. (2004) for over 53,000 star forming SDSS galaxies. For further comparison,
404: the dashed line is an extrapolation of the least-squares fit found by Lee et al. (2006) for dIrrs.
405: Our galaxies are consistent with the Tremonti et al. (2004) fit. We plot [NII]/[OII] derived abundances
406: for the transition dwarfs. \label{Mz}}
407: \end{figure}
408:
409:
410: \end{document}
411: